text
stringlengths
0
2.35k
[2080.76 --> 2082.46] Speaking of loops, I have a scary story.
[2082.46 --> 2097.22] But if that's really a story that I think helped me gain a new appreciation for sort of how to integrate systems, how distributed systems have sort of pitfalls.
[2097.22 --> 2104.84] It's a tradeoff for everything and all the things that we value, right, as best practices for dealing with, you know, integrating systems.
[2104.84 --> 2117.28] So I was working for an organization, that very awesome organization, a nonprofit, which basically helps students, right, especially in underserved communities, sort of prepare to take sort of standardized testing and that kind of thing.
[2117.28 --> 2126.46] So for months leading up to a major sort of testing day, right, students are going to come, sit down into their classrooms.
[2126.78 --> 2132.16] They're going to be logging in and taking an assessment, right, to help them, right, with the real things.
[2132.40 --> 2136.94] And this is like a coordination across multiple schools and everything.
[2137.02 --> 2138.84] So the whole county, right, is doing this thing.
[2138.88 --> 2144.62] So we're talking like, you know, maybe like 3,000 to 4,000 students, right, that are going to all sit down and do this thing.
[2144.92 --> 2145.02] Wow.
[2145.02 --> 2152.06] And basically I'm part of the team that's basically has been working on this sort of integration, right, for months now, right?
[2152.12 --> 2154.24] We have different systems talking to each other and everything else.
[2154.82 --> 2157.98] In development and even in staging, everything works perfectly.
[2159.32 --> 2163.04] Systems can talk to each other, you know, like we're sending, you know, lots of traffic.
[2163.18 --> 2167.90] We're keeping an eye on things and we're observing to the best availability, you know, with the tools that we have.
[2168.50 --> 2174.00] And on production day, basically, which is when students actually sit down to do the thing,
[2174.00 --> 2183.12] what we didn't test against is basically having roughly 4,000 students trying to log in to the system at the same time.
[2185.66 --> 2190.86] You know, and because you have these systems that are talking to each other for authentication and pooling things, whatever it is,
[2190.94 --> 2194.22] basically we just had a thundering herd kind of situation happening.
[2194.22 --> 2201.04] And we didn't account for that because all of our tests, even our integration tests and everything else, they didn't factor in that kind of scale.
[2201.38 --> 2204.74] And I take responsibility for that because I was one of the team leads.
[2204.88 --> 2212.20] And basically one of my questions was supposed to be, what is the expected number of users and clients, right, on the system?
[2212.20 --> 2219.30] And basically we touched on these things, but there were bottlenecks in the system, right, that we should have better accounted for.
[2219.74 --> 2226.68] And thankfully we had enough of an understanding of what was happening with enough observability to be like, oh, crap, we know where the bottleneck is.
[2226.72 --> 2227.64] We need to go do this, whatever.
[2227.82 --> 2233.14] So within a matter of, you know, about an hour and a half or so, while students are waiting there,
[2233.14 --> 2236.32] because they can't really dismiss everybody and send everybody home, right?
[2236.36 --> 2242.34] We're talking like countywide, 4,000 plus students, all this coordination across months, you know.
[2242.64 --> 2248.18] To me, this remains the best and worst moment of my career because I'm like, here I am.
[2248.30 --> 2249.96] I'm supposed to be serving these kids.
[2249.96 --> 2255.60] Like as an, often we are so far removed from the consequences of our code, right, good or bad, right?
[2255.62 --> 2256.68] We're so far removed from it.
[2256.76 --> 2263.00] But here I am, I knew exactly, right, what the impact that my mistake was having on these kids.
[2263.14 --> 2263.34] Right.
[2263.36 --> 2266.56] And they already, right, have been given a short straw in life.
[2266.60 --> 2268.58] And here I am just making that worse, right?
[2268.94 --> 2271.68] So after that incident, I was like, never again.
[2271.72 --> 2272.86] Like, what do I need to do?
[2272.96 --> 2273.98] What do I need to learn?
[2274.16 --> 2275.30] Who do I need to talk to?
[2275.70 --> 2277.76] Like it was, I had to level up.
[2277.76 --> 2283.64] At any point in my career, I can't remember a single incident that has driven me to level up as much as this one.
