WEBVTT Kind: captions Language: en 00:00:04.560 --> 00:00:08.150 align:start position:0% so<00:00:05.280> my<00:00:05.460> name<00:00:05.610> is<00:00:05.730> Martin<00:00:06.090> holes<00:00:06.240> Linda<00:00:06.900> and<00:00:07.710> I 00:00:08.150 --> 00:00:08.160 align:start position:0% so my name is Martin holes Linda and I 00:00:08.160 --> 00:00:11.540 align:start position:0% so my name is Martin holes Linda and I work<00:00:08.370> with<00:00:09.240> the<00:00:09.360> security<00:00:10.040> for<00:00:11.040> the<00:00:11.190> etherium 00:00:11.540 --> 00:00:11.550 align:start position:0% work with the security for the etherium 00:00:11.550 --> 00:00:13.910 align:start position:0% work with the security for the etherium Foundation<00:00:12.030> and<00:00:12.300> I<00:00:12.540> also<00:00:12.719> work<00:00:12.990> with<00:00:13.020> a<00:00:13.260> gift 00:00:13.910 --> 00:00:13.920 align:start position:0% Foundation and I also work with a gift 00:00:13.920 --> 00:00:17.740 align:start position:0% Foundation and I also work with a gift client<00:00:14.719> and<00:00:15.719> I'm<00:00:15.990> going<00:00:16.170> to<00:00:16.230> talk<00:00:16.470> today<00:00:16.920> about 00:00:17.740 --> 00:00:17.750 align:start position:0% client and I'm going to talk today about 00:00:17.750 --> 00:00:19.609 align:start position:0% client and I'm going to talk today about basically<00:00:18.750> things<00:00:19.080> that<00:00:19.200> have<00:00:19.440> happened 00:00:19.609 --> 00:00:19.619 align:start position:0% basically things that have happened 00:00:19.619 --> 00:00:22.160 align:start position:0% basically things that have happened during<00:00:20.070> the<00:00:20.340> last<00:00:20.460> three<00:00:21.030> years<00:00:21.060> so<00:00:21.660> I<00:00:21.690> became 00:00:22.160 --> 00:00:22.170 align:start position:0% during the last three years so I became 00:00:22.170 --> 00:00:25.010 align:start position:0% during the last three years so I became the<00:00:22.350> security<00:00:23.010> guy<00:00:23.160> in<00:00:23.580> basically<00:00:24.540> when<00:00:24.780> I<00:00:24.810> was 00:00:25.010 --> 00:00:25.020 align:start position:0% the security guy in basically when I was 00:00:25.020 --> 00:00:27.470 align:start position:0% the security guy in basically when I was Def<00:00:25.290> Con<00:00:25.560> in<00:00:25.740> Shanghai<00:00:26.220> and<00:00:26.940> we're<00:00:27.060> now<00:00:27.180> in<00:00:27.330> all 00:00:27.470 --> 00:00:27.480 align:start position:0% Def Con in Shanghai and we're now in all 00:00:27.480 --> 00:00:29.270 align:start position:0% Def Con in Shanghai and we're now in all second<00:00:27.780> I'm<00:00:27.960> going<00:00:28.110> to<00:00:28.230> talk<00:00:28.410> about<00:00:28.440> the<00:00:28.890> base 00:00:29.270 --> 00:00:29.280 align:start position:0% second I'm going to talk about the base 00:00:29.280 --> 00:00:34.100 align:start position:0% second I'm going to talk about the base layer<00:00:29.550> the<00:00:30.060> infrastructure<00:00:30.660> of<00:00:31.200> ethereum<00:00:33.110> so 00:00:34.100 --> 00:00:34.110 align:start position:0% layer the infrastructure of ethereum so 00:00:34.110 --> 00:00:37.560 align:start position:0% layer the infrastructure of ethereum so what<00:00:34.710> is<00:00:34.860> the<00:00:35.100> theorem<00:00:35.370> infrastructure<00:00:36.090> and 00:00:37.560 --> 00:00:37.570 align:start position:0% what is the theorem infrastructure and 00:00:37.570 --> 00:00:39.430 align:start position:0% what is the theorem infrastructure and [Music] 00:00:39.430 --> 00:00:39.440 align:start position:0% [Music] 00:00:39.440 --> 00:00:43.160 align:start position:0% [Music] the<00:00:40.440> building<00:00:40.949> blocks<00:00:40.980> of<00:00:41.429> ethereum<00:00:42.170> so 00:00:43.160 --> 00:00:43.170 align:start position:0% the building blocks of ethereum so 00:00:43.170 --> 00:00:46.790 align:start position:0% the building blocks of ethereum so etherium<00:00:44.010> is<00:00:44.249> a<00:00:44.760> blockchain<00:00:45.320> and<00:00:46.320> one<00:00:46.559> of<00:00:46.710> the 00:00:46.790 --> 00:00:46.800 align:start position:0% etherium is a blockchain and one of the 00:00:46.800 --> 00:00:49.149 align:start position:0% etherium is a blockchain and one of the things<00:00:47.039> that<00:00:47.339> can<00:00:47.550> be<00:00:47.609> attacked<00:00:48.089> and<00:00:48.480> it's<00:00:48.929> a 00:00:49.149 --> 00:00:49.159 align:start position:0% things that can be attacked and it's a 00:00:49.159 --> 00:00:53.809 align:start position:0% things that can be attacked and it's a vulnerability<00:00:51.170> which<00:00:52.170> can<00:00:52.409> be<00:00:52.559> is<00:00:52.769> the<00:00:53.609> block 00:00:53.809 --> 00:00:53.819 align:start position:0% vulnerability which can be is the block 00:00:53.819 --> 00:00:57.739 align:start position:0% vulnerability which can be is the block processing<00:00:54.420> so<00:00:55.409> a<00:00:55.979> client<00:00:56.550> gets<00:00:56.819> blocks<00:00:57.239> and 00:00:57.739 --> 00:00:57.749 align:start position:0% processing so a client gets blocks and 00:00:57.749 --> 00:01:00.259 align:start position:0% processing so a client gets blocks and the<00:00:58.170> block<00:00:58.499> consists<00:00:59.370> of<00:00:59.519> headers<00:00:59.969> and 00:01:00.259 --> 00:01:00.269 align:start position:0% the block consists of headers and 00:01:00.269 --> 00:01:02.329 align:start position:0% the block consists of headers and transactions<00:01:00.960> and<00:01:01.350> some<00:01:01.589> other<00:01:01.739> things 00:01:02.329 --> 00:01:02.339 align:start position:0% transactions and some other things 00:01:02.339 --> 00:01:05.030 align:start position:0% transactions and some other things and<00:01:02.749> they<00:01:03.749> need<00:01:03.899> to<00:01:04.019> be<00:01:04.140> validated<00:01:04.860> according 00:01:05.030 --> 00:01:05.040 align:start position:0% and they need to be validated according 00:01:05.040 --> 00:01:10.249 align:start position:0% and they need to be validated according to<00:01:05.489> a<00:01:05.519> very<00:01:06.530> particular<00:01:07.530> set<00:01:08.280> of<00:01:08.820> rules<00:01:09.270> it 00:01:10.249 --> 00:01:10.259 align:start position:0% to a very particular set of rules it 00:01:10.259 --> 00:01:12.380 align:start position:0% to a very particular set of rules it can't<00:01:10.530> be<00:01:10.680> too<00:01:10.890> much<00:01:10.920> validated<00:01:11.880> and<00:01:12.060> it<00:01:12.149> can't 00:01:12.380 --> 00:01:12.390 align:start position:0% can't be too much validated and it can't 00:01:12.390 --> 00:01:16.039 align:start position:0% can't be too much validated and it can't be<00:01:12.540> too<00:01:12.719> little<00:01:12.960> validated<00:01:14.719> so<00:01:15.719> there's<00:01:15.869> very 00:01:16.039 --> 00:01:16.049 align:start position:0% be too little validated so there's very 00:01:16.049 --> 00:01:18.410 align:start position:0% be too little validated so there's very intricate<00:01:16.469> rules<00:01:16.649> and<00:01:17.040> it's<00:01:17.939> extremely 00:01:18.410 --> 00:01:18.420 align:start position:0% intricate rules and it's extremely 00:01:18.420 --> 00:01:21.469 align:start position:0% intricate rules and it's extremely consensus<00:01:18.990> critical<00:01:19.500> because<00:01:20.130> if<00:01:20.460> clients<00:01:21.090> do 00:01:21.469 --> 00:01:21.479 align:start position:0% consensus critical because if clients do 00:01:21.479 --> 00:01:22.789 align:start position:0% consensus critical because if clients do it<00:01:21.570> differently<00:01:21.960> we<00:01:22.229> have<00:01:22.469> an<00:01:22.590> immediate 00:01:22.789 --> 00:01:22.799 align:start position:0% it differently we have an immediate 00:01:22.799 --> 00:01:27.080 align:start position:0% it differently we have an immediate chain<00:01:23.250> split<00:01:24.680> there's<00:01:25.680> this<00:01:25.890> EVM<00:01:26.460> which<00:01:26.909> I 00:01:27.080 --> 00:01:27.090 align:start position:0% chain split there's this EVM which I 00:01:27.090 --> 00:01:28.730 align:start position:0% chain split there's this EVM which I think<00:01:27.329> is<00:01:27.450> the<00:01:27.570> one<00:01:27.750> thing<00:01:28.020> that<00:01:28.289> everyone 00:01:28.730 --> 00:01:28.740 align:start position:0% think is the one thing that everyone 00:01:28.740 --> 00:01:31.310 align:start position:0% think is the one thing that everyone knows<00:01:29.130> exists<00:01:29.670> in<00:01:29.880> ethereum<00:01:30.329> and<00:01:30.539> this<00:01:31.170> is 00:01:31.310 --> 00:01:31.320 align:start position:0% knows exists in ethereum and this is 00:01:31.320 --> 00:01:37.510 align:start position:0% knows exists in ethereum and this is where<00:01:31.560> the<00:01:31.770> smart<00:01:31.860> contracts<00:01:33.530> operate<00:01:34.530> and 00:01:37.510 --> 00:01:37.520 align:start position:0% 00:01:37.520 --> 00:01:40.789 align:start position:0% the<00:01:38.520> smart<00:01:39.090> contracts<00:01:39.630> execute<00:01:40.200> on<00:01:40.409> top<00:01:40.619> of 00:01:40.789 --> 00:01:40.799 align:start position:0% the smart contracts execute on top of 00:01:40.799 --> 00:01:42.800 align:start position:0% the smart contracts execute on top of this<00:01:40.950> thing<00:01:41.280> called<00:01:41.579> the<00:01:41.700> state<00:01:42.060> which<00:01:42.570> is 00:01:42.800 --> 00:01:42.810 align:start position:0% this thing called the state which is 00:01:42.810 --> 00:01:46.399 align:start position:0% this thing called the state which is this<00:01:43.500> global<00:01:44.310> state<00:01:44.700> of<00:01:45.000> where<00:01:45.840> everything<00:01:46.109> is 00:01:46.399 --> 00:01:46.409 align:start position:0% this global state of where everything is 00:01:46.409 --> 00:01:48.980 align:start position:0% this global state of where everything is right<00:01:46.649> now<00:01:46.890> and<00:01:47.399> basically<00:01:47.939> they<00:01:48.270> accumulated 00:01:48.980 --> 00:01:48.990 align:start position:0% right now and basically they accumulated 00:01:48.990 --> 00:01:51.560 align:start position:0% right now and basically they accumulated result<00:01:49.530> of<00:01:49.859> all<00:01:50.070> the<00:01:50.280> transactions<00:01:50.969> in<00:01:51.359> the 00:01:51.560 --> 00:01:51.570 align:start position:0% result of all the transactions in the 00:01:51.570 --> 00:01:54.469 align:start position:0% result of all the transactions in the history<00:01:51.960> of<00:01:51.990> ethereum<00:01:53.000> also<00:01:54.000> highly 00:01:54.469 --> 00:01:54.479 align:start position:0% history of ethereum also highly 00:01:54.479 --> 00:01:57.380 align:start position:0% history of ethereum also highly consensus<00:01:54.960> critical<00:01:55.439> and<00:01:56.310> but<00:01:57.090> aside<00:01:57.329> from 00:01:57.380 --> 00:01:57.390 align:start position:0% consensus critical and but aside from 00:01:57.390 --> 00:01:59.510 align:start position:0% consensus critical and but aside from that<00:01:57.570> they<00:01:58.259> can<00:01:58.439> also<00:01:58.560> be<00:01:58.770> denial<00:01:59.100> of<00:01:59.189> service 00:01:59.510 --> 00:01:59.520 align:start position:0% that they can also be denial of service 00:01:59.520 --> 00:02:02.929 align:start position:0% that they can also be denial of service attacks<00:01:59.880> that<00:02:00.299> causes<00:02:00.719> slow<00:02:00.929> blocks<00:02:01.609> and<00:02:02.609> this 00:02:02.929 --> 00:02:02.939 align:start position:0% attacks that causes slow blocks and this 00:02:02.939 --> 00:02:04.069 align:start position:0% attacks that causes slow blocks and this is<00:02:03.060> what<00:02:03.240> happened<00:02:03.600> during<00:02:03.689> the<00:02:03.869> so-called 00:02:04.069 --> 00:02:04.079 align:start position:0% is what happened during the so-called 00:02:04.079 --> 00:02:08.570 align:start position:0% is what happened during the so-called Shanghai<00:02:04.740> attacks<00:02:06.439> we<00:02:07.439> have<00:02:07.580> networking 00:02:08.570 --> 00:02:08.580 align:start position:0% Shanghai attacks we have networking 00:02:08.580 --> 00:02:11.060 align:start position:0% Shanghai attacks we have networking protocols<00:02:09.060> one<00:02:09.690> of<00:02:09.840> them<00:02:10.020> is<00:02:10.350> the<00:02:10.920> eath 00:02:11.060 --> 00:02:11.070 align:start position:0% protocols one of them is the eath 00:02:11.070 --> 00:02:12.890 align:start position:0% protocols one of them is the eath protocol<00:02:11.610> it's<00:02:11.970> not<00:02:12.090> actually<00:02:12.510> a<00:02:12.540> networking 00:02:12.890 --> 00:02:12.900 align:start position:0% protocol it's not actually a networking 00:02:12.900 --> 00:02:16.990 align:start position:0% protocol it's not actually a networking protocol<00:02:13.319> it's<00:02:13.890> it's<00:02:14.280> the<00:02:15.090> ethereum<00:02:15.810> specific 00:02:16.990 --> 00:02:17.000 align:start position:0% protocol it's it's the ethereum specific 00:02:17.000 --> 00:02:18.220 align:start position:0% protocol it's it's the ethereum specific application 00:02:18.220 --> 00:02:18.230 align:start position:0% application 00:02:18.230 --> 00:02:20.770 align:start position:0% application recall<00:02:18.530> so<00:02:19.340> that<00:02:19.580> two<00:02:19.819> nodes<00:02:20.420> can<00:02:20.750> communicate 00:02:20.770 --> 00:02:20.780 align:start position:0% recall so that two nodes can communicate 00:02:20.780 --> 00:02:22.720 align:start position:0% recall so that two nodes can communicate over<00:02:21.409> each<00:02:21.530> other<00:02:21.769> and<00:02:21.980> one<00:02:22.129> can<00:02:22.310> say<00:02:22.459> I<00:02:22.489> need 00:02:22.720 --> 00:02:22.730 align:start position:0% over each other and one can say I need 00:02:22.730 --> 00:02:25.229 align:start position:0% over each other and one can say I need some<00:02:23.150> blocks<00:02:23.420> and<00:02:23.870> they<00:02:24.560> can<00:02:24.769> do 00:02:25.229 --> 00:02:25.239 align:start position:0% some blocks and they can do 00:02:25.239 --> 00:02:31.000 align:start position:0% some blocks and they can do announcements<00:02:26.239> of<00:02:26.390> new<00:02:27.040> books<00:02:28.040> etc<00:02:28.599> and<00:02:30.010> this 00:02:31.000 --> 00:02:31.010 align:start position:0% announcements of new books etc and this 00:02:31.010 --> 00:02:33.190 align:start position:0% announcements of new books etc and this protocol<00:02:31.700> contains<00:02:32.150> the<00:02:32.420> building<00:02:32.870> blocks 00:02:33.190 --> 00:02:33.200 align:start position:0% protocol contains the building blocks 00:02:33.200 --> 00:02:36.300 align:start position:0% protocol contains the building blocks that<00:02:33.620> clients<00:02:34.370> can<00:02:34.519> use<00:02:34.549> for<00:02:35.090> their<00:02:35.209> syncs 00:02:36.300 --> 00:02:36.310 align:start position:0% that clients can use for their syncs 00:02:36.310 --> 00:02:39.449 align:start position:0% that clients can use for their syncs there<00:02:37.310> are<00:02:37.370> neighboring<00:02:37.730> protocols<00:02:38.299> for<00:02:38.660> LS 00:02:39.449 --> 00:02:39.459 align:start position:0% there are neighboring protocols for LS 00:02:39.459 --> 00:02:42.009 align:start position:0% there are neighboring protocols for LS there's<00:02:40.459> like<00:02:40.670> science<00:02:41.060> and<00:02:41.269> parodies<00:02:41.840> own 00:02:42.009 --> 00:02:42.019 align:start position:0% there's like science and parodies own 