The anatomy of a 2AM mental breakdown

https://zarar.dev/anatomy-of-a-mental-breakdown/

by recroad • 2 years ago

435 261 2 years ago

riiii2• 2 years ago

I heard a sleep expert say that during the night your logic and reasoning abilities are greatly reduced. I think it was in relation to dreaming, you don't want to apply much logic to that stuff.

That's why trying to solve problems in the middle of the night just ends up in stress.

SketchySeaBeast• 2 years ago

It also makes sense why I'll wake up in the middle of the night terrified of things that don't bother me during the day. My fears all are much more immediate in the witching hour, and I can't talk them away.

riiii2• 2 years ago

I'm the same. The only thing that has a chance of working is to remind myself that it's the middle of the night and however terrifying this is I'll deal with it in the morning.

Waterluvian• 2 years ago

Funny enough, I almost exclusively do my best problem solving at night. I'll wake up with solutions, or I'll stay awake and get things done. I'd say 90% of my Master's work was done between midnight and 3am.

But indeed, regardless of time of day, if I just wake up, or am woken up, I'm basically a big, dangerous toddler when it comes to problem solving.

Context and nuance is important, of course. We're all so different.

jrgoff• 2 years ago

That often happened for me in grad school as well. Generally the questions I had trouble with on a take home exam would yield to late night inspiration. And if they didn't yield by a semi-reasonable time, I would go to bed and many times, as I was drifting off, an insight would come to me. One memorable time though, that didn't happen and I woke up several times in the middle of the night from stress dreams where I was trying to solve the problem. And when I thought about the dream, nothing I had been doing in it made any logical sense to actually help me with a solution. Fortunately I woke up early and was able to figure it out in the morning. It was not a very restful night of sleep though.

jimmaswell• 2 years ago

On the other hand, sometimes I find the time right after waking up the clearest, having an empty mental stack. Like having a clear cache.

sergiotapia• 2 years ago

correct, once i notice myself making stupid mistakes i verbally confirm it with myself and say whelp that's the night. superpower. too many times i've woken up and said what the fuck was i thinking?

bityard2• 2 years ago

Pilots call the wee hours of the morning the "window of circadian low." Unless you routinely sleep during normal daytime hours (for a few weeks at least), the circadian low still nudges your brain and body metabolism toward a lower-energy state EVEN if you are fully rested up until that point.

https://www.faa.gov/documentLibrary/media/Advisory_Circular/...

rockyj• 2 years ago

All monitoring comes at a cost and adds complexity. I wish people realized that, I struggle with this in my own team, we keep adding layers upon layer of monitoring, metrics etc.

htrp• 2 years ago

You need to insulate your metrics/monitoring from the critical path so that failures in these providers don't take out your app.

macNchz• 2 years ago

I’ve definitely come across this genre of issue on many sites in the past when checking out the console when a site is broken. Page is just a blank white screen? Oh, looks like the render function was placed after the init for some 3rd party user monitoring, which crashed because the script didn’t load properly. “Complete Checkout” button just does nothing at all? Oh, looks like the code to take my money runs in a callback to some analytics script that my ad blocker blocked. Oops.

ejs• 2 years ago

I had these issues before for plenty of things, it just hurts the most when it's something non-essential. I've had outtages because silly system updates to slack broke and took things down. I run metrics and such out through logs these days because UDP don't care.

ssiddharth• 2 years ago

Ha, that was a stressful yet funny read. The self flagellation bit hits too close to home though. I run a somewhat successful iOS/MacOS app and pushed a release that completely broke about 350k+ installations. Not entirely my fault but doesn't matter as it's my product.

The cold sweats and shame I felt, man... Plus it's on the App Store so there's the review process to deal with which extends the timeline for a fix. Thankfully, they picked it up for review 30 minutes after submission and approved it in a few minutes.

cqqxo4zV46cp• 2 years ago

My first employer as a developer, due to their incompetence, not intelligence, ‘let’ me break our customers’ shit from a young age. As I’ve progressed through my career, and through my transition to leadership, I’ve realised that it was a very valuable experience. I may have stressed about it in the past, but those memories are too distant for me to even reach now. I certainly don’t stress about it now.

