1 00:00:00,00 --> 00:00:03,080 After sustential research, the detection team is ready to advance. 2 00:00:03,090 --> 00:00:07,420 All we need to strengthen our research capabilities 3 00:00:07,430 --> 00:00:11,700 is just a frozen index so we can conduct retrospective research. 4 00:00:11,710 --> 00:00:15,910 It will greatly speed the cadance™ and improve the precision. 5 00:00:15,920 --> 00:00:20,980 Alright, so just spin up a new ElasticSearch cluster. 6 00:00:20,990 --> 00:00:24,430 7 00:00:24,440 --> 00:00:25,330 Look.. 8 00:00:25,340 --> 00:00:30,040 Raz...? 9 00:00:30,050 --> 00:00:33,780 Spinning up a new cluster is not that simple. 10 00:00:33,790 --> 00:00:36,120 It's a 3 month project. 11 00:00:36,130 --> 00:00:53,070 12 00:00:53,080 --> 00:00:59,020 Everyone who's been in the company for more than two years, stay. Everybody else, leave now. 13 00:00:59,030 --> 00:01:13,150 14 00:01:13,160 --> 00:01:14,580 What is this!? 15 00:01:14,590 --> 00:01:17,490 I can't comprehend what you're saying. 16 00:01:17,500 --> 00:01:23,680 We are an org of 100 engineers, we have been using ElasticSearch in production for years! 17 00:01:23,690 --> 00:01:27,620 Almost every team has a dedicated cluster 18 00:01:27,630 --> 00:01:31,450 which has a full freaking copy of the entire dataset of the company! 19 00:01:31,460 --> 00:01:33,930 We have split the clusters so issues won't impact cross-team workloads 20 00:01:33,940 --> 00:01:38,400 But still dev-velocity is almost unusable 21 00:01:38,410 --> 00:01:41,020 It's red if we are not super duper careful. 22 00:01:41,030 --> 00:01:41,280 Look, 23 00:01:41,290 --> 00:01:43,050 we're doing our best to monitor and -- 24 00:01:43,060 --> 00:01:46,110 Monitoring is not enough! The clusters are just too loaded 25 00:01:46,120 --> 00:01:47,060 There's a new 26 00:01:47,070 --> 00:01:48,540 cluster, ES-Analytics 27 00:01:48,550 --> 00:01:51,960 It's only used for Sigmund and it crashes every other day. 170 node cluster! 28 00:01:51,970 --> 00:01:54,200 It's too much! 29 00:01:54,210 --> 00:02:01,600 We need to be able to spin up clusters more easily! Obviously they are too loaded. Everyone needs a cluster 30 00:02:01,610 --> 00:02:03,730 So just have some decent infrastructure 31 00:02:03,740 --> 00:02:08,480 so we could click to deploy a cluster for anyone who requires it. 32 00:02:08,490 --> 00:02:14,940 Without it we would never be able to proceed with our deeper research needs, 33 00:02:14,950 --> 00:02:16,700 our competitors will outsmart us. 34 00:02:16,710 --> 00:02:21,830 Velocity is the key selling point of this company 35 00:02:21,840 --> 00:02:27,700 36 00:02:27,710 --> 00:02:29,580 We have outstanding researchers here 37 00:02:29,590 --> 00:02:31,880 the engineers are fine also. 38 00:02:31,890 --> 00:02:34,390 But first we are human beings. 39 00:02:34,400 --> 00:02:36,670 We need to have quality tools to be able to work. 40 00:02:36,680 --> 00:02:41,440 - 41 00:02:41,450 --> 00:02:43,010 The alternative of just creating a new cluster 42 00:02:43,020 --> 00:02:47,980 is dead on arrival. Precision Eng don't have the capacity to handle it 43 00:02:47,990 --> 00:02:53,060 and other teams don't have the knowhow. 44 00:02:53,070 --> 00:02:56,090 Making RE do it without proper infrastracture is a suicide mission 45 00:02:56,100 --> 00:02:59,450 There's no chance they'll be able to handle it. It's a cluster 5x big compared to ES-Analytics 46 00:02:59,460 --> 00:03:02,890 It'll comsume all of their time. 47 00:03:02,970 --> 00:03:05,190 48 00:03:05,200 --> 00:03:05,580 We will be fine. 49 00:03:05,590 --> 00:03:07,970 We're hiring people to help... we just can't find them 50 00:03:07,980 --> 00:03:14,830 51 00:03:14,840 --> 00:03:16,760 There's no choice. 52 00:03:16,770 --> 00:03:23,740 We will just have to invest the time. 53 00:03:23,750 --> 00:03:26,900 FPGAs are a long, long way ahead. 54 00:03:26,910 --> 00:03:32,060 - 55 00:03:32,070 --> 00:03:35,690 We'll have to double down on Elastic. 56 00:03:35,700 --> 00:03:41,000 - 57 00:03:41,010 --> 00:03:47,360 We'll just have to make sure this infra project is better managed. Agile. 58 00:03:47,370 --> 00:03:48,980 To the point. 59 00:03:48,990 --> 00:03:54,840 60 00:03:54,840 --> 00:03:55,630 So it wouldn't end up like past infra efforts.