00:02:42.019 --> 00:02:45.910 align:start position:0% there's like science and parodies own custom<00:02:42.560> warp<00:02:42.769> sync<00:02:43.930> this<00:02:44.930> piece<00:02:45.650> is<00:02:45.890> not 00:02:45.910 --> 00:02:45.920 align:start position:0% custom warp sync this piece is not 00:02:45.920 --> 00:02:47.800 align:start position:0% custom warp sync this piece is not consensus<00:02:46.549> critical<00:02:46.910> per<00:02:47.060> se<00:02:47.239> but<00:02:47.299> of<00:02:47.599> course 00:02:47.800 --> 00:02:47.810 align:start position:0% consensus critical per se but of course 00:02:47.810 --> 00:02:49.449 align:start position:0% consensus critical per se but of course it's<00:02:47.989> very<00:02:48.019> important<00:02:48.860> that<00:02:49.040> it<00:02:49.159> works 00:02:49.449 --> 00:02:49.459 align:start position:0% it's very important that it works 00:02:49.459 --> 00:02:52.449 align:start position:0% it's very important that it works because<00:02:49.970> otherwise<00:02:50.269> we<00:02:50.720> can<00:02:50.870> have<00:02:51.459> problems 00:02:52.449 --> 00:02:52.459 align:start position:0% because otherwise we can have problems 00:02:52.459 --> 00:02:56.559 align:start position:0% because otherwise we can have problems getting<00:02:52.910> no<00:02:53.329> sync<00:02:54.459> and<00:02:55.459> there<00:02:56.060> can<00:02:56.209> be<00:02:56.239> denial 00:02:56.559 --> 00:02:56.569 align:start position:0% getting no sync and there can be denial 00:02:56.569 --> 00:02:58.660 align:start position:0% getting no sync and there can be denial of<00:02:56.660> service<00:02:57.109> and<00:02:57.440> there<00:02:57.590> can<00:02:57.739> be<00:02:57.769> lost<00:02:58.280> blocks 00:02:58.660 --> 00:02:58.670 align:start position:0% of service and there can be lost blocks 00:02:58.670 --> 00:03:01.270 align:start position:0% of service and there can be lost blocks and<00:02:59.000> largely<00:02:59.239> orcs<00:02:59.629> and<00:02:59.989> generally<00:03:00.650> bad 00:03:01.270 --> 00:03:01.280 align:start position:0% and largely orcs and generally bad 00:03:01.280 --> 00:03:04.900 align:start position:0% and largely orcs and generally bad functioning<00:03:02.150> of<00:03:02.269> ethereum<00:03:02.540> and<00:03:03.760> the<00:03:04.760> eath 00:03:04.900 --> 00:03:04.910 align:start position:0% functioning of ethereum and the eath 00:03:04.910 --> 00:03:06.490 align:start position:0% functioning of ethereum and the eath protocol<00:03:05.480> operates<00:03:06.019> on<00:03:06.140> top<00:03:06.290> of<00:03:06.410> the 00:03:06.490 --> 00:03:06.500 align:start position:0% protocol operates on top of the 00:03:06.500 --> 00:03:08.759 align:start position:0% protocol operates on top of the peer-to-peer<00:03:06.769> protocol<00:03:07.459> the<00:03:07.609> p2p<00:03:07.970> protocol 00:03:08.759 --> 00:03:08.769 align:start position:0% peer-to-peer protocol the p2p protocol 00:03:08.769 --> 00:03:12.670 align:start position:0% peer-to-peer protocol the p2p protocol which<00:03:09.769> sits<00:03:10.040> on<00:03:10.129> top<00:03:10.280> of<00:03:10.459> TCP<00:03:10.670> and<00:03:11.329> it<00:03:12.109> has<00:03:12.379> just 00:03:12.670 --> 00:03:12.680 align:start position:0% which sits on top of TCP and it has just 00:03:12.680 --> 00:03:15.640 align:start position:0% which sits on top of TCP and it has just some<00:03:12.799> basic<00:03:13.190> guarantees<00:03:13.790> and<00:03:14.150> identify<00:03:15.019> no 00:03:15.640 --> 00:03:15.650 align:start position:0% some basic guarantees and identify no 00:03:15.650 --> 00:03:19.089 align:start position:0% some basic guarantees and identify no verifies<00:03:16.519> node<00:03:16.790> identities<00:03:17.829> also<00:03:18.829> and<00:03:19.040> not 00:03:19.089 --> 00:03:19.099 align:start position:0% verifies node identities also and not 00:03:19.099 --> 00:03:20.440 align:start position:0% verifies node identities also and not concerns<00:03:19.519> is<00:03:19.640> critical<00:03:20.030> but<00:03:20.180> it's<00:03:20.359> the 00:03:20.440 --> 00:03:20.450 align:start position:0% concerns is critical but it's the 00:03:20.450 --> 00:03:22.180 align:start position:0% concerns is critical but it's the backbone<00:03:20.900> for<00:03:21.170> the<00:03:21.560> upper<00:03:22.010> layer 00:03:22.180 --> 00:03:22.190 align:start position:0% backbone for the upper layer 00:03:22.190 --> 00:03:25.569 align:start position:0% backbone for the upper layer peer-to-peer<00:03:22.930> protocols<00:03:23.930> so<00:03:24.739> if<00:03:24.980> this<00:03:25.190> were 00:03:25.569 --> 00:03:25.579 align:start position:0% peer-to-peer protocols so if this were 00:03:25.579 --> 00:03:27.640 align:start position:0% peer-to-peer protocols so if this were to<00:03:25.790> fail<00:03:26.030> then<00:03:26.540> of<00:03:26.660> oh<00:03:26.870> yeah<00:03:27.109> you'd<00:03:27.349> have<00:03:27.530> the 00:03:27.640 --> 00:03:27.650 align:start position:0% to fail then of oh yeah you'd have the 00:03:27.650 --> 00:03:33.009 align:start position:0% to fail then of oh yeah you'd have the same<00:03:27.829> failures<00:03:28.639> as<00:03:28.849> the<00:03:29.480> each<00:03:29.630> protocol<00:03:32.019> if 00:03:33.009 --> 00:03:33.019 align:start position:0% same failures as the each protocol if 00:03:33.019 --> 00:03:39.479 align:start position:0% same failures as the each protocol if varium<00:03:33.919> peer-to-peer<00:03:36.130> networking<00:03:38.230> obviously 00:03:39.479 --> 00:03:39.489 align:start position:0% varium peer-to-peer networking obviously 00:03:39.489 --> 00:03:43.449 align:start position:0% varium peer-to-peer networking obviously is<00:03:40.489> built<00:03:41.180> upon<00:03:41.450> the<00:03:41.930> premise<00:03:42.349> that<00:03:42.560> I<00:03:42.799> as<00:03:43.430> a 00:03:43.449 --> 00:03:43.459 align:start position:0% is built upon the premise that I as a 00:03:43.459 --> 00:03:46.930 align:start position:0% is built upon the premise that I as a node<00:03:43.849> can<00:03:44.180> find<00:03:44.510> another<00:03:44.930> node<00:03:45.639> on<00:03:46.639> the 00:03:46.930 --> 00:03:46.940 align:start position:0% node can find another node on the 00:03:46.940 --> 00:03:49.449 align:start position:0% node can find another node on the Internet<00:03:47.419> which<00:03:47.750> is<00:03:47.989> not<00:03:48.260> a<00:03:48.290> trivial<00:03:48.620> task<00:03:49.040> but 00:03:49.449 --> 00:03:49.459 align:start position:0% Internet which is not a trivial task but 00:03:49.459 --> 00:03:51.370 align:start position:0% Internet which is not a trivial task but for<00:03:49.700> that<00:03:49.849> we<00:03:50.060> have<00:03:50.209> the<00:03:50.359> discovery<00:03:50.870> protocol 00:03:51.370 --> 00:03:51.380 align:start position:0% for that we have the discovery protocol 00:03:51.380 --> 00:03:54.460 align:start position:0% for that we have the discovery protocol which<00:03:51.980> currently<00:03:52.430> is<00:03:52.700> in<00:03:52.730> version<00:03:53.389> 4<00:03:53.660> there<00:03:54.440> is 00:03:54.460 --> 00:03:54.470 align:start position:0% which currently is in version 4 there is 00:03:54.470 --> 00:03:58.990 align:start position:0% which currently is in version 4 there is a<00:03:54.560> version<00:03:54.950> 5<00:03:55.310> being<00:03:55.579> worked<00:03:55.790> at<00:03:57.790> and<00:03:58.790> this 00:03:58.990 --> 00:03:59.000 align:start position:0% a version 5 being worked at and this 00:03:59.000 --> 00:04:02.349 align:start position:0% a version 5 being worked at and this uses<00:03:59.329> academia<00:04:00.250> distributed<00:04:01.250> hash<00:04:01.430> table<00:04:01.910> to 00:04:02.349 --> 00:04:02.359 align:start position:0% uses academia distributed hash table to 00:04:02.359 --> 00:04:07.839 align:start position:0% uses academia distributed hash table to announce<00:04:02.810> presence<00:04:03.530> and<00:04:04.599> using<00:04:05.599> this<00:04:06.430> DHT<00:04:07.430> we 00:04:07.839 --> 00:04:07.849 align:start position:0% announce presence and using this DHT we 00:04:07.849 --> 00:04:10.180 align:start position:0% announce presence and using this DHT we can<00:04:08.030> find<00:04:08.299> nodes<00:04:08.660> according<00:04:09.590> to<00:04:09.739> a<00:04:09.769> distance 00:04:10.180 --> 00:04:10.190 align:start position:0% can find nodes according to a distance 00:04:10.190 --> 00:04:12.069 align:start position:0% can find nodes according to a distance metric<00:04:10.430> and<00:04:11.060> they<00:04:11.299> can<00:04:11.480> hone<00:04:11.660> in<00:04:11.810> on<00:04:11.930> each 00:04:12.069 --> 00:04:12.079 align:start position:0% metric and they can hone in on each 00:04:12.079 --> 00:04:13.720 align:start position:0% metric and they can hone in on each other<00:04:12.230> and<00:04:12.500> the<00:04:12.590> distance<00:04:13.010> does<00:04:13.190> has<00:04:13.400> nothing 00:04:13.720 --> 00:04:13.730 align:start position:0% other and the distance does has nothing 00:04:13.730 --> 00:04:15.490 align:start position:0% other and the distance does has nothing to<00:04:13.760> do<00:04:13.940> with<00:04:14.060> geographic<00:04:14.780> location<00:04:15.349> or 00:04:15.490 --> 00:04:15.500 align:start position:0% to do with geographic location or 00:04:15.500 --> 00:04:19.750 align:start position:0% to do with geographic location or network<00:04:16.430> distance<00:04:17.829> the<00:04:18.829> discovery<00:04:19.250> protocol 00:04:19.750 --> 00:04:19.760 align:start position:0% network distance the discovery protocol 00:04:19.760 --> 00:04:23.469 align:start position:0% network distance the discovery protocol is<00:04:20.150> not<00:04:20.449> based<00:04:20.660> on<00:04:20.720> these<00:04:20.900> p<00:04:21.109> but<00:04:21.289> UDP<00:04:22.330> so<00:04:23.330> it 00:04:23.469 --> 00:04:23.479 align:start position:0% is not based on these p but UDP so it 00:04:23.479 --> 00:04:25.810 align:start position:0% is not based on these p but UDP so it has<00:04:23.630> no<00:04:23.900> built-in<00:04:24.470> guarantees<00:04:25.070> from<00:04:25.760> the 00:04:25.810 --> 00:04:25.820 align:start position:0% has no built-in guarantees from the 00:04:25.820 --> 00:04:28.210 align:start position:0% has no built-in guarantees from the lower<00:04:26.090> layers<00:04:26.450> about 00:04:28.210 --> 00:04:28.220 align:start position:0% lower layers about 00:04:28.220 --> 00:04:30.830 align:start position:0% lower layers about for<00:04:29.220> spoofing<00:04:29.610> protection<00:04:30.120> it's<00:04:30.300> trivial<00:04:30.810> to 00:04:30.830 --> 00:04:30.840 align:start position:0% for spoofing protection it's trivial to 00:04:30.840 --> 00:04:38.740 align:start position:0% for spoofing protection it's trivial to spoof<00:04:31.169> by<00:04:31.349> piece<00:04:31.590> and<00:04:32.870> this<00:04:33.870> protocol<00:04:36.020> can<00:04:37.020> be 00:04:38.740 --> 00:04:38.750 align:start position:0% spoof by piece and this protocol can be 00:04:38.750 --> 00:04:41.059 align:start position:0% spoof by piece and this protocol can be successful<00:04:39.750> to<00:04:39.900> a<00:04:39.930> couple<00:04:40.560> different<00:04:40.919> types 00:04:41.059 --> 00:04:41.069 align:start position:0% successful to a couple different types 00:04:41.069 --> 00:04:42.950 align:start position:0% successful to a couple different types of<00:04:41.220> vulnerabilities<00:04:41.300> one<00:04:42.300> of<00:04:42.449> them<00:04:42.659> is 00:04:42.950 --> 00:04:42.960 align:start position:0% of vulnerabilities one of them is 00:04:42.960 --> 00:04:45.770 align:start position:0% of vulnerabilities one of them is Eclipse<00:04:43.409> attacks<00:04:44.069> where<00:04:44.849> a<00:04:44.879> node<00:04:45.389> is 00:04:45.770 --> 00:04:45.780 align:start position:0% Eclipse attacks where a node is 00:04:45.780 --> 00:04:47.450 align:start position:0% Eclipse attacks where a node is partitioned<00:04:46.379> off<00:04:46.650> from<00:04:46.919> the<00:04:47.039> rest<00:04:47.220> of<00:04:47.340> the 00:04:47.450 --> 00:04:47.460 align:start position:0% partitioned off from the rest of the 00:04:47.460 --> 00:04:48.800 align:start position:0% partitioned off from the rest of the network<00:04:47.639> and<00:04:48.120> they<00:04:48.360> only<00:04:48.479> wants<00:04:48.780> to 00:04:48.800 --> 00:04:48.810 align:start position:0% network and they only wants to 00:04:48.810 --> 00:04:51.580 align:start position:0% network and they only wants to communicate<00:04:49.020> with<00:04:49.560> our<00:04:49.889> nodes<00:04:50.460> controlled<00:04:50.879> by 00:04:51.580 --> 00:04:51.590 align:start position:0% communicate with our nodes controlled by 00:04:51.590 --> 00:04:53.439 align:start position:0% communicate with our nodes controlled by the<00:04:52.590> attacker 00:04:53.439 --> 00:04:53.449 align:start position:0% the attacker 00:04:53.449 --> 00:04:56.300 align:start position:0% the attacker this<00:04:54.449> protocol<00:04:54.840> can<00:04:54.990> also<00:04:55.139> be<00:04:55.319> used<00:04:55.440> to<00:04:55.680> do 00:04:56.300 --> 00:04:56.310 align:start position:0% this protocol can also be used to do 00:04:56.310 --> 00:04:58.249 align:start position:0% this protocol can also be used to do distributed<00:04:57.150> denial<00:04:57.449> of<00:04:57.539> service<00:04:57.599> attacks 00:04:58.249 --> 00:04:58.259 align:start position:0% distributed denial of service attacks 00:04:58.259 --> 00:05:00.350 align:start position:0% distributed denial of service attacks against<00:04:58.830> third<00:04:59.129> parties<00:04:59.370> that<00:04:59.789> are<00:04:59.970> nothing 00:05:00.350 --> 00:05:00.360 align:start position:0% against third parties that are nothing 00:05:00.360 --> 00:05:04.939 align:start position:0% against third parties that are nothing to<00:05:00.599> do<00:05:00.719> with<00:05:00.810> ethereum<00:05:02.930> and<00:05:03.949> amplification 00:05:04.939 --> 00:05:04.949 align:start position:0% to do with ethereum and amplification 00:05:04.949 --> 00:05:07.309 align:start position:0% to do with ethereum and amplification attacks<00:05:05.639> which<00:05:05.969> kind<00:05:06.690> of<00:05:06.750> is<00:05:06.870> part<00:05:07.110> of<00:05:07.169> the 00:05:07.309 --> 00:05:07.319 align:start position:0% attacks which kind of is part of the 00:05:07.319 --> 00:05:12.890 align:start position:0% attacks which kind of is part of the DDoS<00:05:08.750> and<00:05:09.750> an<00:05:10.099> amplification<00:05:11.099> attack<00:05:12.000> can<00:05:12.719> be 00:05:12.890 --> 00:05:12.900 align:start position:0% DDoS and an amplification attack can be 00:05:12.900 --> 00:05:17.020 align:start position:0% DDoS and an amplification attack can be said<00:05:13.349> to<00:05:13.469> be<00:05:13.530> an<00:05:13.889> attack<00:05:14.219> where<00:05:14.490> I<00:05:14.719> have<00:05:15.719> a 00:05:17.020 --> 00:05:17.030 align:start position:0% said to be an attack where I have a 00:05:17.030 --> 00:05:19.909 align:start position:0% said to be an attack where I have a small<00:05:18.030> resource<00:05:18.719> that<00:05:19.199> can<00:05:19.440> be<00:05:19.590> a<00:05:19.620> small 00:05:19.909 --> 00:05:19.919 align:start position:0% small resource that can be a small 00:05:19.919 --> 00:05:22.279 align:start position:0% small resource that can be a small network<00:05:20.550> bandwidth<00:05:21.000> and<00:05:21.389> I<00:05:21.629> could<00:05:21.810> sell<00:05:22.020> my 00:05:22.279 --> 00:05:22.289 