I’ll, maybe controversially, sometimes allow my (early-career) team members to break prod, if I can see it happening ahead of time, but am confident that we’ll be able to recover quickly. It’s common knowledge that being given room to fail is important. But many leaders draw the line at failures that actually hit customers. If one finds oneself in the very common and very fortunate position to be building software that isn’t in charge of landing planes or anything similarly important, they should definitely let their team experience prod failures, even if it’s at the expense of so-and-so from Spokane Washington not being able to use their product for a few minutes.

mynameisvlad4• 2 years ago

Within my first 6 months at a FAANG, I accidentally took down our service for half the world for 30-60 minutes.

That was honestly one of the best things that could’ve happened to me and I still use the story for new hires to this day. It’s a very humbling reminder that nobody is perfect and we all make mistakes. And I’m still here, so it’s never the end of the world.

sandspar• 2 years ago

Nice writing. Art can be useful for helping us cope.

curiousllama• 2 years ago

Love the detailed emotional reaction to scrambling to fix an outage. Nothing quite like attempting calm, dispassionate debugging while actively wrestling your own brain

tempfile• 2 years ago

Every external service you integrate is adding a small, non-zero, compounding probability of finding yourself in exactly this situation.

datavirtue• 2 years ago

Not to mention the performance hit to your actual customers when it all "works."

shadowgovt• 2 years ago

This also serves as a cautionary tale to small-business web people.

You can start a web service business solo (or with a small handful of folks). But the web doesn't shut down overnight, so either have a plan to get 24-hour support onboarded early or accept that you're going to lose a lot of sleep.

(And if you think that's fun, wait until you trip over a regulatory hurdle and you get to come out of that 2AM code-bash to a meeting with some federal or state agent at 9AM...)

JohnMakin4• 2 years ago

Working as a SRE for a year in a large global company broke me out of this "panic" mode described in this post. To a business, every problem seems like a world-ending event. It's very easy to give in to panic in those situations. However, in reality, it's rarely that bad, and even if it is, you'll probably survive without harm. The key in these situations, and what I try to do (totally relate to breaking out in a sweat, that still happens to me, just happened yesterday) is to take 5-10 minutes before doing any action to try to fix it and sketch it out, think about it as clearly as you can. Fear interferes with your ability to reason rationally. Mashing buttons in a panic can make your problems spiral even worse (seen that happen). Disrupting that fear circuit any way you can is important. Splashing my face and hands with extremely cold water is my trick.

Then after you go through a few of these, you'll realize it really isn't too bad and you've dealt with bad situations before and you'll gain the confidence to know you can deal with it, even when there's no one you can reach out to for help.

H8crilA• 2 years ago

And ultimately you carry none of the risk. It's not your company, and the company can (and will) cut you off at a random time. Unless it is your company :)

minkles2• 2 years ago

This is the entire risk: being fired and financially in trouble. I spent 5 years eliminating that risk. Now I don't give a single fuck. There is no fear. They get better, more rational work and I have security.

sdenton4• 2 years ago

OTOH, a lot of people who really needed to be able to change their event pricing at 2am (+/- two timezones) on jumpcomedy.com were really let down. Probably some of them died.

Imagine how much more the damage would have been if someone was telling this solo developer to stop trying to make 'fetch' happen.

agumonkey2• 2 years ago

After a few times I realized that everytime I step into blind-ish panic over a mistake, I should get as far as possible from the machines. Otherwise I try one bad idea after the other and create an actual catastrophic failure.

caseyy6• 2 years ago

What is this response called? I see it in many people. It’s like trying to catch a boiling mug of coffee when it falls — panic and not thinking leads to severe consequences; inaction would be better.

It seems to relate to fight or flight, as opposed to freezing. All three are not great and can significantly escalate a bad situation.

nine_zeros2• 2 years ago

The easiest way to not panic is to always remind yourself, "You don't need to be a hero. It is not your company".

If the company wants to derisk their business, they can hire more people and create systematic safeguards in their services over a long period of time. The return on this investment is theirs. If they don't make this investment, they don't get the returns - simple as that.

If you are afraid of "losing your job" - remember that you might lose your job even if you were the hero because of reasons not really in your control.

The only time to really panic is when your family is under crisis. To prevent your family from going in crisis, you should build safeguards in terms of health, income etc.

Business owners can do the same if they want safeguards. If not, fine, things will just break. And it is NOT an individual engineers problem to fix all of their poor decisions.

bityard2• 2 years ago

For me, in situations like this, the fear and anxiety comes from the part of my brain screaming, "but what if I can NEVER fix this?" Which is of course completely nonsense. There is ALWAYS a solution somewhere, even if it's VERY hard to find, or even if it's not the one you want.