[2283.74 --> 2286.00] Because the impact was so real.
[2286.18 --> 2287.84] It was so in my face.
[2287.96 --> 2288.88] It's just undeniable.
[2290.18 --> 2290.72] That was scary.
[2290.72 --> 2294.68] That is the $1,001 bill.
[2296.00 --> 2296.78] Yeah, that's a face.
[2296.98 --> 2299.60] Listen, I would have gladly like handed over $1,000.
[2300.10 --> 2301.26] Like out of my pocket.
[2301.50 --> 2303.30] Like to be like, look, whatever this is.
[2303.34 --> 2303.88] To the kids.
[2304.02 --> 2305.26] Make it go away right now.
[2305.28 --> 2306.02] Just give it to the kids.
[2306.12 --> 2306.44] To the kids.
[2307.64 --> 2308.00] Right.
[2308.56 --> 2310.40] Uncle Johnny's messed up again.
[2310.54 --> 2312.94] Come and collect your $20 bills, everyone.
[2313.54 --> 2316.68] No college for you, but here's some, I don't know.
[2316.78 --> 2317.60] Yeah, that's horrific.
[2317.60 --> 2324.04] Well, yeah, but when the stakes are that high, Johnny, like that is like, ooh, that is scary.
[2324.44 --> 2324.62] Yeah.
[2324.86 --> 2325.08] Yeah.
[2325.18 --> 2326.50] And you feel awful.
[2326.74 --> 2331.04] Like awful, awful, awful for being responsible for that.
[2331.26 --> 2331.48] Oof.
[2331.48 --> 2332.02] Mm-hmm.
[2332.90 --> 2337.12] What do you think about sort of engineers sort of being, having that sort of sense of consequence?
[2337.42 --> 2338.88] Because it comes in multiple directions.
[2339.04 --> 2343.34] You get salespeople going, if you don't do this, we're going to lose a million dollar deal.
[2343.52 --> 2345.62] It's not like the engineer is going to receive a million dollars, right?
[2345.64 --> 2346.96] They're just going to get their normal salary.
[2347.12 --> 2348.18] They're getting the normal pay.
[2348.72 --> 2350.12] That's a little bit abstract.
[2350.28 --> 2351.04] It's a ton of stress.
[2351.04 --> 2351.60] Mm-hmm.
[2351.60 --> 2355.60] And likewise, when you're working incidents, you know, someone will turn and go, it's affected
[2355.60 --> 2360.74] this air traffic control signal, or it's affected this sort of kids or hospital, or traffic
[2360.74 --> 2361.92] lights are down, or whatever.
[2362.66 --> 2364.04] Kind of unhealthy, isn't it?
[2364.52 --> 2370.02] It's like, we've got to keep it abstract enough that, because I get that it changes us.
[2370.38 --> 2372.32] All of those incidents changed all of us.
[2372.64 --> 2373.82] They're all horrifying moments.
[2373.82 --> 2378.52] But I'm also just like, that's the path to burnout, to sort of accepting the consequence
[2378.52 --> 2379.66] for all of these things.
[2379.92 --> 2380.02] Yeah.
[2380.16 --> 2384.90] So some value in being somewhat abstracted from the consequences.
[2385.18 --> 2385.58] Is that what you mean?
[2385.88 --> 2386.20] Yeah.
[2386.48 --> 2389.66] If you consider it too much, it just weighs so heavily.
[2389.88 --> 2390.80] It's too serious.
[2390.96 --> 2392.04] It's too much of a...
[2392.04 --> 2395.26] And the things you need to do to actually get out of those situations, they become even
[2395.26 --> 2396.46] more horrifying and scary.
[2396.70 --> 2398.02] What if I prolong this?
[2398.10 --> 2399.28] What if I make it worse?
[2399.28 --> 2402.08] And sometimes you've just got to be a bit fearless.
[2402.08 --> 2406.48] And you can't if you've got that sort of burden on you that we put on ourselves.
[2406.98 --> 2409.64] I feel like this is where systems can be helpful, though.
[2409.72 --> 2414.94] Because I think we, as an industry, are pretty bad at understanding that there can be bad