align:start position:0% network bandwidth and I could sell my 00:05:22.289 --> 00:05:25.219 align:start position:0% network bandwidth and I could sell my and<00:05:22.590> bytes<00:05:23.520> out<00:05:23.789> to<00:05:23.969> the<00:05:24.090> network<00:05:24.509> but<00:05:24.840> the 00:05:25.219 --> 00:05:25.229 align:start position:0% and bytes out to the network but the 00:05:25.229 --> 00:05:27.879 align:start position:0% and bytes out to the network but the effect<00:05:25.590> that<00:05:25.680> is<00:05:26.009> reached<00:05:26.759> on<00:05:27.000> the<00:05:27.030> victim<00:05:27.599> is 00:05:27.879 --> 00:05:27.889 align:start position:0% effect that is reached on the victim is 00:05:27.889 --> 00:05:32.499 align:start position:0% effect that is reached on the victim is amplified<00:05:28.889> by<00:05:29.129> a<00:05:29.159> factor<00:05:29.610> of<00:05:29.639> something<00:05:29.909> else 00:05:32.499 --> 00:05:32.509 align:start position:0% 00:05:32.509 --> 00:05:35.870 align:start position:0% so<00:05:33.509> basically<00:05:33.900> this<00:05:34.680> is<00:05:34.889> one<00:05:35.069> way<00:05:35.219> to<00:05:35.250> develop 00:05:35.870 --> 00:05:35.880 align:start position:0% so basically this is one way to develop 00:05:35.880 --> 00:05:39.170 align:start position:0% so basically this is one way to develop the<00:05:35.969> attack<00:05:36.270> service<00:05:36.569> surface<00:05:37.349> I<00:05:38.180> omitted 00:05:39.170 --> 00:05:39.180 align:start position:0% the attack service surface I omitted 00:05:39.180 --> 00:05:41.120 align:start position:0% the attack service surface I omitted some<00:05:39.569> parts<00:05:39.960> there's<00:05:40.349> also<00:05:40.620> the<00:05:40.860> sync 00:05:41.120 --> 00:05:41.130 align:start position:0% some parts there's also the sync 00:05:41.130 --> 00:05:46.399 align:start position:0% some parts there's also the sync protocols<00:05:43.520> but<00:05:44.520> there<00:05:44.940> were<00:05:45.029> only<00:05:45.090> four<00:05:45.690> forth 00:05:46.399 --> 00:05:46.409 align:start position:0% protocols but there were only four forth 00:05:46.409 --> 00:05:49.550 align:start position:0% protocols but there were only four forth things<00:05:46.979> in<00:05:47.250> this<00:05:47.400> slide<00:05:47.969> template<00:05:48.659> so<00:05:48.810> I<00:05:48.900> had 00:05:49.550 --> 00:05:49.560 align:start position:0% things in this slide template so I had 00:05:49.560 --> 00:05:52.760 align:start position:0% things in this slide template so I had to<00:05:49.680> divvy<00:05:49.889> it<00:05:49.979> up<00:05:50.039> like<00:05:50.340> this<00:05:51.380> so<00:05:52.380> I'm<00:05:52.590> going 00:05:52.760 --> 00:05:52.770 align:start position:0% to divvy it up like this so I'm going 00:05:52.770 --> 00:05:58.010 align:start position:0% to divvy it up like this so I'm going from<00:05:53.009> the<00:05:53.099> bottom<00:05:53.400> up<00:05:53.550> here<00:05:54.860> and<00:05:56.060> during<00:05:57.060> 2018 00:05:58.010 --> 00:05:58.020 align:start position:0% from the bottom up here and during 2018 00:05:58.020 --> 00:06:00.709 align:start position:0% from the bottom up here and during 2018 and<00:05:58.169> 2019<00:05:58.800> there<00:05:59.610> was<00:05:59.789> actually<00:06:00.240> quite<00:06:00.270> a<00:06:00.539> lot 00:06:00.709 --> 00:06:00.719 align:start position:0% and 2019 there was actually quite a lot 00:06:00.719 --> 00:06:02.990 align:start position:0% and 2019 there was actually quite a lot research<00:06:01.199> into<00:06:01.650> the<00:06:01.770> discovery<00:06:02.250> protocol<00:06:02.729> and 00:06:02.990 --> 00:06:03.000 align:start position:0% research into the discovery protocol and 00:06:03.000 --> 00:06:07.760 align:start position:0% research into the discovery protocol and there<00:06:03.120> were<00:06:03.240> two<00:06:04.099> papers<00:06:05.099> sent<00:06:06.090> to<00:06:06.629> us<00:06:06.840> before 00:06:07.760 --> 00:06:07.770 align:start position:0% there were two papers sent to us before 00:06:07.770 --> 00:06:09.499 align:start position:0% there were two papers sent to us before they<00:06:07.919> were<00:06:07.979> released<00:06:08.580> to<00:06:08.729> the<00:06:08.789> general<00:06:08.849> public 00:06:09.499 --> 00:06:09.509 align:start position:0% they were released to the general public 00:06:09.509 --> 00:06:13.909 align:start position:0% they were released to the general public about<00:06:10.279> the<00:06:11.279> Klipsch<00:06:11.490> attacks<00:06:11.879> and<00:06:12.919> some 00:06:13.909 --> 00:06:13.919 align:start position:0% about the Klipsch attacks and some 00:06:13.919 --> 00:06:16.700 align:start position:0% about the Klipsch attacks and some people<00:06:14.250> also<00:06:14.370> reported<00:06:14.940> by<00:06:15.270> the<00:06:15.330> bouncy<00:06:15.960> bug 00:06:16.700 --> 00:06:16.710 align:start position:0% people also reported by the bouncy bug 00:06:16.710 --> 00:06:19.939 align:start position:0% people also reported by the bouncy bug bounty<00:06:17.129> and<00:06:17.930> vulnerabilities<00:06:18.949> amplification 00:06:19.939 --> 00:06:19.949 align:start position:0% bounty and vulnerabilities amplification 00:06:19.949 --> 00:06:22.790 align:start position:0% bounty and vulnerabilities amplification DDoS<00:06:20.310> attacks<00:06:20.639> and<00:06:21.060> Felix<00:06:21.509> and<00:06:21.840> Frank<00:06:22.229> on<00:06:22.620> the 00:06:22.790 --> 00:06:22.800 align:start position:0% DDoS attacks and Felix and Frank on the 00:06:22.800 --> 00:06:26.360 align:start position:0% DDoS attacks and Felix and Frank on the guest<00:06:22.949> team<00:06:23.190> found<00:06:23.550> even<00:06:23.729> more<00:06:24.029> issues<00:06:25.370> and 00:06:26.360 --> 00:06:26.370 align:start position:0% guest team found even more issues and 00:06:26.370 --> 00:06:28.909 align:start position:0% guest team found even more issues and one<00:06:26.610> of<00:06:26.699> the<00:06:26.789> eclipse<00:06:27.150> attacks<00:06:27.539> that<00:06:28.319> was 00:06:28.909 --> 00:06:28.919 align:start position:0% one of the eclipse attacks that was 00:06:28.919 --> 00:06:33.140 align:start position:0% one of the eclipse attacks that was found<00:06:29.159> in<00:06:29.279> the<00:06:29.370> first<00:06:29.610> paper<00:06:29.849> is<00:06:30.779> that<00:06:31.560> gif<00:06:32.430> as 00:06:33.140 --> 00:06:33.150 align:start position:0% found in the first paper is that gif as 00:06:33.150 --> 00:06:34.730 align:start position:0% found in the first paper is that gif as a<00:06:33.210> client 00:06:34.730 --> 00:06:34.740 align:start position:0% a client 00:06:34.740 --> 00:06:38.500 align:start position:0% a client had<00:06:35.370> dedicated<00:06:35.940> half<00:06:36.870> the<00:06:37.680> connection 00:06:38.500 --> 00:06:38.510 align:start position:0% had dedicated half the connection 00:06:38.510 --> 00:06:41.360 align:start position:0% had dedicated half the connection numbers<00:06:39.510> to<00:06:40.080> be<00:06:40.200> incoming<00:06:40.650> and<00:06:40.830> half<00:06:41.040> to<00:06:41.250> be 00:06:41.360 --> 00:06:41.370 align:start position:0% numbers to be incoming and half to be 00:06:41.370 --> 00:06:44.510 align:start position:0% numbers to be incoming and half to be outgoing<00:06:41.760> however<00:06:42.630> if<00:06:43.020> it<00:06:43.800> hadn't<00:06:44.310> already 00:06:44.510 --> 00:06:44.520 align:start position:0% outgoing however if it hadn't already 00:06:44.520 --> 00:06:47.270 align:start position:0% outgoing however if it hadn't already failed<00:06:45.120> done<00:06:45.990> all<00:06:46.230> the<00:06:46.380> outgoing<00:06:46.830> connections 00:06:47.270 --> 00:06:47.280 align:start position:0% failed done all the outgoing connections 00:06:47.280 --> 00:06:50.510 align:start position:0% failed done all the outgoing connections it<00:06:47.910> allowed<00:06:48.420> the<00:06:48.870> incoming<00:06:49.320> connections<00:06:49.920> to 00:06:50.510 --> 00:06:50.520 align:start position:0% it allowed the incoming connections to 00:06:50.520 --> 00:06:53.570 align:start position:0% it allowed the incoming connections to fill<00:06:50.850> up<00:06:50.970> all<00:06:51.120> the<00:06:51.300> slots<00:06:51.450> so<00:06:52.350> an<00:06:53.190> attacker 00:06:53.570 --> 00:06:53.580 align:start position:0% fill up all the slots so an attacker 00:06:53.580 --> 00:06:55.910 align:start position:0% fill up all the slots so an attacker could<00:06:53.790> simply<00:06:54.180> just<00:06:54.240> over<00:06:54.750> we're<00:06:54.960> overwhelmed 00:06:55.910 --> 00:06:55.920 align:start position:0% could simply just over we're overwhelmed 00:06:55.920 --> 00:06:58.280 align:start position:0% could simply just over we're overwhelmed I<00:06:56.130> guess<00:06:56.610> no<00:06:56.910> with<00:06:57.300> incoming<00:06:57.630> connection<00:06:58.110> and 00:06:58.280 --> 00:06:58.290 align:start position:0% I guess no with incoming connection and 00:06:58.290 --> 00:07:02.170 align:start position:0% I guess no with incoming connection and have<00:06:59.190> all<00:06:59.550> the<00:06:59.760> connections<00:07:00.210> to<00:07:00.360> the<00:07:00.450> thing 00:07:02.170 --> 00:07:02.180 align:start position:0% have all the connections to the thing 00:07:02.180 --> 00:07:04.100 align:start position:0% have all the connections to the thing they<00:07:03.180> could<00:07:03.330> also<00:07:03.450> see<00:07:03.720> that<00:07:03.750> there<00:07:04.020> were 00:07:04.100 --> 00:07:04.110 align:start position:0% they could also see that there were 00:07:04.110 --> 00:07:07.190 align:start position:0% they could also see that there were other<00:07:04.260> ways<00:07:04.500> to<00:07:05.010> fill<00:07:05.400> out<00:07:05.580> slots<00:07:06.000> by<00:07:06.780> pinging 00:07:07.190 --> 00:07:07.200 align:start position:0% other ways to fill out slots by pinging 00:07:07.200 --> 00:07:10.250 align:start position:0% other ways to fill out slots by pinging pinging<00:07:07.590> them<00:07:08.130> and<00:07:08.430> looting<00:07:09.420> the<00:07:09.600> identity 00:07:10.250 --> 00:07:10.260 align:start position:0% pinging them and looting the identity 00:07:10.260 --> 00:07:13.310 align:start position:0% pinging them and looting the identity database<00:07:11.180> and<00:07:12.180> the<00:07:12.300> third<00:07:12.570> attack<00:07:12.870> they<00:07:13.080> found 00:07:13.310 --> 00:07:13.320 align:start position:0% database and the third attack they found 00:07:13.320 --> 00:07:16.850 align:start position:0% database and the third attack they found was<00:07:13.560> that<00:07:13.590> if<00:07:14.040> you<00:07:14.220> manage<00:07:15.150> to<00:07:15.410> disturb<00:07:16.410> the 00:07:16.850 --> 00:07:16.860 align:start position:0% was that if you manage to disturb the 00:07:16.860 --> 00:07:21.950 align:start position:0% was that if you manage to disturb the timing<00:07:17.550> at<00:07:18.270> the<00:07:18.420> time<00:07:18.750> that<00:07:19.530> a<00:07:19.860> node<00:07:20.460> has<00:07:20.960> other 00:07:21.950 --> 00:07:21.960 align:start position:0% timing at the time that a node has other 00:07:21.960 --> 00:07:24.920 align:start position:0% timing at the time that a node has other peers<00:07:22.500> will<00:07:22.860> reject<00:07:23.550> him<00:07:23.670> and<00:07:24.210> forget<00:07:24.780> about 00:07:24.920 --> 00:07:24.930 align:start position:0% peers will reject him and forget about 00:07:24.930 --> 00:07:28.510 align:start position:0% peers will reject him and forget about him<00:07:25.290> and<00:07:25.440> he<00:07:26.040> will<00:07:26.220> reject<00:07:26.550> other<00:07:26.760> peers<00:07:27.180> and 00:07:28.510 --> 00:07:28.520 align:start position:0% him and he will reject other peers and 00:07:28.520 --> 00:07:30.830 align:start position:0% him and he will reject other peers and only<00:07:29.520> the<00:07:29.640> malicious<00:07:30.000> payers<00:07:30.240> will<00:07:30.510> accept 00:07:30.830 --> 00:07:30.840 align:start position:0% only the malicious payers will accept 00:07:30.840 --> 00:07:36.380 align:start position:0% only the malicious payers will accept him<00:07:32.360> as<00:07:33.360> for<00:07:33.540> amplification<00:07:34.580> attacks<00:07:35.580> here's 00:07:36.380 --> 00:07:36.390 align:start position:0% him as for amplification attacks here's 00:07:36.390 --> 00:07:39.790 align:start position:0% him as for amplification attacks here's one<00:07:37.160> example<00:07:38.160> that<00:07:38.340> Jeff<00:07:38.700> was<00:07:38.940> susceptible<00:07:39.270> to 00:07:39.790 --> 00:07:39.800 align:start position:0% one example that Jeff was susceptible to 00:07:39.800 --> 00:07:47.090 align:start position:0% one example that Jeff was susceptible to so<00:07:41.750> if<00:07:42.750> a<00:07:42.960> malicious<00:07:43.410> node<00:07:44.750> spoof<00:07:45.750> and<00:07:46.260> spoof 00:07:47.090 --> 00:07:47.100 align:start position:0% so if a malicious node spoof and spoof 00:07:47.100 --> 00:07:50.600 align:start position:0% so if a malicious node spoof and spoof pinged<00:07:47.550> the<00:07:47.910> target<00:07:48.590> the<00:07:49.590> target<00:07:49.800> would<00:07:50.100> reach 00:07:50.600 --> 00:07:50.610 align:start position:0% pinged the target the target would reach 00:07:50.610 --> 00:07:53.060 align:start position:0% pinged the target the target would reach out<00:07:50.910> to<00:07:51.210> the<00:07:51.330> actual<00:07:51.630> victim<00:07:52.080> with<00:07:52.770> the<00:07:52.860> ping 00:07:53.060 --> 00:07:53.070 align:start position:0% out to the actual victim with the ping 00:07:53.070 --> 00:07:55.580 align:start position:0% out to the actual victim with the ping with<00:07:53.940> the<00:07:54.030> poems<00:07:54.360> or<00:07:54.540> in<00:07:54.660> the<00:07:54.780> new<00:07:55.110> ping<00:07:55.380> and 00:07:55.580 --> 00:07:55.590 align:start position:0% with the poems or in the new ping and 00:07:55.590 --> 00:07:57.680 align:start position:0% with the poems or in the new ping and the<00:07:56.250> malicious<00:07:56.610> node<00:07:56.850> could<00:07:57.090> send<00:07:57.420> in<00:07:57.540> the 00:07:57.680 --> 00:07:57.690 align:start position:0% the malicious node could send in the 00:07:57.690 --> 00:08:01.550 align:start position:0% the malicious node could send in the poem<00:07:58.880> pretending<00:07:59.880> to<00:08:00.030> have<00:08:00.120> seen<00:08:00.390> the<00:08:00.570> first 00:08:01.550 --> 00:08:01.560 align:start position:0% poem pretending to have seen the first 00:08:01.560 --> 00:08:04.790 align:start position:0% poem pretending to have seen the first outgoing<00:08:02.220> ping<00:08:02.460> from<00:08:02.730> the<00:08:02.850> target<00:08:03.180> and<00:08:03.800> he's 00:08:04.790 --> 00:08:04.800 align:start position:0% outgoing ping from the target and he's 00:08:04.800 --> 00:08:06.740 align:start position:0% outgoing ping from the target and he's supposed<00:08:05.190> to<00:08:05.310> supply<00:08:05.580> reply<00:08:06.000> token<00:08:06.480> but 00:08:06.740 --> 00:08:06.750 align:start position:0% supposed to supply reply token but 00:08:06.750 --> 00:08:08.960 align:start position:0% supposed to supply reply token but didn't<00:08:07.320> actually<00:08:07.500> validate<00:08:08.130> the<00:08:08.280> reply<00:08:08.550> token 00:08:08.960 --> 00:08:08.970 align:start position:0% didn't actually validate the reply token 00:08:08.970 --> 00:08:13.150 align:start position:0% didn't actually