My first serious job as an adult was avionics repair. I knew how to do SOME component-level troubleshooting but I wasn't very good at it and always got lost in schematics pretty quickly. When I ran across a broken device that had a problem I couldn't solve within 30 minutes of investigation, my first instinct was to say, "welp, this is beyond my abilities" and ship the card or box off as permanently broken, or try to pawn it off to a more senior tech in the shop.

Fortunately, I had a mentor who disabused me of that tendency. He taught me that EVERY fault has a cause. Every time you think you've exhausted all options, what it really means is that you've only exhausted all CONVENIENT options and you just have to suck it up and get cracking on the harder paths.

everdrive2• 2 years ago

Bear in mind that when something "breaks" a business can freak out, but they're not freaking out about lots of other arguably more important problems:

- The software they purchased doesn't do anything because it was never staffed or configured correctly.

- Employees as a whole lose thousands of hours a year to terrible UX or meaningless requirements.

- Some capability doesn't actually do anything, but no one cares enough to notice.

- Useless meetings which waste collective employee hours daily.

- Capability serves no purpose, but just exists to meet audit requirements.

- Executives enrich themselves and waste company money constantly.

- Etc.

Downtime doesn't seem any worse than these problems, but it gets far more attention and panic. It feels a bit like terrorism vs. heart disease. It's important not to sweat it, because a company does not care about your sleep or mental health. It will push you as far as it can. I'm not calling companies malevolent, but they're a bit like a bully in that regard: they push you as much as you yield.

ekanes• 2 years ago

It feels a bit like terrorism vs. heart disease.

Fantastic analogy! One is scary but practically non-existent, and the other will bring early death to many people you know.

JohnMakin4• 2 years ago

One of the funnier situations (funny now, wasn't so much the first time I saw this) that I run into at new gigs or contracts is when a business has absolutely zero monitoring or alerting. Go into their backend, it's predictably a dumpster fire. Start plugging in monitoring and the business realizes everything is on fire and PANICS. It's very difficult to explain to someone who definitely doesn't want to hear that it's actually been broken for a long time, they just didn't care enough to notice or to invest in doing it right the first time.

lifeisstillgood2• 2 years ago

That is a great perspective - thank you

caseyy6• 2 years ago

It feels a bit like terrorism [..] It’s important not to sweat it

Hmm… interesting thought. It’s interesting how labeling something a terror attack makes the public lose critical thinking and act incoherently for a bit. It’s like a stunning shock response. The very same this author describes.

Do you think some managers intentionally play up the severity of problems to push their teams into this panic mode?

People in customer service know clients definitely do.

It is interesting how these concepts connect.

vlunkr3• 2 years ago

The zero-downtime culture is pretty out of control. Your flight can be delayed for hours, your roads can be closed for weeks, your internet can be out for hours because someone cut a line, that's all ok, but someone can't get into your website for 3 minutes to check the status of their order and we all must lose our minds and publicly apologize and explain in detail what happened and why it will never happen again.

fragmede6• 2 years ago

There are even times to prefer to site being down to being up. If you've been hacked and the site is serving CSAM/malware/leaking PII maliciously, as the SRE on deck, your job is to keep it down!

Keyframe• 2 years ago

not saying you're wrong, because you absolutely aren't, but there's one key difference - usually freakouts are when it's client-facing related in combination with no control and/or information on what's going on and 'they' have to answer customer calls. That's when panic mode is activated.

cgh• 2 years ago

It's important not to sweat it, because a company does not care about your sleep or mental health. It will push you as far as it can.

One time a few years ago, a coworker called me to tell me the stress we were under was causing her hair to fall out. I mentioned I'd lost significant weight and wasn't sleeping well, especially since another coworker had quit and now I was shouldering his responsibilities (programming and managing a junior). She said she was going to quit. I wished her the best. Around six months later, I had to quit too. It took around a year for me to get back to normal sleep patterns and for my heart to stop racing in the night.

Oh yeah, another guy from that team had a "cardiac event" due to stress and had to take a month off.

burningChrome• 2 years ago

> Bear in mind that when something "breaks" a business can freak out, but they're not freaking out about lots of other arguably more important problems

THIS.

Worked for a huge corporation when I first started contracting as a developer. This was a daily occurrence. Someone would push some stuff to prod and it would break their very large e-commerce site. Sometimes it was minor stuff, other times it would cripple the entire site. Most of the time, nobody GAF at all.

I remember sitting at my desk one day and my senior dev walking by my desk like Lumbergh in Office Space. Coffee in hand and just casually said to me, "Looks like someone broke the build, most of the video game pages are throwing 404's right now, going to lunch, see you soon." and then just walked away.

I also found out one day one of the CDN folks that was running the Akamai account quit and then a month later, we found out one of our secondary content servers had been down since the day he left. When dude left, he never transferred any of his knowledge to the other team members so once they found out the server was down, it took another two days to get in contact with someone at Akamai who was handling our account.

At my previous job if something went down on our site, it was a four alarm fire, war room, and all hands on deck to get it resolved or heads would roll. It was so dysfunctional to work somewhere when something broke or stopped working, nobody was in any hurry to fix it. Several times I just thought, "Is this what they mean when they say "the inmates are running the asylum""?

lukan• 2 years ago

"Splashing my face and hands with extremely cold water is my trick."

Yes, literally cooling down. Combined with conscious breathing. And then go on doing, what you can do.

cma• 2 years ago

It's the mammalian diving reflex: https://en.wikipedia.org/wiki/Diving_reflex

sameoldtune• 2 years ago

I was once the primary party responsible for an accounting bug that would have cost Amazon around 100 million dollars in unselable inventory. I stayed up for two nights writing the most heinous regexes to scrape logs(1) and fix the months-long discrepancies. Ended up fixing the problem by writing a script that re-submitted an event for every inbounded shipment to every Amazon warehouse for the last few months. I took the next week off to recover physically and mentally.

Afterward I made a promise to myself to never get into such a panic again. But funny enough, when I feel like panic is immanent I just remind myself that whatever problem I have is absolute peanuts compared to that event. So for me at least having one truly bad event helps me put things into context.

(1) the logs had JSON-escaped JSON and I am slightly proud of those regexes because it is the only time I hopefully ever have to use 8 backslashes in a row https://xkcd.com/1638/

sb8244• 2 years ago

Some of the worst mistakes that I saw were from over-reaction in an active incident.

One of my programming mantras is "no black magic." If I don't understand why something works, then it's not done.

I take this same approach to an incident. If someone can't coherently identify why their suggestion will have an impact, I don't think they should do it. Now there may come a time that you need to just pull the trigger on something, but as I think back I'm not sure that was ever the case in the end.

It was wild to see the top brass—normally very cool and composed—start suggesting arbitrary potential fixes during an incident.

Vegenoid2• 2 years ago

It's a good reminder that if things get bad, people will just start burning things to try and appease the gods.

JohnMakin4• 2 years ago

I have a similar mantra - "if you don't know why a fix worked, you may not have fixed it."

I'm willing to throw shit at the wall early in the triaging process, but only when they are low-impact and "simple" things. stuff like -

have we tried clearing cache?

have we checked DNS resolver for errors?

have we restarted the server?

etc. I try to find the "dumb" problems before jumping to some wild fix. In one of the worst outages of my career, a team I was working for tried to do a full database restore, which had never been done in production, based on a guess. At 3am on a saturday. I push back really hard at stuff like that.

DylanSp• 2 years ago

To quote Gene Kranz during the start of problems on Apollo 13: "Let's not make things worse by guessing."

aPoCoMiLogin• 2 years ago

One of my programming mantras is "no black magic." If I don't understand why something works, then it's not done.

so many ppl nowadays don't care about knowing how thing works, they just push shit and cross the fingers.

ChrisMarshallNY2• 2 years ago

One of my programming mantras is "no black magic."

This. * 1000

I am very grateful that my earliest training was as an RF bench technician. It taught me how to find and fix really weird problems, without freaking out.

However, it does appear that this particular event may have been caused by reliance on a dependency that may not have merited that reliance.

I'm really, really careful about dependencies. That's an attitude that wins me few friends, in this crowd, but it's been a long time, since I've had 2AM freakouts.

hibikir• 2 years ago

My proudest SRE moment was when, in a team that dealt with a key system where seconds meant millions, I was able to find a precursor to the problem, so instead of having to intervene on a page in 3 minutes before disaster, the alarm now had 20 minutes of headroom.