validate the reply token so<00:08:09.720> at<00:08:10.050> that<00:08:10.200> point<00:08:10.440> target<00:08:11.070> believed<00:08:12.110> that 00:08:13.150 --> 00:08:13.160 align:start position:0% so at that point target believed that 00:08:13.160 --> 00:08:16.790 align:start position:0% so at that point target believed that malicious<00:08:14.160> node<00:08:14.340> resided<00:08:14.940> on<00:08:15.210> IP<00:08:15.990> address<00:08:16.410> 1<00:08:16.620> 2 00:08:16.790 --> 00:08:16.800 align:start position:0% malicious node resided on IP address 1 2 00:08:16.800 --> 00:08:21.320 align:start position:0% malicious node resided on IP address 1 2 3<00:08:16.980> 4<00:08:17.310> and<00:08:18.080> if<00:08:19.080> the<00:08:19.230> militia<00:08:19.470> of<00:08:19.740> them<00:08:20.150> sends<00:08:21.150> a 00:08:21.320 --> 00:08:21.330 align:start position:0% 3 4 and if the militia of them sends a 00:08:21.330 --> 00:08:23.500 align:start position:0% 3 4 and if the militia of them sends a so-called<00:08:21.720> final<00:08:22.230> request<00:08:22.830> which<00:08:23.130> is<00:08:23.160> only 00:08:23.500 --> 00:08:23.510 align:start position:0% so-called final request which is only 00:08:23.510 --> 00:08:27.800 align:start position:0% so-called final request which is only 213<00:08:24.510> bytes<00:08:24.780> or<00:08:24.990> so<00:08:25.700> it<00:08:26.700> causes<00:08:27.060> a<00:08:27.240> final 00:08:27.800 --> 00:08:27.810 align:start position:0% 213 bytes or so it causes a final 00:08:27.810 --> 00:08:30.800 align:start position:0% 213 bytes or so it causes a final response<00:08:28.400> called<00:08:29.400> neighbors<00:08:29.760> to<00:08:30.330> be<00:08:30.480> sent<00:08:30.690> to 00:08:30.800 --> 00:08:30.810 align:start position:0% response called neighbors to be sent to 00:08:30.810 --> 00:08:33.320 align:start position:0% response called neighbors to be sent to the<00:08:30.870> victim<00:08:31.050> which<00:08:31.410> is<00:08:31.590> 1.5<00:08:32.340> kilobytes 00:08:33.320 --> 00:08:33.330 align:start position:0% the victim which is 1.5 kilobytes 00:08:33.330 --> 00:08:35.750 align:start position:0% the victim which is 1.5 kilobytes there's<00:08:33.780> an<00:08:33.900> amplification<00:08:34.230> like<00:08:34.920> an 00:08:35.750 --> 00:08:35.760 align:start position:0% there's an amplification like an 00:08:35.760 --> 00:08:39.620 align:start position:0% there's an amplification like an amplification<00:08:36.270> of<00:08:36.600> about<00:08:36.870> 7<00:08:38.240> on<00:08:39.240> these 00:08:39.620 --> 00:08:39.630 align:start position:0% amplification of about 7 on these 00:08:39.630 --> 00:08:44.870 align:start position:0% amplification of about 7 on these attacks<00:08:39.990> which<00:08:40.620> is<00:08:40.800> not<00:08:41.520> trivial<00:08:43.880> more 00:08:44.870 --> 00:08:44.880 align:start position:0% attacks which is not trivial more 00:08:44.880 --> 00:08:47.540 align:start position:0% attacks which is not trivial more complex<00:08:45.390> example<00:08:45.930> could<00:08:46.230> be<00:08:46.350> made<00:08:46.530> also<00:08:46.950> where 00:08:47.540 --> 00:08:47.550 align:start position:0% complex example could be made also where 00:08:47.550 --> 00:08:48.530 align:start position:0% complex example could be made also where you<00:08:47.670> set<00:08:48.030> up<00:08:48.180> the<00:08:48.330> ball 00:08:48.530 --> 00:08:48.540 align:start position:0% you set up the ball 00:08:48.540 --> 00:08:51.829 align:start position:0% you set up the ball ending<00:08:48.779> with<00:08:49.380> two<00:08:49.620> notes<00:08:50.100> and<00:08:50.459> then<00:08:51.120> had<00:08:51.480> one 00:08:51.829 --> 00:08:51.839 align:start position:0% ending with two notes and then had one 00:08:51.839 --> 00:08:55.069 align:start position:0% ending with two notes and then had one of<00:08:51.990> them<00:08:52.199> lie<00:08:52.709> about<00:08:53.339> the<00:08:53.940> location<00:08:54.810> of<00:08:54.899> the 00:08:55.069 --> 00:08:55.079 align:start position:0% of them lie about the location of the 00:08:55.079 --> 00:08:58.579 align:start position:0% of them lie about the location of the first<00:08:55.320> one<00:08:55.620> and<00:08:56.009> the<00:08:56.730> first<00:08:56.940> one<00:08:57.149> then<00:08:57.570> do<00:08:58.410> is 00:08:58.579 --> 00:08:58.589 align:start position:0% first one and the first one then do is 00:08:58.589 --> 00:09:02.540 align:start position:0% first one and the first one then do is moved<00:08:58.949> find<00:08:59.610> out<00:08:59.820> requests<00:09:00.420> from<00:09:00.779> that<00:09:01.550> IP 00:09:02.540 --> 00:09:02.550 align:start position:0% moved find out requests from that IP 00:09:02.550 --> 00:09:05.389 align:start position:0% moved find out requests from that IP address<00:09:02.759> which<00:09:03.180> he'd<00:09:03.360> lied<00:09:03.570> about<00:09:03.720> and<00:09:04.399> ankles 00:09:05.389 --> 00:09:05.399 align:start position:0% address which he'd lied about and ankles 00:09:05.399 --> 00:09:09.050 align:start position:0% address which he'd lied about and ankles the<00:09:05.519> same<00:09:05.670> effect<00:09:07.310> moving<00:09:08.310> upwards<00:09:08.790> in<00:09:09.029> the 00:09:09.050 --> 00:09:09.060 align:start position:0% the same effect moving upwards in the 00:09:09.060 --> 00:09:11.269 align:start position:0% the same effect moving upwards in the network<00:09:09.779> stack<00:09:10.019> to<00:09:10.199> ethan<00:09:10.620> peer-to-peer<00:09:10.860> la 00:09:11.269 --> 00:09:11.279 align:start position:0% network stack to ethan peer-to-peer la 00:09:11.279 --> 00:09:15.949 align:start position:0% network stack to ethan peer-to-peer la lair<00:09:12.920> in<00:09:13.920> the<00:09:14.040> beginning<00:09:14.370> of<00:09:14.459> 2017<00:09:15.420> we<00:09:15.750> saw 00:09:15.949 --> 00:09:15.959 align:start position:0% lair in the beginning of 2017 we saw 00:09:15.959 --> 00:09:18.290 align:start position:0% lair in the beginning of 2017 we saw that<00:09:16.139> the<00:09:16.620> network<00:09:17.100> traffic<00:09:17.339> was<00:09:17.790> really<00:09:18.089> up 00:09:18.290 --> 00:09:18.300 align:start position:0% that the network traffic was really up 00:09:18.300 --> 00:09:20.990 align:start position:0% that the network traffic was really up here<00:09:18.810> so<00:09:19.410> much<00:09:19.620> noise<00:09:19.889> and<00:09:20.579> we<00:09:20.699> started 00:09:20.990 --> 00:09:21.000 align:start position:0% here so much noise and we started 00:09:21.000 --> 00:09:22.519 align:start position:0% here so much noise and we started analyzing<00:09:21.149> and<00:09:21.690> looking<00:09:21.930> into<00:09:22.019> what<00:09:22.259> caused 00:09:22.519 --> 00:09:22.529 align:start position:0% analyzing and looking into what caused 00:09:22.529 --> 00:09:24.680 align:start position:0% analyzing and looking into what caused that<00:09:22.560> and<00:09:23.100> found<00:09:23.250> that<00:09:23.459> during<00:09:23.759> about<00:09:24.209> two 00:09:24.680 --> 00:09:24.690 align:start position:0% that and found that during about two 00:09:24.690 --> 00:09:28.389 align:start position:0% that and found that during about two hours<00:09:25.100> and<00:09:26.100> note<00:09:26.399> would<00:09:26.670> typically<00:09:27.060> reject 00:09:28.389 --> 00:09:28.399 align:start position:0% hours and note would typically reject 00:09:28.399 --> 00:09:30.680 align:start position:0% hours and note would typically reject 165,000<00:09:29.399> transactions<00:09:30.180> because<00:09:30.360> they<00:09:30.630> were 00:09:30.680 --> 00:09:30.690 align:start position:0% 165,000 transactions because they were 00:09:30.690 --> 00:09:33.710 align:start position:0% 165,000 transactions because they were invalid<00:09:31.410> so<00:09:32.100> 19<00:09:32.610> invalid<00:09:33.000> transactions<00:09:33.480> every 00:09:33.710 --> 00:09:33.720 align:start position:0% invalid so 19 invalid transactions every 00:09:33.720 --> 00:09:40.360 align:start position:0% invalid so 19 invalid transactions every second<00:09:34.110> just<00:09:35.089> streaming<00:09:36.089> into<00:09:36.420> the<00:09:36.540> client 00:09:40.360 --> 00:09:40.370 align:start position:0% 00:09:40.370 --> 00:09:43.569 align:start position:0% problems<00:09:41.370> were<00:09:41.579> that<00:09:41.930> one<00:09:42.930> of<00:09:42.990> the<00:09:43.079> clients 00:09:43.569 --> 00:09:43.579 align:start position:0% problems were that one of the clients 00:09:43.579 --> 00:09:46.610 align:start position:0% problems were that one of the clients parrot<00:09:44.579> I<00:09:44.699> didn't<00:09:45.060> remember<00:09:45.750> who<00:09:46.110> they'd<00:09:46.290> sent 00:09:46.610 --> 00:09:46.620 align:start position:0% parrot I didn't remember who they'd sent 00:09:46.620 --> 00:09:48.710 align:start position:0% parrot I didn't remember who they'd sent transactions<00:09:47.220> to<00:09:47.430> and<00:09:47.699> another<00:09:47.940> issue<00:09:48.449> was 00:09:48.710 --> 00:09:48.720 align:start position:0% transactions to and another issue was 00:09:48.720 --> 00:09:51.650 align:start position:0% transactions to and another issue was that<00:09:48.750> they<00:09:49.649> didn't<00:09:50.009> actually<00:09:50.660> garbage 00:09:51.650 --> 00:09:51.660 align:start position:0% that they didn't actually garbage 00:09:51.660 --> 00:09:54.290 align:start position:0% that they didn't actually garbage collect<00:09:52.079> invalid<00:09:52.589> transactions<00:09:53.250> asked<00:09:53.730> state 00:09:54.290 --> 00:09:54.300 align:start position:0% collect invalid transactions asked state 00:09:54.300 --> 00:09:55.970 align:start position:0% collect invalid transactions asked state progressed<00:09:54.930> over<00:09:55.230> blocks 00:09:55.970 --> 00:09:55.980 align:start position:0% progressed over blocks 00:09:55.980 --> 00:09:58.730 align:start position:0% progressed over blocks so<00:09:56.490> we've<00:09:56.790> if<00:09:57.690> they<00:09:57.930> were<00:09:57.990> included<00:09:58.529> they<00:09:58.680> were 00:09:58.730 --> 00:09:58.740 align:start position:0% so we've if they were included they were 00:09:58.740 --> 00:10:00.620 align:start position:0% so we've if they were included they were removed<00:09:59.430> but<00:09:59.639> if<00:09:59.850> they<00:10:00.029> became<00:10:00.300> invalid 00:10:00.620 --> 00:10:00.630 align:start position:0% removed but if they became invalid 00:10:00.630 --> 00:10:02.960 align:start position:0% removed but if they became invalid because<00:10:01.050> of<00:10:01.350> they<00:10:01.440> projected<00:10:01.980> progressional 00:10:02.960 --> 00:10:02.970 align:start position:0% because of they projected progressional 00:10:02.970 --> 00:10:05.329 align:start position:0% because of they projected progressional blocks<00:10:03.180> and<00:10:03.630> they<00:10:04.410> were<00:10:04.529> not<00:10:04.680> removed<00:10:05.040> so<00:10:05.220> they 00:10:05.329 --> 00:10:05.339 align:start position:0% blocks and they were not removed so they 00:10:05.339 --> 00:10:08.150 align:start position:0% blocks and they were not removed so they were<00:10:05.459> just<00:10:05.490> constantly<00:10:06.240> being<00:10:06.510> recent<00:10:07.380> and 00:10:08.150 --> 00:10:08.160 align:start position:0% were just constantly being recent and 00:10:08.160 --> 00:10:11.840 align:start position:0% were just constantly being recent and after<00:10:08.819> we<00:10:08.970> looked<00:10:09.720> into<00:10:10.050> that<00:10:10.199> and<00:10:10.589> and<00:10:10.850> nail 00:10:11.840 --> 00:10:11.850 align:start position:0% after we looked into that and and nail 00:10:11.850 --> 00:10:14.059 align:start position:0% after we looked into that and and nail out<00:10:12.149> iron<00:10:12.600> out<00:10:12.839> those<00:10:13.050> issues<00:10:13.500> you<00:10:13.529> could<00:10:13.829> see 00:10:14.059 --> 00:10:14.069 align:start position:0% out iron out those issues you could see 00:10:14.069 --> 00:10:16.430 align:start position:0% out iron out those issues you could see that<00:10:14.279> the<00:10:14.490> the<00:10:15.089> graphs<00:10:15.510> on<00:10:15.779> these<00:10:15.930> monitoring 00:10:16.430 --> 00:10:16.440 align:start position:0% that the the graphs on these monitoring 00:10:16.440 --> 00:10:18.650 align:start position:0% that the the graphs on these monitoring notes<00:10:16.680> which<00:10:16.920> we<00:10:17.100> put<00:10:17.430> up<00:10:17.579> started<00:10:18.180> levelling 00:10:18.650 --> 00:10:18.660 align:start position:0% notes which we put up started levelling 00:10:18.660 --> 00:10:21.980 align:start position:0% notes which we put up started levelling out<00:10:18.810> and<00:10:19.230> that<00:10:19.920> the<00:10:20.160> incoming<00:10:20.850> traffic<00:10:21.420> on<00:10:21.959> a 00:10:21.980 --> 00:10:21.990 align:start position:0% out and that the incoming traffic on a 00:10:21.990 --> 00:10:24.740 align:start position:0% out and that the incoming traffic on a node<00:10:22.319> went<00:10:23.040> down<00:10:23.279> by<00:10:23.550> a<00:10:23.579> factor<00:10:24.029> of<00:10:24.089> 20<00:10:24.569> to<00:10:24.720> 30 00:10:24.740 --> 00:10:24.750 align:start position:0% node went down by a factor of 20 to 30 00:10:24.750 --> 00:10:32.540 align:start position:0% node went down by a factor of 20 to 30 during<00:10:25.970> January<00:10:26.970> to<00:10:27.149> March<00:10:27.360> in<00:10:27.889> 2017<00:10:30.920> and<00:10:31.920> also 00:10:32.540 --> 00:10:32.550 align:start position:0% during January to March in 2017 and also 00:10:32.550 --> 00:10:36.530 align:start position:0% during January to March in 2017 and also the<00:10:32.760> only<00:10:33.329> is<00:10:33.480> this<00:10:34.019> protocol<00:10:34.500> level<00:10:35.510> there<00:10:36.510> is 00:10:36.530 --> 00:10:36.540 align:start position:0% the only is this protocol level there is 00:10:36.540 --> 00:10:38.449 align:start position:0% the only is this protocol level there is the<00:10:36.810> sync<00:10:37.050> protocols<00:10:37.589> as<00:10:37.769> I<00:10:37.889> mentioned<00:10:38.339> and 00:10:38.449 --> 00:10:38.459 align:start position:0% the sync protocols as I mentioned and 00:10:38.459 --> 00:10:41.629 align:start position:0% the sync protocols as I mentioned and earlier<00:10:39.120> in<00:10:39.240> April<00:10:39.720> this<00:10:39.870> year<00:10:39.930> we<00:10:40.860> saw<00:10:41.160> a<00:10:41.459> new 00:10:41.629 --> 00:10:41.639 align:start position:0% earlier in April this year we saw a new 00:10:41.639 --> 00:10:43.009 align:start position:0% earlier in April this year we saw a new kind<00:10:41.910> of<00:10:42.000> attack<00:10:42.269> which<00:10:42.510> we<00:10:42.630> hadn't<00:10:42.870> seen 00:10:43.009 --> 00:10:43.019 align:start position:0% kind of attack which we hadn't seen 00:10:43.019 --> 00:10:45.019 align:start position:0% kind of attack which we hadn't seen before<00:10:43.050> which<00:10:43.439> was<00:10:43.649> a<00:10:43.709> target<00:10:44.130> attack<00:10:44.699> or<00:10:44.939> the 00:10:45.019 --> 00:10:45.029 align:start position:0% before which was a target attack or the 00:10:45.029 --> 00:10:48.199 align:start position:0% before which was a target attack or the fasting<00:10:45.689> protocol<00:10:46.579> which<00:10:47.579> was<00:10:47.699> a<00:10:47.759> grieving 00:10:48.199 --> 00:10:48.209 align:start position:0% fasting protocol which was a grieving 00:10:48.209 --> 00:10:51.050 align:start position:0% fasting