We all slept so much better after that

pknomad2• 2 years ago

One of my VP's (one of the coolest dude I know) would often say "slow is smooth and smooth is fast".

You're absolutely right that fear interferes with one's ability to reason rationally but I'd also add that fear is highly infectious. If ICs sees their leaders/managers/colleagues panicking, they will often panic too. Thankfully my VP was always level-headed and prioritized clarity over action.

DowagerDave• 2 years ago

> One of my VP's (one of the coolest dude I know) would often say "slow is smooth and smooth is fast".

Apparently you worked for a Navy Seal, so yeah, likely a very cool dude :)

dmvdoug• 2 years ago

I must not fear. Fear is the mind-killer. Fear is the little-death that brings total obliteration. I will face my fear. I will permit it to pass over me and through me. And when it has gone past I will turn the inner eye to see its path. Where the fear has gone there will be nothing. Only I will remain.

iwontberude• 2 years ago

Same experience working for a major streaming provider as SRE. Even when it's a million dollars per minute, the stress never will help you to get the bleeding to stop. It's very helpful everywhere else in life. I can handle social, physical, and other stressors better after this exposure. Nothing in business is as bad as your biology can fool you to believe.

ldayley• 2 years ago

Author: Thank you for writing this! I love reading about how people overcome challenges like this, especially under pressure (and usually overnight!). I am better for hearing not just the technical post mortem but also the human perspective that is usually sanitized from stories like this. This is the kind of technical narrative only a small/solo dev or entrepreneur can share freely.

adamc• 2 years ago

Great reminder of the people behind services, as well as a nice account of the debugging process. The reality is that pressure doesn't make you debug problems any faster... usually, it interferes with your thinking. You have to try to ignore the consequences and stay as calm as possible.

But most of us have been in some situation similar, if not quite as bad. (Running your own company is going to be uniquely stressful.)

uaas2• 2 years ago

Having a(n accurate) service graph of all your (internal and external) dependencies is a game changer in troubleshooting issues like this.

thruway516• 2 years ago

Do you need another dependency for that? A dependency to manage dependency hell.

uaas2• 2 years ago

Optimally, that’s already part of your observability stack, and might well be a built-in feature. In a certain scale/landscape it easily pays off.

jonnycat• 2 years ago

Great post, but kind of buries the lede: PostHog is having a CrowdStrike moment.

timgl2• 2 years ago

PostHog cofounder here. This affected users that did not have a specific version of the JS library pinned and deployed a new version, or were using the snippet, and had network capture enabled, (a feature we introduced very recently and is only enabled on 3% of projects), and had recordings enabled on that particular session (for most customers, only a small percentage of sessions are recorded due to sampling or billing limits)

This outage was definitely disruptive and we shouldn't have let this happen. We will be doing a full post mortem write up, but this affected a small percentage of our users, so the comparison with Crowdstrike isn't fair.

Etheryte• 2 years ago

You're trying to phrase this as if those conditions make it any less bad, but they don't. This affected users that were using the latest version and used... features? Give me a break. Every product has bugs, but trying to downplay the issue after you've just read a distressed user of yours struggle with it is definitely not what you should be doing.

VBprogrammer• 2 years ago

Not your customer, just a random person on the Internet, but I hope you can see that a lot of that is through luck more than judgement.

I personally would have like to see a bit more contrition rather than trying to minimise the issue.

ziddoap2• 2 years ago

Random guy here. This affected users that

- Used your recommended way of implementing PostHog [1]

- Used a feature of the product

The comparison to CrowdStrike is not fair, you're right. But this attempt to shed responsibility still leaves a sour taste.

[1] See "This is the simplest way to get PostHog up and running. It only takes a few minutes." from your website, which is the first method suggested when clicking the "installation" tab

anonfordays2• 2 years ago

This affected users that did not have a specific version of the JS library pinned and deployed a new version

Par for the course honestly. The amount of garbage that gets called "production" these days is mindboggling. No blue/green or canary deployments, shipping code that has nothing pinned, no clear rollback, etc. This is what happens when anyone can become an EngineerTM after a two week JavaScript boot camp.

misja111• 2 years ago

"Maybe PostHog, I have the api_key blanked out locally to reduce costs"

Come on, if POST requests work locally and not on PROD, isn't this an obvious place to start?

mrsilencedogood• 2 years ago

The author mentioned significantly more notable prod/dev differences than the posthog API key, which I suspect is where they looked first and second. So no, not an obvious place to start.

charles_f• 2 years ago

Such an entertaining read, conveyed very well the sense of stress and abandonment felt by the author. It adds to it that this was written fresh and right in the moment, and feels as an expiation.

I'm struggling to find the lesson to take out of that. Limit your dependencies? Have a safe mode that deactivates everything optional?

foodevl• 2 years ago

This is no good. Let me just try reverting to a version from a month ago. Nothing. Three months ago? Nothing. Still failing. A year ago? Zilch.

Reverting your own code, but still using a broken PostHog update from that same day? For me, the lesson is to make sure that I can revert everything, including dependencies.

roywiggins2• 2 years ago

It seems that PostHog just always loads the latest version of this piece of itself:

https://github.com/PostHog/posthog/issues/24471#issuecomment...

Though you can opt to bundle it yourself:

https://github.com/PostHog/posthog/issues/24471#issuecomment...

slashdave• 2 years ago

I definitely want to figure out in detail what happened here so I can add a test to prevent a similar change in future!

Whoa! Good idea!

Could have been worse. At least the change didn't expose a hidden exploit.

ricardobeat2• 2 years ago

Ouch. That just adds insult to injury.

phkahler• 2 years ago

> It seems that PostHog just always loads the latest version of this piece of itself:

Now there's a supply chain attack vector...

philsnow2• 2 years ago

Years ago, IT at the company I was working at force-pushed a browser extension that did this same trick, but the extension vendor in question didn't even bother loading over https.

Edit: the extension's manifest gave it nearly every permission, on every web site, including internal ones

amsterdorn• 2 years ago

This is the key lesson. Your own deploys mean nothing if you link to another CDN for parts of your application.

You handled it well OP, the silver lining of incidents like this is the grab bag of valuable takeaways!

bluepnume• 2 years ago

Looks like the bug was in a monkey-patched `window.fetch`

https://github.com/PostHog/posthog-js/blob/759829c67fcb8720f...

The biggest lesson here is, if you're writing a popular library that monkey-patches global functions, it needs to be really well tested.

There's a difference between "I'll throw posthog calls in a try/catch just in case" and "With posthog I literally can't make fetch() calls with POST"

pbasista• 2 years ago

Thank you for pointing this out. I did not read the post in detail but I was wondering how could a monitoring library cause the entire application to go down. At worst, I thought, it should have failed to process the monitoring events, assuming that it was integrated in a reasonable way.

PostHog patching a very important global function is a feature that should be well-documented so that the people who are using it are aware of it and can be reasonably expected to have it in mind when debugging these seemingly unexplainable issues.

ataru• 2 years ago

I think it worked as defined, it hogged the POST requests?

ricardobeat2• 2 years ago

I was poking around to understand how this was not caught in a test - any ordinary fetch call could have triggered the error, and besides how poor coverage it has for all the ways `fetch` can be used, it seems excessive mocking may have played a part: https://github.com/PostHog/posthog-js/blob/main/src/__tests_...

The whole fetch and XHR functions are mocked and become no-ops, so obviously this won't catch any issues when interacting with the underlying (native or otherwise) libraries. They have Cypress set up so I don't see why you'd want to mock the browser APIs.

sethammons2• 2 years ago

I have seen so many mocked tests where you end up asserting the logic in the mock works; effectively testing 1=1.

The number of issues that can be prevented with an acceptance level test that has a user log in and do one simple interaction is amazing. Where I can convince the powers that be, PRs to main are gated by a build that runs, among others, that simple kind of AC test. If it was merged to main, you _know_ it will not totally break production.

We had regular outages with our internal emailing system at a small e-commerce shop. I stepped in and added one test that actually sent an email to a known sink that we could verify and had that test run pre-deploy. We went to zero email outages. Tests had the occasional flake that auto-retried. Also, if your acceptance tests are flaky, how do you know your software isn't? Bad excuse to avoid acceptance level testing

x0x02• 2 years ago

it's common with all these analytics toolkits. When you're monkeying with such a core api, I have no idea how you can actually test everything.

eg heap analytics still (as of this month) bad touches something inside hotwire and randomly entirely breaks hotwire causing every click to do a full page load. ime, it affects 30-60% of page loads. You can fix it (but thanks for 50+ hours of debugging) but making heap load after all hotwire js.

jjnoakes• 2 years ago

I have no idea how you can actually test everything.