protocol which was a grieving attack<00:10:48.509> where<00:10:48.870> as<00:10:49.620> soon<00:10:49.860> as<00:10:50.040> a<00:10:50.069> node<00:10:50.399> joined 00:10:51.050 --> 00:10:51.060 align:start position:0% attack where as soon as a node joined 00:10:51.060 --> 00:10:54.139 align:start position:0% attack where as soon as a node joined the<00:10:51.269> network<00:10:51.920> the<00:10:52.920> attacker<00:10:53.579> who<00:10:53.970> was 00:10:54.139 --> 00:10:54.149 align:start position:0% the network the attacker who was 00:10:54.149 --> 00:10:57.199 align:start position:0% the network the attacker who was scanning<00:10:54.509> the<00:10:54.779> DHT<00:10:55.079> tables<00:10:55.829> a<00:10:56.069> new<00:10:56.250> node<00:10:56.519> would 00:10:57.199 --> 00:10:57.209 align:start position:0% scanning the DHT tables a new node would 00:10:57.209 --> 00:10:59.750 align:start position:0% scanning the DHT tables a new node would instantly<00:10:57.870> connect<00:10:58.319> to<00:10:58.439> that<00:10:58.589> node<00:10:58.860> and<00:10:59.069> tell 00:10:59.750 --> 00:10:59.760 align:start position:0% instantly connect to that node and tell 00:10:59.760 --> 00:11:00.650 align:start position:0% instantly connect to that node and tell him 00:11:00.650 --> 00:11:00.660 align:start position:0% him 00:11:00.660 --> 00:11:02.840 align:start position:0% him Hey<00:11:01.170> I<00:11:01.350> had<00:11:01.590> the<00:11:01.710> blockchain<00:11:01.890> here<00:11:02.430> I'm<00:11:02.520> full 00:11:02.840 --> 00:11:02.850 align:start position:0% Hey I had the blockchain here I'm full 00:11:02.850 --> 00:11:05.450 align:start position:0% Hey I had the blockchain here I'm full of<00:11:02.880> sinked<00:11:03.240> at<00:11:03.390> block<00:11:03.570> five<00:11:03.840> and<00:11:04.320> the<00:11:05.070> victim 00:11:05.450 --> 00:11:05.460 align:start position:0% of sinked at block five and the victim 00:11:05.460 --> 00:11:07.910 align:start position:0% of sinked at block five and the victim having<00:11:05.820> no<00:11:06.060> other<00:11:06.210> source<00:11:06.630> of<00:11:06.840> truth<00:11:07.140> would 00:11:07.910 --> 00:11:07.920 align:start position:0% having no other source of truth would 00:11:07.920 --> 00:11:09.860 align:start position:0% having no other source of truth would tell<00:11:08.460> him<00:11:08.670> yeah<00:11:08.850> please<00:11:09.150> send<00:11:09.480> me<00:11:09.600> over<00:11:09.780> the 00:11:09.860 --> 00:11:09.870 align:start position:0% tell him yeah please send me over the 00:11:09.870 --> 00:11:11.840 align:start position:0% tell him yeah please send me over the states<00:11:10.110> on<00:11:10.260> the<00:11:10.290> blocks<00:11:10.560> and<00:11:10.890> that<00:11:11.400> of<00:11:11.580> course 00:11:11.840 --> 00:11:11.850 align:start position:0% states on the blocks and that of course 00:11:11.850 --> 00:11:14.900 align:start position:0% states on the blocks and that of course happens<00:11:12.240> very<00:11:12.540> quickly<00:11:12.810> since<00:11:13.590> what<00:11:13.890> it<00:11:14.190> was<00:11:14.340> a 00:11:14.900 --> 00:11:14.910 align:start position:0% happens very quickly since what it was a 00:11:14.910 --> 00:11:18.530 align:start position:0% happens very quickly since what it was a very<00:11:15.390> small<00:11:15.690> chain<00:11:15.990> and<00:11:16.910> the<00:11:17.910> victim<00:11:18.360> would 00:11:18.530 --> 00:11:18.540 align:start position:0% very small chain and the victim would 00:11:18.540 --> 00:11:22.160 align:start position:0% very small chain and the victim would then<00:11:18.720> be<00:11:19.140> fast<00:11:19.620> synched<00:11:20.160> to<00:11:20.310> block<00:11:20.880> five<00:11:21.420> or<00:11:21.900> so 00:11:22.160 --> 00:11:22.170 align:start position:0% then be fast synched to block five or so 00:11:22.170 --> 00:11:24.410 align:start position:0% then be fast synched to block five or so and<00:11:22.440> after<00:11:23.100> that<00:11:23.220> switch<00:11:23.580> over<00:11:23.910> to<00:11:23.940> block<00:11:24.240> by 00:11:24.410 --> 00:11:24.420 align:start position:0% and after that switch over to block by 00:11:24.420 --> 00:11:26.540 align:start position:0% and after that switch over to block by block<00:11:24.600> sync<00:11:24.990> effectively<00:11:25.620> doing<00:11:25.800> a<00:11:25.950> full<00:11:26.190> sync 00:11:26.540 --> 00:11:26.550 align:start position:0% block sync effectively doing a full sync 00:11:26.550 --> 00:11:29.090 align:start position:0% block sync effectively doing a full sync to<00:11:26.970> book<00:11:27.180> one<00:11:27.420> the<00:11:27.720> remaining<00:11:28.140> eight<00:11:28.830> point 00:11:29.090 --> 00:11:29.100 align:start position:0% to book one the remaining eight point 00:11:29.100 --> 00:11:32.570 align:start position:0% to book one the remaining eight point five<00:11:29.280> million<00:11:29.520> blocks<00:11:30.950> so<00:11:31.950> it<00:11:32.340> was<00:11:32.490> an 00:11:32.570 --> 00:11:32.580 align:start position:0% five million blocks so it was an 00:11:32.580 --> 00:11:34.460 align:start position:0% five million blocks so it was an interesting<00:11:32.760> attack<00:11:33.150> because<00:11:33.510> in<00:11:33.870> no<00:11:34.200> way 00:11:34.460 --> 00:11:34.470 align:start position:0% interesting attack because in no way 00:11:34.470 --> 00:11:37.430 align:start position:0% interesting attack because in no way does<00:11:34.770> it<00:11:34.830> benefit<00:11:35.430> the<00:11:35.490> attacker<00:11:36.000> he<00:11:36.900> needs<00:11:37.170> to 00:11:37.430 --> 00:11:37.440 align:start position:0% does it benefit the attacker he needs to 00:11:37.440 --> 00:11:40.010 align:start position:0% does it benefit the attacker he needs to actively<00:11:38.010> run<00:11:38.340> nodes<00:11:38.700> and<00:11:39.120> run<00:11:39.330> some<00:11:39.540> custom 00:11:40.010 --> 00:11:40.020 align:start position:0% actively run nodes and run some custom 00:11:40.020 --> 00:11:43.820 align:start position:0% actively run nodes and run some custom software<00:11:40.260> to<00:11:41.160> scan<00:11:41.430> the<00:11:41.610> DHT<00:11:42.300> and<00:11:42.650> do<00:11:43.650> this 00:11:43.820 --> 00:11:43.830 align:start position:0% software to scan the DHT and do this 00:11:43.830 --> 00:11:48.320 align:start position:0% software to scan the DHT and do this attack<00:11:44.220> and<00:11:45.770> gain<00:11:46.770> absolutely<00:11:47.580> nothing<00:11:47.970> other 00:11:48.320 --> 00:11:48.330 align:start position:0% attack and gain absolutely nothing other 00:11:48.330 --> 00:11:50.240 align:start position:0% attack and gain absolutely nothing other than<00:11:48.660> what's<00:11:48.960> the<00:11:49.230> etherion<00:11:49.770> world<00:11:49.980> burn<00:11:50.220> a 00:11:50.240 --> 00:11:50.250 align:start position:0% than what's the etherion world burn a 00:11:50.250 --> 00:11:58.480 align:start position:0% than what's the etherion world burn a bit<00:11:53.210> a<00:11:54.620> bug<00:11:55.620> bounty<00:11:56.000> from<00:11:57.110> junior-m 00:11:58.480 --> 00:11:58.490 align:start position:0% bit a bug bounty from junior-m 00:11:58.490 --> 00:12:01.610 align:start position:0% bit a bug bounty from junior-m was<00:11:59.490> that<00:11:59.910> there<00:12:00.330> was<00:12:00.480> a<00:12:00.510> vulnerability<00:12:01.290> in 00:12:01.610 --> 00:12:01.620 align:start position:0% was that there was a vulnerability in 00:12:01.620 --> 00:12:04.310 align:start position:0% was that there was a vulnerability in the<00:12:01.650> protocol<00:12:02.280> and<00:12:02.400> gap<00:12:02.610> where<00:12:03.540> someone<00:12:03.930> queer 00:12:04.310 --> 00:12:04.320 align:start position:0% the protocol and gap where someone queer 00:12:04.320 --> 00:12:07.040 align:start position:0% the protocol and gap where someone queer state<00:12:04.740> entries<00:12:05.280> with<00:12:05.610> a<00:12:05.640> so-called<00:12:06.150> kind<00:12:06.900> of<00:12:06.990> a 00:12:07.040 --> 00:12:07.050 align:start position:0% state entries with a so-called kind of a 00:12:07.050 --> 00:12:09.590 align:start position:0% state entries with a so-called kind of a null<00:12:07.790> requests<00:12:08.790> were<00:12:08.940> crossing<00:12:09.150> the<00:12:09.300> Tri 00:12:09.590 --> 00:12:09.600 align:start position:0% null requests were crossing the Tri 00:12:09.600 --> 00:12:11.900 align:start position:0% null requests were crossing the Tri metal<00:12:09.960> route<00:12:10.110> an<00:12:10.380> instantly<00:12:11.220> crash<00:12:11.610> and 00:12:11.900 --> 00:12:11.910 align:start position:0% metal route an instantly crash and 00:12:11.910 --> 00:12:16.610 align:start position:0% metal route an instantly crash and appeared<00:12:12.390> so<00:12:13.050> these<00:12:13.890> kinds<00:12:14.400> of<00:12:14.490> bugs<00:12:14.730> if<00:12:15.690> even 00:12:16.610 --> 00:12:16.620 align:start position:0% appeared so these kinds of bugs if even 00:12:16.620 --> 00:12:18.500 align:start position:0% appeared so these kinds of bugs if even an<00:12:16.710> attacker<00:12:17.070> has<00:12:17.400> them<00:12:17.670> he<00:12:17.820> can<00:12:17.970> iteratively 00:12:18.500 --> 00:12:18.510 align:start position:0% an attacker has them he can iteratively 00:12:18.510 --> 00:12:22.040 align:start position:0% an attacker has them he can iteratively shut<00:12:19.380> down<00:12:19.620> the<00:12:19.890> etherium<00:12:20.340> Network<00:12:20.840> but<00:12:21.840> just 00:12:22.040 --> 00:12:22.050 align:start position:0% shut down the etherium Network but just 00:12:22.050 --> 00:12:25.430 align:start position:0% shut down the etherium Network but just connecting<00:12:22.530> to<00:12:22.560> them<00:12:23.780> mapping<00:12:24.780> the<00:12:24.810> the 00:12:25.430 --> 00:12:25.440 align:start position:0% connecting to them mapping the the 00:12:25.440 --> 00:12:27.170 align:start position:0% connecting to them mapping the the amount<00:12:25.590> of<00:12:25.800> the<00:12:25.860> DHT<00:12:26.310> and<00:12:26.550> connected<00:12:26.910> them<00:12:27.030> is 00:12:27.170 --> 00:12:27.180 align:start position:0% amount of the DHT and connected them is 00:12:27.180 --> 00:12:30.170 align:start position:0% amount of the DHT and connected them is shut<00:12:27.360> them<00:12:27.540> down<00:12:27.780> iteratively<00:12:28.850> which<00:12:29.850> can<00:12:30.060> be 00:12:30.170 --> 00:12:30.180 align:start position:0% shut them down iteratively which can be 00:12:30.180 --> 00:12:34.520 align:start position:0% shut them down iteratively which can be really<00:12:30.510> really<00:12:30.660> bad<00:12:32.960> as<00:12:33.960> for<00:12:34.320> block 00:12:34.520 --> 00:12:34.530 align:start position:0% really really bad as for block 00:12:34.530 --> 00:12:39.140 align:start position:0% really really bad as for block processing<00:12:35.660> and<00:12:36.660> in<00:12:37.160> 2018<00:12:38.160> there<00:12:38.730> was<00:12:38.850> a<00:12:38.880> live 00:12:39.140 --> 00:12:39.150 align:start position:0% processing and in 2018 there was a live 00:12:39.150 --> 00:12:42.130 align:start position:0% processing and in 2018 there was a live consensus<00:12:39.780> floor<00:12:40.050> that<00:12:40.260> happened<00:12:40.620> on<00:12:40.710> Roxton 00:12:42.130 --> 00:12:42.140 align:start position:0% consensus floor that happened on Roxton 00:12:42.140 --> 00:12:45.950 align:start position:0% consensus floor that happened on Roxton Robson<00:12:43.140> is<00:12:43.230> one<00:12:43.350> of<00:12:43.440> the<00:12:43.500> test<00:12:43.710> nets<00:12:44.300> and<00:12:45.300> what 00:12:45.950 --> 00:12:45.960 align:start position:0% Robson is one of the test nets and what 00:12:45.960 --> 00:12:47.630 align:start position:0% Robson is one of the test nets and what happened<00:12:46.380> was<00:12:46.440> that<00:12:46.500> someone<00:12:46.950> had<00:12:47.070> sent<00:12:47.340> one 00:12:47.630 --> 00:12:47.640 align:start position:0% happened was that someone had sent one 00:12:47.640 --> 00:12:52.040 align:start position:0% happened was that someone had sent one of<00:12:47.670> these<00:12:47.850> EEP<00:12:48.750> 86<00:12:50.450> transactions<00:12:51.450> on<00:12:51.690> Rob 00:12:52.040 --> 00:12:52.050 align:start position:0% of these EEP 86 transactions on Rob 00:12:52.050 --> 00:12:52.840 align:start position:0% of these EEP 86 transactions on Rob stone 00:12:52.840 --> 00:12:52.850 align:start position:0% stone 00:12:52.850 --> 00:12:56.600 align:start position:0% stone now<00:12:53.850> EP<00:12:54.300> 86<00:12:54.810> and<00:12:55.170> was<00:12:55.500> Anita<00:12:55.770> proposed<00:12:56.370> a<00:12:56.490> new 00:12:56.600 --> 00:12:56.610 align:start position:0% now EP 86 and was Anita proposed a new 00:12:56.610 --> 00:12:58.820 align:start position:0% now EP 86 and was Anita proposed a new transaction<00:12:56.940> format<00:12:57.590> for<00:12:58.590> account 00:12:58.820 --> 00:12:58.830 align:start position:0% transaction format for account 00:12:58.830 --> 00:13:00.740 align:start position:0% transaction format for account abstraction<00:12:59.490> and<00:12:59.730> it<00:13:00.000> was<00:13:00.180> never<00:13:00.450> actually 00:13:00.740 --> 00:13:00.750 align:start position:0% abstraction and it was never actually 00:13:00.750 --> 00:13:02.960 align:start position:0% abstraction and it was never actually roll<00:13:01.410> out<00:13:01.620> it<00:13:01.890> was<00:13:02.010> never<00:13:02.220> actually<00:13:02.400> included 00:13:02.960 --> 00:13:02.970 align:start position:0% roll out it was never actually included 00:13:02.970 --> 00:13:05.750 align:start position:0% roll out it was never actually included in<00:13:03.090> any<00:13:03.300> hard<00:13:03.540> work<00:13:03.980> but<00:13:04.980> parity<00:13:05.580> had 00:13:05.750 --> 00:13:05.760 align:start position:0% in any hard work but parity had 00:13:05.760 --> 00:13:07.760 align:start position:0% in any hard work but parity had implemented<00:13:06.270> it<00:13:06.570> and<00:13:06.870> they<00:13:07.020> mistakenly 00:13:07.760 --> 00:13:07.770 align:start position:0% implemented it and they mistakenly 00:13:07.770 --> 00:13:11.180 align:start position:0% implemented it and they mistakenly enabled<00:13:08.610> it<00:13:09.200> in<00:13:10.200> the<00:13:10.410> client<00:13:10.770> and<00:13:10.950> not<00:13:11.040> only 00:13:11.180 --> 00:13:11.190 align:start position:0% enabled it in the client and not only 00:13:11.190 --> 00:13:12.410 align:start position:0% enabled it in the client and not only for<00:13:11.400> test<00:13:11.610> nuts 00:13:12.410 --> 00:13:12.420 align:start position:0% for test nuts 00:13:12.420 --> 00:13:16.880 align:start position:0% for test nuts so<00:13:12.780> this<00:13:12.960> was<00:13:13.140> kind<00:13:13.530> of<00:13:13.680> a<00:13:13.890> real<00:13:15.890> critical 00:13:16.880 --> 00:13:16.890 align:start position:0% so this was kind of a real critical 00:13:16.890 --> 00:13:19.490 align:start position:0% so this was kind of a real critical issue<00:13:17.100> and<00:13:17.640> we<00:13:17.940> were<00:13:18.090> just<00:13:18.300> waiting<00:13:18.810> for 00:13:19.490 --> 00:13:19.500 align:start position:0% issue and we were just waiting for 00:13:19.500 --> 00:13:21.230 align:start position:0% issue and we were just waiting for someone<00:13:19.860> hacked<00:13:20.160> this<00:13:20.430> someone<00:13:20.790> did<00:13:21.030> this 00:13:21.230 --> 00:13:21.240 align:start position:0% someone hacked this someone did this 00:13:21.240 --> 00:13:23.930 align:start position:0% someone hacked this someone did this found<00:13:21.810> the<00:13:21.990> floor<00:13:22.340> exploited<00:13:23.340> the<00:13:23.460> test<00:13:23.700> net 00:13:23.930 --> 00:13:23.940 align:start position:0% found the floor exploited the test net 00:13:23.940 --> 00:13:25.310 align:start position:0% found the floor exploited the test net and<00:13:24.150> we<00:13:24.240> were<00:13:24.390> just<00:13:24.540> waiting<00:13:24.720> for<00:13:24.930> it<00:13:25.080> to<00:13:25.170> hit 00:13:25.310 --> 00:13:25.320 align:start position:0% and we were just waiting for it to hit 00:13:25.320 --> 00:13:28.610 align:start position:0% and we were just waiting for it to hit maintenance<00:13:27.050> fortunately<00:13:28.050> it<00:13:28.200> did<00:13:28.380> not 00:13:28.610 --> 00:13:28.620 align:start position:0% maintenance fortunately it did not 00:13:28.620 --> 00:13:32.570 align:start position:0% maintenance fortunately it did not apparently<00:13:29.010> quickly<00:13:29.670> released<00:13:29.970> a<00:13:30.540> fix<00:13:31.170> and<00:13:31.590> we 00:13:32.570 --> 00:13:32.580 align:start position:0% apparently quickly released a fix and we 00:13:32.580 --> 00:13:35.090 align:start position:0% apparently quickly released a fix and we could<00:13:32.760> see<00:13:33.000> eventually<00:13:33.540> that<00:13:34.020> actually<00:13:34.650> there 00:13:35.090 --> 00:13:35.100 align:start position:0% could see eventually that actually there 00:13:35.100 --> 00:13:37.010 align:start position:0% could see eventually that actually there was<00:13:35.250> no<00:13:35.430> attacker<00:13:36.090> per<00:13:36.540> se 00:13:37.010 --> 00:13:37.020 align:start position:0% was no attacker per se 00:13:37.020 --> 00:13:39.950 align:start position:0% was no attacker per se it<00:13:37.860> was<00:13:38.040> basically<00:13:38.430> the<00:13:38.670> researchers<00:13:39.570> who 00:13:39.950 --> 00:13:39.960 align:start position:0% it was basically the researchers who 00:13:39.960 --> 00:13:43.250 align:start position:0% it was basically the researchers who were<00:13:40.200> playing<00:13:41.070> about<00:13:41.370> a<00:13:41.400> bit<00:13:41.670> with<00:13:41.970> these<00:13:42.840> kind 00:13:43.250 --> 00:13:43.260 align:start position:0% were playing about a bit with these kind 00:13:43.260 --> 00:13:46.490 align:start position:0% were playing about a bit with these kind of<00:13:43.380> transactions<00:13:44.870> so<00:13:45.870> not<00:13:46.020> all<00:13:46.230> instances 00:13:46.490 --> 00:13:46.500 align:start position:0% of transactions so not all instances 00:13:46.500 --> 00:13:51.830 align:start position:0% of transactions so not all instances caused<00:13:46.890> by<00:13:47.070> malicious<00:13:47.460> intents<00:13:49.760> and<00:13:50.760> in<00:13:51.720> the 00:13:51.830 --> 00:13:51.840 align:start position:0% caused by malicious intents and in the 00:13:51.840 --> 00:13:55.490 align:start position:0% caused by malicious intents and in the block<00:13:52.020> processing<00:13:52.620> area<00:13:53.720> there<00:13:54.720> was<00:13:54.840> also<00:13:55.050> in 00:13:55.490 --> 00:13:55.500 align:start position:0% block processing area there was also in 00:13:55.500 --> 00:13:58.430 align:start position:0% block processing area there was also in March<00:13:55.860> a<00:13:56.280> very<00:13:56.580> long-lived<00:13:57.180> consensus<00:13:58.050> bug<00:13:58.320> I 00:13:58.430 --> 00:13:58.440 align:start position:0% March a very long-lived consensus bug I 00:13:58.440 --> 00:14:00.320 align:start position:0% March a very long-lived consensus bug I think<00:13:58.740> that<00:13:58.950> might<00:13:59.550> have<00:13:59.790> been<00:13:59.820> the<00:14:00.090> most 00:14:00.320 --> 00:14:00.330 align:start position:0% think that might have been the most 00:14:00.330 --> 00:14:04.880 align:start position:0% think that might have been the most long-lived<00:14:00.810> consensus<00:14:01.380> bug<00:14:01.970> we<00:14:02.970> had<00:14:03.530> found<00:14:04.530> in 00:14:04.880 --> 00:14:04.890 align:start position:0% long-lived consensus bug we had found in 00:14:04.890 --> 00:14:07.250 align:start position:0% long-lived consensus bug we had found in the<00:14:05.130> block<00:14:05.310> processing<00:14:05.880> engine<00:14:06.270> whereby<00:14:07.020> both 00:14:07.250 --> 00:14:07.260 align:start position:0% the block processing engine whereby both 00:14:07.260 --> 00:14:13.010 align:start position:0% the block processing engine whereby both gas<00:14:07.860> and<00:14:08.360> parity<00:14:09.360> and<00:14:09.650> did<00:14:10.650> it<00:14:10.800> erroneously<00:14:12.020> so 00:14:13.010 --> 00:14:13.020 align:start position:0% gas and parity and did it erroneously so 00:14:13.020 --> 00:14:15.770 align:start position:0% gas and parity and did it erroneously so blocks<00:14:13.260> have<00:14:13.440> time<00:14:13.680> stamps<00:14:13.830> a<00:14:14.390> normal<00:14:15.390> time<00:14:15.630> a 00:14:15.770 --> 00:14:15.780 align:start position:0% blocks have time stamps a normal time a 00:14:15.780 --> 00:14:18.290 align:start position:0% blocks have time stamps a normal time a normal<00:14:16.110> block<00:14:16.800> cannot<00:14:17.430> be<00:14:17.670> too<00:14:17.850> far<00:14:18.120> in<00:14:18.210> the 00:14:18.290 --> 00:14:18.300 align:start position:0% normal block cannot be too far in the 00:14:18.300 --> 00:14:19.940 align:start position:0% normal block cannot be too far in the future<00:14:18.480> because<00:14:18.960> the<00:14:19.080> client<00:14:19.350> won't<00:14:19.470> import 00:14:19.940 --> 00:14:19.950 align:start position:0% future because the client won't import 00:14:19.950 --> 00:14:23.930 align:start position:0% future because the client won't import it<00:14:20.360> however<00:14:21.380> an<00:14:22.380> orphan<00:14:22.860> block<00:14:23.070> or<00:14:23.310> an<00:14:23.580> Omer 00:14:23.930 --> 00:14:23.940 align:start position:0% it however an orphan block or an Omer 00:14:23.940 --> 00:14:26.990 align:start position:0% it however an orphan block or an Omer block<00:14:24.920> doesn't<00:14:25.920> have<00:14:26.070> any<00:14:26.250> restrictions<00:14:26.850> on 00:14:26.990 --> 00:14:27.000 align:start position:0% block doesn't have any restrictions on 00:14:27.000 --> 00:14:29.840 align:start position:0% block doesn't have any restrictions on the<00:14:27.180> time<00:14:27.390> stamp<00:14:27.540> they<00:14:27.840> can<00:14:28.020> be<00:14:28.250> how<00:14:29.250> far<00:14:29.550> much 00:14:29.840 --> 00:14:29.850 align:start position:0% the time stamp they can be how far much 00:14:29.850 --> 00:14:31.700 align:start position:0% the time stamp they can be how far much in<00:14:30.000> the<00:14:30.090> future<00:14:30.240> they<00:14:30.570> want<00:14:30.780> and<00:14:31.140> if<00:14:31.560> they 00:14:31.700 --> 00:14:31.710 align:start position:0% in the future they want and if they 00:14:31.710 --> 00:14:37.340 align:start position:0% in the future they want and if they exceeded<00:14:32.100> the<00:14:32.280> UN<00:14:32.550> 64<00:14:33.210> yes<00:14:33.780> would<00:14:34.140> just<00:14:36.320> yeah 00:14:37.340 --> 00:14:37.350 align:start position:0% exceeded the UN 64 yes would just yeah 00:14:37.350 --> 00:14:40.760 align:start position:0% exceeded the UN 64 yes would just yeah wrap<00:14:37.710> it<00:14:37.860> around<00:14:37.950> on<00:14:38.370> you<00:14:38.970> in<00:14:39.090> 64<00:14:39.630> and<00:14:39.810> parity 00:14:40.760 --> 00:14:40.770 align:start position:0% wrap it around on you in 64 and parity 00:14:40.770 --> 00:14:43.430 align:start position:0% wrap it around on you in 64 and parity would<00:14:40.920> cap<00:14:41.220> it<00:14:41.430> so<00:14:42.210> both<00:14:42.480> would<00:14:42.780> be<00:14:42.840> incorrect 00:14:43.430 --> 00:14:43.440 align:start position:0% would cap it so both would be incorrect 00:14:43.440 --> 00:14:48.190 align:start position:0% would cap it so both would be incorrect and<00:14:43.830> a<00:14:44.370> calculating<00:14:45.120> correct<00:14:45.420> block<00:14:46.140> cache 00:14:48.190 --> 00:14:48.200 align:start position:0% and a calculating correct block cache 00:14:48.200 --> 00:14:52.100 align:start position:0% and a calculating correct block cache and<00:14:49.200> it<00:14:49.980> would<00:14:50.160> cause<00:14:50.340> an<00:14:50.520> a<00:14:51.210> consensus 00:14:52.100 --> 00:14:52.110 align:start position:0% and it would cause an a consensus 00:14:52.110 --> 00:14:55.670 align:start position:0% and it would cause an a consensus failure<00:14:52.640> with<00:14:53.640> the<00:14:53.670> quirk<00:14:54.120> that<00:14:54.480> none<00:14:55.350> of<00:14:55.500> them 00:14:55.670 --> 00:14:55.680 align:start position:0% failure with the quirk that none of them 00:14:55.680 --> 00:14:57.280 align:start position:0% failure with the quirk that none of them would<00:14:56.070> be<00:14:56.100> correct 00:14:57.280 --> 00:14:57.290 align:start position:0% would be correct 00:14:57.290 --> 00:14:59.870 align:start position:0% would be correct this<00:14:58.290> would<00:14:58.530> be<00:14:58.650> kind<00:14:58.890> of<00:14:58.950> hard<00:14:59.250> to<00:14:59.340> exploit 00:14:59.870 --> 00:14:59.880 align:start position:0% this would be kind of hard to exploit 00:14:59.880 --> 00:15:01.970 align:start position:0% this would be kind of hard to exploit because<00:15:00.210> you<00:15:00.480> would<00:15:00.630> need<00:15:00.870> some<00:15:01.200> mind<00:15:01.680> both 00:15:01.970 --> 00:15:01.980 align:start position:0% because you would need some mind both 00:15:01.980 --> 00:15:06.110 align:start position:0% because you would need some mind both the<00:15:02.190> quark<00:15:02.520> uncle<00:15:03.120> and<00:15:03.330> a<00:15:03.660> block<00:15:04.640> within<00:15:05.640> a 00:15:06.110 --> 00:15:06.120 align:start position:0% the quark uncle and a block within a 00:15:06.120 --> 00:15:08.690 align:start position:0% the quark uncle and a block within a period<00:15:06.690> of<00:15:06.840> 7<00:15:07.290> blocks<00:15:07.530> and<00:15:07.980> then<00:15:08.190> roll<00:15:08.400> it<00:15:08.550> out 00:15:08.690 --> 00:15:08.700 align:start position:0% period of 7 blocks and then roll it out 00:15:08.700 --> 00:15:10.580 align:start position:0% period of 7 blocks and then roll it out to<00:15:08.880> maintenance<00:15:09.150> so<00:15:09.840> you'd<00:15:10.050> have<00:15:10.080> to<00:15:10.290> have 00:15:10.580 --> 00:15:10.590 align:start position:0% to maintenance so you'd have to have 00:15:10.590 --> 00:15:12.500 align:start position:0% to maintenance so you'd have to have quite<00:15:11.220> a<00:15:11.280> lot<00:15:11.490> of<00:15:11.610> mining<00:15:11.970> power<00:15:12.180> to<00:15:12.390> actually 00:15:12.500 --> 00:15:12.510 align:start position:0% quite a lot of mining power to actually 00:15:12.510 --> 00:15:17.150 align:start position:0% quite a lot of mining power to actually do<00:15:12.810> this<00:15:12.930> kind<00:15:13.170> of<00:15:13.260> attack<00:15:15.680> I'm<00:15:16.680> going<00:15:16.980> to<00:15:17.070> have 00:15:17.150 --> 00:15:17.160 align:start position:0% do this kind of attack I'm going to have 00:15:17.160 --> 00:15:20.360 align:start position:0% do this kind of attack I'm going to have to<00:15:17.250> skip<00:15:17.520> a<00:15:17.580> bit<00:15:17.850> in<00:15:18.030> the<00:15:18.150> EVM<00:15:18.420> portion<00:15:19.160> most<00:15:20.160> of 00:15:20.360 --> 00:15:20.370 align:start position:0% to skip a bit in the EVM portion most of 00:15:20.370 --> 00:15:24.950 align:start position:0% to skip a bit in the EVM portion most of you<00:15:21.710> might<00:15:22.710> know<00:15:22.950> some<00:15:23.220> of<00:15:23.340> these<00:15:23.460> so<00:15:24.270> Shanghai 00:15:24.950 --> 00:15:24.960 align:start position:0% you might know some of these so Shanghai 00:15:24.960 --> 00:15:26.060 align:start position:0% you might know some of these so Shanghai attacks<00:15:25.290> arrived<00:15:25.620> which 00:15:26.060 --> 00:15:26.070 align:start position:0% attacks arrived which 00:15:26.070 --> 00:15:29.150 align:start position:0% attacks arrived which mentioned<00:15:26.400> a<00:15:26.460> bit<00:15:26.610> caused<00:15:27.300> a<00:15:27.330> bit<00:15:27.660> of<00:15:28.160> problems 00:15:29.150 --> 00:15:29.160 align:start position:0% mentioned a bit caused a bit of problems 00:15:29.160 --> 00:15:34.340 align:start position:0% mentioned a bit caused a bit of problems for<00:15:29.310> us<00:15:30.260> during<00:15:31.260> the<00:15:31.560> about<00:15:31.890> the<00:15:32.070> month<00:15:33.350> there 00:15:34.340 --> 00:15:34.350 align:start position:0% for us during the about the month there 00:15:34.350 --> 00:15:36.550 align:start position:0% for us during the about the month there were<00:15:34.440> denial<00:15:34.770> of<00:15:34.860> service<00:15:35.160> attacks<00:15:35.520> because 00:15:36.550 --> 00:15:36.560 align:start position:0% were denial of service attacks because 00:15:36.560 --> 00:15:39.710 align:start position:0% were denial of service attacks because first<00:15:37.560> of<00:15:37.830> all<00:15:37.950> in<00:15:38.760> optimal<00:15:39.330> suboptimal 00:15:39.710 --> 00:15:39.720 align:start position:0% first of all in optimal suboptimal 00:15:39.720 --> 00:15:41.810 align:start position:0% first of all in optimal suboptimal client<00:15:40.290> with<00:15:40.380> limitations<00:15:40.710> and<00:15:41.070> secondly 00:15:41.810 --> 00:15:41.820 align:start position:0% client with limitations and secondly 00:15:41.820 --> 00:15:45.500 align:start position:0% client with limitations and secondly because<00:15:42.030> of<00:15:42.210> gas<00:15:43.110> miss<00:15:44.040> pricings<00:15:44.640> and<00:15:44.970> there 00:15:45.500 --> 00:15:45.510 align:start position:0% because of gas miss pricings and there 00:15:45.510 --> 00:15:47.090 align:start position:0% because of gas miss pricings and there were<00:15:45.600> two<00:15:45.810> hard<00:15:46.080> Forks<00:15:46.440> that<00:15:46.560> followed<00:15:46.950> on 00:15:47.090 --> 00:15:47.100 align:start position:0% were two hard Forks that followed on 00:15:47.100 --> 00:15:50.060 align:start position:0% were two hard Forks that followed on that<00:15:47.310> and<00:15:47.640> there<00:15:48.600> was<00:15:48.720> a<00:15:48.780> naina<00:15:49.080> chain<00:15:49.560> splits 00:15:50.060 --> 00:15:50.070 align:start position:0% that and there was a naina chain splits 00:15:50.070 --> 00:15:53.720 align:start position:0% that and there was a naina chain splits because<00:15:50.700> of<00:15:51.180> how<00:15:52.050> the<00:15:52.430> reversion<00:15:53.430> was 00:15:53.720 --> 00:15:53.730 align:start position:0% because of how the reversion was 00:15:53.730 --> 00:15:58.130 align:start position:0% because of how the reversion was implemented<00:15:55.670> and<00:15:56.670> how<00:15:56.730> that<00:15:57.570> whole<00:15:57.780> change 00:15:58.130 --> 00:15:58.140 align:start position:0% implemented and how that whole change 00:15:58.140 --> 00:16:01.280 align:start position:0% implemented and how that whole change that<00:15:58.440> happened<00:15:59.270> we<00:16:00.270> know<00:16:00.420> how<00:16:00.570> it<00:16:00.720> happened<00:16:01.170> we 00:16:01.280 --> 00:16:01.290 align:start position:0% that happened we know how it happened we 00:16:01.290 --> 00:16:03.320 align:start position:0% that happened we know how it happened we haven't<00:16:01.590> really<00:16:02.040> figured<00:16:02.370> out<00:16:02.580> a<00:16:02.640> good<00:16:03.000> rule 00:16:03.320 --> 00:16:03.330 align:start position:0% haven't really figured out a good rule 00:16:03.330 --> 00:16:05.090 align:start position:0% haven't really figured out a good rule to<00:16:03.570> explain<00:16:04.020> how<00:16:04.200> it<00:16:04.350> happened<00:16:04.380> and<00:16:04.800> put<00:16:04.980> it 00:16:05.090 --> 00:16:05.100 align:start position:0% to explain how it happened and put it 00:16:05.100 --> 00:16:07.760 align:start position:0% to explain how it happened and put it into<00:16:05.370> the<00:16:05.490> yellow<00:16:05.640> paper<00:16:06.440> right<00:16:07.440> now<00:16:07.470> the 00:16:07.760 --> 00:16:07.770 align:start position:0% into the yellow paper right now the 00:16:07.770 --> 00:16:10.070 align:start position:0% into the yellow paper right now the situation<00:16:08.430> is<00:16:09.000> that<00:16:09.030> all<00:16:09.330> the<00:16:09.450> clients<00:16:09.900> have 00:16:10.070 --> 00:16:10.080 align:start position:0% situation is that all the clients have 00:16:10.080 --> 00:16:14.480 align:start position:0% situation is that all the clients have particular<00:16:10.890> right<00:16:11.490> and<00:16:11.760> DS<00:16:12.090> special<00:16:12.810> rules<00:16:13.490> so 00:16:14.480 --> 00:16:14.490 align:start position:0% particular right and DS special rules so 00:16:14.490 --> 00:16:16.970 align:start position:0% particular right and DS special rules so you<00:16:14.940> can<00:16:15.120> see<00:16:15.300> here<00:16:15.540> a<00:16:15.570> bit<00:16:15.690> of<00:16:15.930> code<00:16:16.470> from<00:16:16.800> the 00:16:16.970 --> 00:16:16.980 align:start position:0% you can see here a bit of code from the 00:16:16.980 --> 00:16:19.040 align:start position:0% you can see here a bit of code from the Goethe<00:16:17.400> areum<00:16:17.790> codebase<00:16:18.240> and<00:16:18.270> get<00:16:18.750> up<00:16:18.870> code 00:16:19.040 --> 00:16:19.050 align:start position:0% Goethe areum codebase and get up code 00:16:19.050 --> 00:16:22.190 align:start position:0% Goethe areum codebase and get up code from<00:16:19.260> the<00:16:19.440> aleph<00:16:20.400> codebase<00:16:20.970> and<00:16:21.330> they 00:16:22.190 --> 00:16:22.200 align:start position:0% from the aleph codebase and they 00:16:22.200 --> 00:16:25.790 align:start position:0% from the aleph codebase and they basically<00:16:22.590> just<00:16:22.950> oh<00:16:23.840> we<00:16:24.840> fix<00:16:25.200> this<00:16:25.410> a<00:16:25.470> little 00:16:25.790 --> 00:16:25.800 align:start position:0% basically just oh we fix this a little 00:16:25.800 --> 00:16:27.560 align:start position:0% basically just oh we fix this a little bit<00:16:26.040> it's<00:16:26.220> not<00:16:26.340> it's<00:16:26.760> not<00:16:26.910> don't<00:16:27.210> look<00:16:27.360> at<00:16:27.480> it 00:16:27.560 --> 00:16:27.570 align:start position:0% bit it's not it's not don't look at it 00:16:27.570 --> 00:16:29.830 align:start position:0% bit it's not it's not don't look at it this<00:16:27.720> is<00:16:27.780> ugly<00:16:28.280> we<00:16:29.280> just<00:16:29.520> know<00:16:29.730> that 00:16:29.830 --> 00:16:29.840 align:start position:0% this is ugly we just know that 00:16:29.840 --> 00:16:32.150 align:start position:0% this is ugly we just know that fortunately<00:16:30.840> this<00:16:31.020> is<00:16:31.200> this<00:16:31.500> exact<00:16:31.950> situation 00:16:32.150 --> 00:16:32.160 align:start position:0% fortunately this is this exact situation 00:16:32.160 --> 00:16:35.380 align:start position:0% fortunately this is this exact situation will<00:16:32.760> never<00:16:32.970> again<00:16:33.240> arise<00:16:34.040> because<00:16:35.040> the 00:16:35.380 --> 00:16:35.390 align:start position:0% will never again arise because the 00:16:35.390 --> 00:16:37.760 align:start position:0% will never again arise because the conditions<00:16:36.390> that<00:16:36.630> would<00:16:36.900> enable<00:16:37.290> this<00:16:37.560> to 00:16:37.760 --> 00:16:37.770 align:start position:0% conditions that would enable this to 00:16:37.770 --> 00:16:39.590 align:start position:0% conditions that would enable this to rise<00:16:37.980> can<00:16:38.400> never<00:16:38.610> again<00:16:39.000> happen<00:16:39.390> on<00:16:39.480> the 00:16:39.590 --> 00:16:39.600 align:start position:0% rise can never again happen on the 00:16:39.600 --> 00:16:44.690 align:start position:0% rise can never again happen on the maintenance<00:16:42.170> in<00:16:43.170> January<00:16:43.710> I<00:16:44.010> had<00:16:44.160> to<00:16:44.400> actually 00:16:44.690 --> 00:16:44.700 align:start position:0% maintenance in January I had to actually 00:16:44.700 --> 00:16:46.430 align:start position:0% maintenance in January I had to actually remove<00:16:44.940> this<00:16:45.120> slide<00:16:45.360> because<00:16:45.570> each<00:16:45.990> see<00:16:46.260> was 00:16:46.430 --> 00:16:46.440 align:start position:0% remove this slide because each see was 00:16:46.440 --> 00:16:52.120 align:start position:0% remove this slide because each see was still<00:16:46.770> vulnerable<00:16:47.190> last<00:16:47.820> Friday<00:16:50.690> in<00:16:51.690> February 00:16:52.120 --> 00:16:52.130 align:start position:0% still vulnerable last Friday in February 00:16:52.130 --> 00:16:54.650 align:start position:0% still vulnerable last Friday in February 2017<00:16:53.130> that<00:16:53.430> was<00:16:53.580> really<00:16:53.910> interesting<00:16:54.360> a 00:16:54.650 --> 00:16:54.660 align:start position:0% 2017 that was really interesting a 00:16:54.660 --> 00:16:57.290 align:start position:0% 2017 that was really interesting a consensus<00:16:55.200> bug<00:16:55.440> which<00:16:55.950> was<00:16:56.760> a<00:16:56.790> combo 00:16:57.290 --> 00:16:57.300 align:start position:0% consensus bug which was a combo 00:16:57.300 --> 00:16:59.030 align:start position:0% consensus bug which was a combo consensus<00:16:57.870> bug<00:16:58.080> that<00:16:58.260> could<00:16:58.410> either<00:16:58.620> be<00:16:58.770> used 00:16:59.030 --> 00:16:59.040 align:start position:0% consensus bug that could either be used 00:16:59.040 --> 00:17:02.150 align:start position:0% consensus bug that could either be used to<00:16:59.550> cause<00:16:59.820> an<00:17:00.030> instant<00:17:00.420> chain<00:17:00.630> split<00:17:00.960> or<00:17:01.650> it 00:17:02.150 --> 00:17:02.160 align:start position:0% to cause an instant chain split or it 00:17:02.160 --> 00:17:05.530 align:start position:0% to cause an instant chain split or it could<00:17:02.400> be<00:17:02.550> used<00:17:02.700> to<00:17:02.970> crash<00:17:03.450> all<00:17:03.750> the<00:17:03.990> gif<00:17:04.170> nodes 00:17:05.530 --> 00:17:05.540 align:start position:0% could be used to crash all the gif nodes 00:17:05.540 --> 00:17:07.880 align:start position:0% could be used to crash all the gif nodes so<00:17:06.540> the<00:17:06.720> stack<00:17:06.990> requirements<00:17:07.740> were 00:17:07.880 --> 00:17:07.890 align:start position:0% so the stack requirements were 00:17:07.890 --> 00:17:10.010 align:start position:0% so the stack requirements were misconfigured<00:17:08.640> for<00:17:08.970> the<00:17:09.060> swap<00:17:09.329> top<00:17:09.689> and 00:17:10.010 --> 00:17:10.020 align:start position:0% misconfigured for the swap top and 00:17:10.020 --> 00:17:13.429 align:start position:0% misconfigured for the swap top and balance<00:17:10.680> of<00:17:10.860> op<00:17:11.280> codes<00:17:11.839> and<00:17:12.839> the<00:17:12.959> thing<00:17:13.170> is<00:17:13.290> if 00:17:13.429 --> 00:17:13.439 align:start position:0% balance of op codes and the thing is if 00:17:13.439 --> 00:17:15.319 align:start position:0% balance of op codes and the thing is if the<00:17:13.560> stack<00:17:13.829> requirements<00:17:14.430> are<00:17:14.520> off<00:17:14.670> if<00:17:15.060> it 00:17:15.319 --> 00:17:15.329 align:start position:0% the stack requirements are off if it 00:17:15.329 --> 00:17:17.360 align:start position:0% the stack requirements are off if it leads<00:17:15.510> to<00:17:15.689> I'm<00:17:16.350> not<00:17:16.500> talking<00:17:16.709> about<00:17:16.740> the<00:17:16.949> EVM 00:17:17.360 --> 00:17:17.370 align:start position:0% leads to I'm not talking about the EVM 00:17:17.370 --> 00:17:20.510 align:start position:0% leads to I'm not talking about the EVM stack<00:17:17.730> not<00:17:17.970> like<00:17:18.270> an<00:17:18.420> ALICE<00:17:18.689> type<00:17:19.220> if<00:17:20.220> it's<00:17:20.430> an 00:17:20.510 --> 00:17:20.520 align:start position:0% stack not like an ALICE type if it's an 00:17:20.520 --> 00:17:23.390 align:start position:0% stack not like an ALICE type if it's an underflow<00:17:21.170> death<00:17:22.170> would<00:17:22.410> immediately<00:17:22.740> panic 00:17:23.390 --> 00:17:23.400 align:start position:0% underflow death would immediately panic 00:17:23.400 --> 00:17:25.040 align:start position:0% underflow death would immediately panic in<00:17:23.550> all<00:17:23.699> the<00:17:23.820> nodes<00:17:24.030> that<00:17:24.480> were<00:17:24.630> processing 00:17:25.040 --> 00:17:25.050 align:start position:0% in all the nodes that were processing 00:17:25.050 --> 00:17:26.540 align:start position:0% in all the nodes that were processing this<00:17:25.290> block<00:17:25.560> all<00:17:25.770> the<00:17:25.829> gift<00:17:26.040> notes<00:17:26.280> would 00:17:26.540 --> 00:17:26.550 align:start position:0% this block all the gift notes would 00:17:26.550 --> 00:17:28.010 align:start position:0% this block all the gift notes would immediately<00:17:27.060> drop<00:17:27.270> off<00:17:27.420> drop<00:17:27.780> off<00:17:27.900> the 00:17:28.010 --> 00:17:28.020 align:start position:0% immediately drop off drop off the 00:17:28.020 --> 00:17:30.350 align:start position:0% immediately drop off drop off the network<00:17:28.199> if<00:17:29.010> it<00:17:29.190> that<00:17:29.370> was<00:17:29.520> like<00:17:29.700> overflow 00:17:30.350 --> 00:17:30.360 align:start position:0% network if it that was like overflow 00:17:30.360 --> 00:17:32.210 align:start position:0% network if it that was like overflow that<00:17:30.870> will<00:17:31.110> lead<00:17:31.230> to<00:17:31.260> a<00:17:31.380> consensus<00:17:32.010> issue 00:17:32.210 --> 00:17:32.220 align:start position:0% that will lead to a consensus issue 00:17:32.220 --> 00:17:33.950 align:start position:0% that will lead to a consensus issue where<00:17:32.670> GAF<00:17:32.850> had<00:17:33.060> a<00:17:33.090> different<00:17:33.480> state<00:17:33.720> route 00:17:33.950 --> 00:17:33.960 align:start position:0% where GAF had a different state route 00:17:33.960 --> 00:17:34.700 align:start position:0% where GAF had a different state route and<00:17:34.170> parity 00:17:34.700 --> 00:17:34.710 align:start position:0% and parity 00:17:34.710 --> 00:17:39.169 align:start position:0% and parity and<00:17:35.480> you<00:17:36.480> would<00:17:36.600> split<00:17:36.840> the<00:17:36.870> chain<00:17:37.880> so<00:17:38.880> any 00:17:39.169 --> 00:17:39.179 align:start position:0% and you would split the chain so any 00:17:39.179 --> 00:17:40.669 align:start position:0% and you would split the chain so any attacker<00:17:39.539> which<00:17:39.809> found<00:17:40.080> this<00:17:40.289> could<00:17:40.529> just 00:17:40.669 --> 00:17:40.679 align:start position:0% attacker which found this could just 00:17:40.679 --> 00:17:43.279 align:start position:0% attacker which found this could just choose<00:17:41.159> which<00:17:41.909> route<00:17:42.149> to<00:17:42.360> go<00:17:42.510> and<00:17:42.750> it<00:17:43.110> would 00:17:43.279 --> 00:17:43.289 align:start position:0% choose which route to go and it would 00:17:43.289 --> 00:17:45.230 align:start position:0% choose which route to go and it would have<00:17:43.350> been<00:17:43.500> trivial<00:17:43.799> to<00:17:44.100> exploit<00:17:44.460> and<00:17:44.760> trivial 00:17:45.230 --> 00:17:45.240 align:start position:0% have been trivial to exploit and trivial 00:17:45.240 --> 00:17:48.190 align:start position:0% have been trivial to exploit and trivial to<00:17:45.390> detect<00:17:45.840> if<00:17:46.140> it'd<00:17:46.320> have<00:17:46.440> been<00:17:46.470> past<00:17:46.860> naively 00:17:48.190 --> 00:17:48.200 align:start position:0% to detect if it'd have been past naively 00:17:48.200 --> 00:17:53.029 align:start position:0% to detect if it'd have been past naively so<00:17:49.200> what<00:17:49.350> we<00:17:49.500> did<00:17:49.679> was<00:17:49.860> we<00:17:50.240> read<00:17:51.799> refactor<00:17:52.799> the 00:17:53.029 --> 00:17:53.039 align:start position:0% so what we did was we read refactor the 00:17:53.039 --> 00:17:56.240 align:start position:0% so what we did was we read refactor the entire<00:17:53.450> code<00:17:54.450> that<00:17:54.720> that's<00:17:55.200> configured<00:17:55.860> the 00:17:56.240 --> 00:17:56.250 align:start position:0% entire code that that's configured the 00:17:56.250 --> 00:17:59.810 align:start position:0% entire code that that's configured the EVM<00:17:56.669> of<00:17:56.970> codes<00:17:57.390> and<00:17:57.659> two<00:17:58.529> weeks<00:17:58.710> later<00:17:58.830> we 00:17:59.810 --> 00:17:59.820 align:start position:0% EVM of codes and two weeks later we 00:17:59.820 --> 00:18:03.019 align:start position:0% EVM of codes and two weeks later we reversed<00:18:00.240> the<00:18:00.630> fix<00:18:01.309> reversed<00:18:02.309> the<00:18:02.580> the<00:18:02.820> whole 00:18:03.019 --> 00:18:03.029 align:start position:0% reversed the fix reversed the the whole 00:18:03.029 --> 00:18:12.049 align:start position:0% reversed the fix reversed the the whole refactoring<00:18:03.720> but<00:18:03.960> left<00:18:04.140> the<00:18:04.320> fix<00:18:04.559> in<00:18:09.950> in<00:18:11.059> 2017 00:18:12.049 --> 00:18:12.059 align:start position:0% refactoring but left the fix in in 2017 00:18:12.059 --> 00:18:14.840 align:start position:0% refactoring but left the fix in in 2017 some<00:18:13.020> attackers<00:18:13.440> were<00:18:13.559> hitting<00:18:13.860> the<00:18:14.010> gas<00:18:14.340> jump 00:18:14.840 --> 00:18:14.850 align:start position:0% some attackers were hitting the gas jump 00:18:14.850 --> 00:18:17.210 align:start position:0% some attackers were hitting the gas jump test<00:18:15.090> analysis<00:18:15.809> so<00:18:16.440> we<00:18:16.590> saw<00:18:16.770> on<00:18:16.919> our<00:18:17.039> live 00:18:17.210 --> 00:18:17.220 align:start position:0% test analysis so we saw on our live 00:18:17.220 --> 00:18:19.399 align:start position:0% test analysis so we saw on our live monitoring<00:18:17.669> system<00:18:18.000> that<00:18:18.029> some<00:18:18.960> blocks<00:18:19.200> took 00:18:19.399 --> 00:18:19.409 align:start position:0% monitoring system that some blocks took 00:18:19.409 --> 00:18:28.310 align:start position:0% monitoring system that some blocks took like<00:18:19.590> 24<00:18:20.100> seconds<00:18:21.080> to<00:18:22.080> go<00:18:22.260> through<00:18:24.950> and<00:18:27.320> yeah 00:18:28.310 --> 00:18:28.320 align:start position:0% like 24 seconds to go through and yeah 00:18:28.320 --> 00:18:31.220 align:start position:0% like 24 seconds to go through and yeah