While a worthy goal (if unattainable sometimes), in this case, folks would settle for "test anything", which would have caught the regression.

refulgentis• 2 years ago

The author needs to relax and Posthog needs more discipline. (and a rename)

f1shy• 2 years ago

I would love to read from the author what are the lessonS learned.

Use better tools? Know better your tools? Know better how to debug? Add yet another tool to detect the error?

In all big companies where I worked, at the end of such an event, it boiled down to answer the 3 questions: - what happened?

- why did it happen?

- what do we do so it does not ever happen again?

mobeigi• 2 years ago

These breakdowns happen to everyone and its really bad when its just you against the world. I've been lucky that my last few major outages have all been team efforts with anywhere from 2-10 people working on the issue. Albeit, this is a perk of working in a large enterprise.

With more than one person you can bounce ideas off each other and share the pain so to speak. It's highly desirable.

simpaticoder2• 2 years ago

This person's stress was caused by a single line of code in PostHog. This is the reversion: https://github.com/PostHog/posthog-js/pull/1371/commits/7598...

Highlights two lessons. 1. If you ship it, you own it. Therefore the less you ship, the better. Keep dependencies to a minimum. 2. Keep non-critical things out of the critical path. A failing AC compressor should not prevent your engine from running. Very difficult to achieve in the browser, but worth attempting.

M4v3R• 2 years ago

These are valuable lessons for sure, but then someone from the marketing comes and demands you add PostHog or any other tracking script to the site and won't take no for an answer.

simpaticoder2• 2 years ago

Then communicate clearly the trade-off the project leader is making.

threecheese• 2 years ago

Even worse, it appears that PostHog dynamically updates their part of their code at runtime - not bundling it at build time. Their docs note an Advanced Option where all dependencies are bundled in the build. I mean, I get why, and maybe I am misunderstanding but as a user I would expect lazy loading of executable code to be an optimization rather than the default. And used only if fully bundling was a serious delivery delay.

__MatrixMan__2• 2 years ago

From the PR:

fetch() broken on August 19: TypeError: ...

Not broken at this version, broken on August 19. This is why I'm terrified of putting anything on the web. It is a dark scary place where runtime dependency on servers that you don't control is considered normal.

One day I'll start my own p2p thing with just a bunch of TUI's and I'll only manage to convince six people to use it each for less than a month and then I'll have to go get a real job again but at least I won't have been at the mercy of PostHog.

scottlamb• 2 years ago

Not broken at this version, broken on August 19. This is why I'm terrified of putting anything on the web. It is a dark scary place where runtime dependency on servers that you don't control is considered normal.

Yeah, that is terrifying. In a nearby comment [1], a PostHog co-founder wrote this affected sites which "did not have a specific version of the JS library pinned and deployed a new version, or were using the snippet". I gather from that is it possible to pin the version, and this incident highlights the value of doing so.

[1] https://news.ycombinator.com/item?id=41301008

__MatrixMan__2• 2 years ago

I'd prefer a something where such references are resolved by cryptographic hash so that there's never any ambiguity re: what you're actually getting.

Unison does this I believe.

linuxrebe1• 2 years ago

Based on the way you were troubleshooting it. You can tell you're a programmer first. You went to your code, you went to your logs. Both reasonable, both potential causes of the problem. Both ignore the primary clue that you had. It worked on localhost.

As an SRE/devops/platform engineer or whatever the title of the day is people want to give. I would have zeroed in on the difference between the working system. And the non-working system. Either adding and then removing, or removing and then adding back the differences one at a time. Until something worked. What I see is two things. 1) you have an environment where it does work. 2) the failing environment was working, then started failing.

Is my method superior to yours, no. It just is being stated to highlight the difference in the way we look at a problem. Both of a zero in on what we know. I know systems, you know code.