so<00:18:28.830> we<00:18:29.309> had<00:18:29.429> to<00:18:29.549> rework<00:18:29.880> the<00:18:30.059> whole<00:18:30.179> jump<00:18:31.049> test 00:18:31.220 --> 00:18:31.230 align:start position:0% so we had to rework the whole jump test 00:18:31.230 --> 00:18:33.440 align:start position:0% so we had to rework the whole jump test analysis<00:18:31.799> very<00:18:32.010> quickly<00:18:32.399> bring<00:18:33.149> it<00:18:33.270> down<00:18:33.419> a 00:18:33.440 --> 00:18:33.450 align:start position:0% analysis very quickly bring it down a 00:18:33.450 --> 00:18:34.970 align:start position:0% analysis very quickly bring it down a couple<00:18:33.809> of<00:18:33.870> orders<00:18:34.169> of<00:18:34.320> magnitude<00:18:34.830> in 00:18:34.970 --> 00:18:34.980 align:start position:0% couple of orders of magnitude in 00:18:34.980 --> 00:18:37.639 align:start position:0% couple of orders of magnitude in processing<00:18:35.490> time<00:18:35.730> and<00:18:36.169> there<00:18:37.169> were<00:18:37.289> some 00:18:37.639 --> 00:18:37.649 align:start position:0% processing time and there were some 00:18:37.649 --> 00:18:40.159 align:start position:0% processing time and there were some ideas<00:18:38.520> that<00:18:39.090> this<00:18:39.210> was<00:18:39.450> possibly<00:18:40.049> the 00:18:40.159 --> 00:18:40.169 align:start position:0% ideas that this was possibly the 00:18:40.169 --> 00:18:42.769 align:start position:0% ideas that this was possibly the Shanghai<00:18:40.679> attacker<00:18:41.070> doing<00:18:41.730> this<00:18:41.909> we<00:18:42.600> don't 00:18:42.769 --> 00:18:42.779 align:start position:0% Shanghai attacker doing this we don't 00:18:42.779 --> 00:18:47.180 align:start position:0% Shanghai attacker doing this we don't know<00:18:43.020> I<00:18:44.750> don't<00:18:45.750> have<00:18:45.960> to<00:18:46.140> have<00:18:46.649> time<00:18:46.890> to<00:18:47.070> go 00:18:47.180 --> 00:18:47.190 align:start position:0% know I don't have to have time to go 00:18:47.190 --> 00:18:48.769 align:start position:0% know I don't have to have time to go through<00:18:47.220> the<00:18:47.580> whole<00:18:47.700> byzantium<00:18:48.390> but<00:18:48.570> in 00:18:48.769 --> 00:18:48.779 align:start position:0% through the whole byzantium but in 00:18:48.779 --> 00:18:50.930 align:start position:0% through the whole byzantium but in bicenthium<00:18:49.260> when<00:18:50.159> we<00:18:50.309> were<00:18:50.429> approaching<00:18:50.760> that 00:18:50.930 --> 00:18:50.940 align:start position:0% bicenthium when we were approaching that 00:18:50.940 --> 00:18:53.779 align:start position:0% bicenthium when we were approaching that we<00:18:51.539> really<00:18:51.990> started<00:18:52.440> getting<00:18:53.010> payoff<00:18:53.549> from 00:18:53.779 --> 00:18:53.789 align:start position:0% we really started getting payoff from 00:18:53.789 --> 00:18:56.720 align:start position:0% we really started getting payoff from passing<00:18:54.510> and<00:18:54.750> it<00:18:55.409> found<00:18:55.679> all<00:18:56.100> these<00:18:56.370> issues 00:18:56.720 --> 00:18:56.730 align:start position:0% passing and it found all these issues 00:18:56.730 --> 00:18:59.000 align:start position:0% passing and it found all these issues and<00:18:57.270> you<00:18:57.360> can<00:18:57.990> see<00:18:58.110> the<00:18:58.260> last<00:18:58.500> one<00:18:58.649> here 00:18:59.000 --> 00:18:59.010 align:start position:0% and you can see the last one here 00:18:59.010 --> 00:19:01.909 align:start position:0% and you can see the last one here happened<00:18:59.789> on<00:18:59.850> October<00:18:59.970> 25<00:19:00.870> that's<00:19:01.289> actually 00:19:01.909 --> 00:19:01.919 align:start position:0% happened on October 25 that's actually 00:19:01.919 --> 00:19:04.360 align:start position:0% happened on October 25 that's actually nine<00:19:02.159> days<00:19:02.370> after<00:19:02.580> buy<00:19:03.149> something<00:19:03.570> for<00:19:03.750> kids 00:19:04.360 --> 00:19:04.370 align:start position:0% nine days after buy something for kids 00:19:04.370 --> 00:19:08.450 align:start position:0% nine days after buy something for kids and<00:19:05.370> not<00:19:05.700> only<00:19:05.880> that<00:19:06.059> not<00:19:06.899> until<00:19:07.200> February<00:19:07.799> did 00:19:08.450 --> 00:19:08.460 align:start position:0% and not only that not until February did 00:19:08.460 --> 00:19:10.370 align:start position:0% and not only that not until February did we<00:19:08.610> uncover<00:19:08.940> the<00:19:09.179> hopefully<00:19:09.659> last<00:19:09.870> bicenthium 00:19:10.370 --> 00:19:10.380 align:start position:0% we uncover the hopefully last bicenthium 00:19:10.380 --> 00:19:11.000 align:start position:0% we uncover the hopefully last bicenthium bug 00:19:11.000 --> 00:19:11.010 align:start position:0% bug 00:19:11.010 --> 00:19:12.769 align:start position:0% bug it<00:19:11.100> was<00:19:11.250> Cassidy<00:19:11.760> drew<00:19:11.940> found<00:19:12.240> an<00:19:12.390> issue<00:19:12.510> with 00:19:12.769 --> 00:19:12.779 align:start position:0% it was Cassidy drew found an issue with 00:19:12.779 --> 00:19:17.990 align:start position:0% it was Cassidy drew found an issue with our<00:19:13.520> biometric<00:19:14.520> pairing<00:19:16.520> which<00:19:17.520> was<00:19:17.730> we 00:19:17.990 --> 00:19:18.000 align:start position:0% our biometric pairing which was we 00:19:18.000 --> 00:19:19.909 align:start position:0% our biometric pairing which was we couldn't<00:19:18.240> actually<00:19:18.690> fix<00:19:19.200> it<00:19:19.409> until<00:19:19.590> a<00:19:19.799> couple 00:19:19.909 --> 00:19:19.919 align:start position:0% couldn't actually fix it until a couple 00:19:19.919 --> 00:19:21.409 align:start position:0% couldn't actually fix it until a couple of<00:19:20.039> months<00:19:20.220> later<00:19:20.460> when<00:19:20.760> we<00:19:20.850> swapped<00:19:21.149> out<00:19:21.390> the 00:19:21.409 --> 00:19:21.419 align:start position:0% of months later when we swapped out the 00:19:21.419 --> 00:19:24.620 align:start position:0% of months later when we swapped out the whole<00:19:21.779> VM<00:19:22.710> library<00:19:23.370> for<00:19:23.640> an<00:19:23.850> assembly<00:19:24.179> version 00:19:24.620 --> 00:19:24.630 align:start position:0% whole VM library for an assembly version 00:19:24.630 --> 00:19:29.210 align:start position:0% whole VM library for an assembly version written<00:19:24.809> by<00:19:25.110> CloudFlare<00:19:27.710> some<00:19:28.710> takeaways 00:19:29.210 --> 00:19:29.220 align:start position:0% written by CloudFlare some takeaways 00:19:29.220 --> 00:19:32.659 align:start position:0% written by CloudFlare some takeaways from<00:19:29.429> this<00:19:29.669> is<00:19:30.000> that<00:19:30.860> monitoring<00:19:31.860> and<00:19:32.100> metrics 00:19:32.659 --> 00:19:32.669 align:start position:0% from this is that monitoring and metrics 00:19:32.669 --> 00:19:35.690 align:start position:0% from this is that monitoring and metrics and<00:19:32.909> graphing<00:19:33.390> up<00:19:33.539> things<00:19:33.990> really<00:19:34.710> helps<00:19:35.460> to 00:19:35.690 --> 00:19:35.700 align:start position:0% and graphing up things really helps to 00:19:35.700 --> 00:19:38.120 align:start position:0% and graphing up things really helps to see<00:19:35.940> the<00:19:36.120> overall<00:19:36.419> network<00:19:36.840> health<00:19:37.140> and<00:19:37.409> see 00:19:38.120 --> 00:19:38.130 align:start position:0% see the overall network health and see 00:19:38.130 --> 00:19:39.680 align:start position:0% see the overall network health and see what<00:19:38.370> problems<00:19:38.700> there<00:19:39.210> are<00:19:39.390> and<00:19:39.570> what 00:19:39.680 --> 00:19:39.690 align:start position:0% what problems there are and what 00:19:39.690 --> 00:19:41.600 align:start position:0% what problems there are and what problems<00:19:40.140> there<00:19:40.289> might<00:19:40.470> be<00:19:40.500> in<00:19:40.860> PTC<00:19:41.309> loggers 00:19:41.600 --> 00:19:41.610 align:start position:0% problems there might be in PTC loggers 00:19:41.610 --> 00:19:44.240 align:start position:0% problems there might be in PTC loggers giving<00:19:41.909> a<00:19:41.970> talk<00:19:42.179> later<00:19:42.450> about<00:19:43.100> monitoring<00:19:44.100> in 00:19:44.240 --> 00:19:44.250 align:start position:0% giving a talk later about monitoring in 00:19:44.250 --> 00:19:45.370 align:start position:0% giving a talk later about monitoring in live<00:19:44.399> with<00:19:44.549> her 00:19:45.370 --> 00:19:45.380 align:start position:0% live with her 00:19:45.380 --> 00:19:47.950 align:start position:0% live with her here's<00:19:45.740> a<00:19:45.860> chart<00:19:46.190> that<00:19:46.340> can<00:19:46.730> display<00:19:47.450> very 00:19:47.950 --> 00:19:47.960 align:start position:0% here's a chart that can display very 00:19:47.960 --> 00:19:50.470 align:start position:0% here's a chart that can display very detailed<00:19:48.620> what<00:19:48.950> happens<00:19:49.490> when<00:19:49.880> we<00:19:50.000> process 00:19:50.470 --> 00:19:50.480 align:start position:0% detailed what happens when we process 00:19:50.480 --> 00:19:52.930 align:start position:0% detailed what happens when we process the<00:19:50.690> blocks<00:19:50.990> and<00:19:51.350> exactly<00:19:52.190> where<00:19:52.460> the<00:19:52.640> time<00:19:52.910> is 00:19:52.930 --> 00:19:52.940 align:start position:0% the blocks and exactly where the time is 00:19:52.940 --> 00:19:54.760 align:start position:0% the blocks and exactly where the time is spent<00:19:53.150> and<00:19:53.690> these<00:19:53.810> kind<00:19:54.080> of<00:19:54.170> things<00:19:54.410> really 00:19:54.760 --> 00:19:54.770 align:start position:0% spent and these kind of things really 00:19:54.770 --> 00:19:57.990 align:start position:0% spent and these kind of things really help<00:19:55.160> us<00:19:55.370> to<00:19:55.880> find<00:19:56.330> what's<00:19:56.960> going<00:19:57.290> on 00:19:57.990 --> 00:19:58.000 align:start position:0% help us to find what's going on 00:19:58.000 --> 00:20:00.250 align:start position:0% help us to find what's going on similarly<00:19:59.000> for<00:19:59.210> the<00:19:59.300> network<00:19:59.630> traffic's<00:20:00.140> you 00:20:00.250 --> 00:20:00.260 align:start position:0% similarly for the network traffic's you 00:20:00.260 --> 00:20:02.830 align:start position:0% similarly for the network traffic's you can<00:20:00.740> pair<00:20:00.980> with<00:20:01.130> a<00:20:01.160> naive<00:20:01.910> one<00:20:02.120> which<00:20:02.330> I<00:20:02.600> showed 00:20:02.830 --> 00:20:02.840 align:start position:0% can pair with a naive one which I showed 00:20:02.840 --> 00:20:06.660 align:start position:0% can pair with a naive one which I showed earlier<00:20:02.960> here's<00:20:03.350> a<00:20:03.410> much<00:20:03.590> more<00:20:03.740> detailed<00:20:03.950> one 00:20:06.660 --> 00:20:06.670 align:start position:0% 00:20:06.670 --> 00:20:09.010 align:start position:0% yeah<00:20:07.670> I<00:20:07.700> mean<00:20:07.880> of<00:20:08.120> course<00:20:08.300> it's<00:20:08.540> really<00:20:08.870> good 00:20:09.010 --> 00:20:09.020 align:start position:0% yeah I mean of course it's really good 00:20:09.020 --> 00:20:12.910 align:start position:0% yeah I mean of course it's really good tooling<00:20:09.470> to<00:20:10.010> to<00:20:10.580> find<00:20:11.150> out<00:20:11.410> what<00:20:12.410> happened 00:20:12.910 --> 00:20:12.920 align:start position:0% tooling to to find out what happened 00:20:12.920 --> 00:20:14.650 align:start position:0% tooling to to find out what happened when<00:20:13.160> something<00:20:13.610> happened<00:20:14.030> and<00:20:14.240> why<00:20:14.630> it 00:20:14.650 --> 00:20:14.660 align:start position:0% when something happened and why it 00:20:14.660 --> 00:20:18.240 align:start position:0% when something happened and why it happened<00:20:14.810> and<00:20:16.630> we<00:20:17.630> need<00:20:17.780> good<00:20:17.810> communication 00:20:18.240 --> 00:20:18.250 align:start position:0% happened and we need good communication 00:20:18.250 --> 00:20:20.890 align:start position:0% happened and we need good communication clients<00:20:19.250> between<00:20:19.400> teams<00:20:19.970> to<00:20:20.600> quickly 00:20:20.890 --> 00:20:20.900 align:start position:0% clients between teams to quickly 00:20:20.900 --> 00:20:23.320 align:start position:0% clients between teams to quickly coordinate<00:20:21.560> to<00:20:21.980> fix<00:20:22.220> stuff<00:20:22.520> when<00:20:22.760> when 00:20:23.320 --> 00:20:23.330 align:start position:0% coordinate to fix stuff when when 00:20:23.330 --> 00:20:26.860 align:start position:0% coordinate to fix stuff when when goes<00:20:23.510> down<00:20:23.570> on<00:20:23.870> the<00:20:23.960> minute<00:20:24.260> and<00:20:25.660> one<00:20:26.660> thing 00:20:26.860 --> 00:20:26.870 align:start position:0% goes down on the minute and one thing 00:20:26.870 --> 00:20:28.210 align:start position:0% goes down on the minute and one thing that's<00:20:27.140> been<00:20:27.290> really<00:20:27.470> important<00:20:27.920> from<00:20:28.040> day 00:20:28.210 --> 00:20:28.220 align:start position:0% that's been really important from day 00:20:28.220 --> 00:20:29.920 align:start position:0% that's been really important from day one<00:20:28.430> and<00:20:28.640> I<00:20:28.730> think<00:20:28.940> it<00:20:29.030> really<00:20:29.210> has<00:20:29.630> been<00:20:29.870> great 00:20:29.920 --> 00:20:29.930 align:start position:0% one and I think it really has been great 00:20:29.930 --> 00:20:32.050 align:start position:0% one and I think it really has been great is<00:20:30.380> that<00:20:30.650> we<00:20:30.770> have<00:20:30.950> there's<00:20:31.160> no<00:20:31.190> multi<00:20:31.730> culture 00:20:32.050 --> 00:20:32.060 align:start position:0% is that we have there's no multi culture 00:20:32.060 --> 00:20:33.640 align:start position:0% is that we have there's no multi culture where<00:20:32.240> we<00:20:32.330> don't<00:20:32.540> just<00:20:32.780> rely<00:20:33.050> on<00:20:33.080> one<00:20:33.380> client 00:20:33.640 --> 00:20:33.650 align:start position:0% where we don't just rely on one client 00:20:33.650 --> 00:20:36.400 align:start position:0% where we don't just rely on one client to<00:20:34.010> hold<00:20:34.310> up<00:20:34.430> the<00:20:34.550> network<00:20:34.730> and<00:20:35.090> I<00:20:35.150> hope<00:20:35.870> that 00:20:36.400 --> 00:20:36.410 align:start position:0% to hold up the network and I hope that 00:20:36.410 --> 00:20:39.910 align:start position:0% to hold up the network and I hope that that<00:20:36.770> node<00:20:37.490> multi-culture<00:20:38.330> will<00:20:38.630> continue<00:20:39.320> on 00:20:39.910 --> 00:20:39.920 align:start position:0% that node multi-culture will continue on 00:20:39.920 --> 00:20:44.310 align:start position:0% that node multi-culture will continue on ethereum<00:20:40.570> that's<00:20:41.570> it<00:20:41.750> for<00:20:41.930> me<00:20:42.020> thank<00:20:42.200> you 00:20:44.310 --> 00:20:44.320 align:start position:0% ethereum that's it for me thank you 00:20:44.320 --> 00:20:45.580 align:start position:0% ethereum that's it for me thank you [Applause] 00:20:45.580 --> 00:20:45.590 align:start position:0% [Applause] 00:20:45.590 --> 00:20:54.159 align:start position:0% [Applause] [Music] 00:20:54.159 --> 00:20:54.169 align:start position:0% 00:20:54.169 --> 00:20:56.230 align:start position:0% you