9659• 2 years ago

many years ago, i was working as an electronic technician. we had a stack of processor boards (from a Perkin Elmer 7/32) that were removed from service. broken. many different revisions, and only schematics for one revision of each board. i thought it was hopeless. an older wiser tech taught me how.

plug a good board on an extender. run a diagnostic that fails in a loop. using a scope, look at every pin on the connector. write down what you see. replace with a bad board. repeat.

which signals are different? chase them back. if the schematic does not match, get out a voltmeter and your eyes and draw a schematic that reflects how the board is wired.

he called this "good card - bad card".

and it worked. not going to make any claims about cost effectiveness, but we fixed every board. and my troubleshooting skills in digital electronics were greatly improved.

this was a 'fireman' kind of job. waiting for the system to break, so it didn't matter if 2 techs put a week into 1 circuit board.

tlarkworthy• 2 years ago

... and thats why you pin dependancies

sillysaurusx3• 2 years ago

For what it's worth, I'm not sure this is a mental breakdown, and it might give the wrong impression to people who do legitimately suffer breakdowns related to stress about tech.

For me it's only happened once. It was an anxiety attack, and I'm very lucky my wife was there to talk me through it and help me understand what was happening. She's had them many times, but it was my first (and thankfully only).

It turns out that this sort of thing happens to people, and that there's nothing wrong with it. It doesn't mean you're defective or weak. That's a really important point to internalize.

Xanax is worth having on hand, since that was what finally ended it for me and I was able to drift off to sleep.

I guess my point is, there's a difference between having intrusive thoughts vs something that debilitates you and that you legitimately can't control, such as an anxiety attack or a panic attack. You won't be getting any work done if those happen, and that's ok.

hughes• 2 years ago

Yeah I was actually disappointed to find an article about a routine dependency debugging story. I've felt on the precipice of a breakdown a few times before and was really hoping this would be a more relevant article.

mynameisvlad4• 2 years ago

https://www.webmd.com/mental-health/signs-nervous-breakdown

Not all breakdowns come in the form of panic and anxiety attacks. Those are certainly a way that breakdowns can manifest, but it’s not the only way. Stress manifests in wildly different ways for different people and even stressors.

You weren’t in his head, experiencing what he was experiencing, so it’s pretty much impossible to “diagnose” from the outside. It certainly sounds like he was functionally paralyzed for hours, even if he didn’t have a full blown panic attack.

sillysaurusx3• 2 years ago

That's a good point. I didn't mean to gatekeep, which is what I ended up doing. Thank you.

If he was functionally paralyzed for hours, then that absolutely qualifies. I was reading through it and thinking "If you can debug stuff, you're probably not having a mental breakdown" and wanted to highlight that some people do break down (which is why it's called a breakdown) and that it's ok.

justinclift• 2 years ago

Xanax is worth having on hand

Doing a search online seems to indicate that Xanax can be addictive?

https://www.drugs.com/xanax.html

Doesn't seem to be the kind of thing to take lightly?

sillysaurusx3• 2 years ago

You're correct. Don't take it lightly. But have it on hand for an emergency, if you're prone to panic or anxiety attacks.

Those are relatively rare, so you shouldn't end up taking it at a frequency where addiction can manifest. But everyone's different; keep open communications with your doctor.

caseyy6• 2 years ago

Doesn’t usually cause addiction if used in emergencies only and quite a few people in tech do.

I do feel more stressed 12-24 hours after taking benzos though. This leads some people to take a higher dose and you can see where that leads.

But no, no addiction in me or anyone I know using it “as needed”.

Some doctors prescribe it daily for a long period of time, which is reckless if other drugs are available for the same problem, like sertraline for anxiety. But then, some doctors prescribe oxy for pain when ibuprofen will do. Malicious stuff. All the addiction warnings are to steer people away from these drugs in these scenarios.

Benzos 3x a day for a month will be tough to shake. Benzos 3x a month to 3x a year — hard to imagine problems with that.

StefanBatory2• 2 years ago

I think it's just a default warning; any drug can me. I am taking Sertraline for anxiety and even while I went through it with my psychiatrist about if it can be addictive (his statement was - no, not really - and for me that was true, I was able to stop taking it without any side effects), there's still that risk.

As with any medical thing, better consult it with a specialist anyway.

kayodelycaon3• 2 years ago

It really doesn't take much to be effective if you don't take it regularly. It takes over a week for it to be addictive.

One of the potential side-effects of the medication is after you stop taking it your baseline anxiety will be worse for a while as your body resets. Similar to caffeine and feeling tired.

Depending on how long you're trying to use it you may need to slowly increase the dose to offset increased baseline. This isn't really dangerous using for say.. one week every two months.

It is possible to get psychologically dependent by frequently using it instead of strategies you would learn in therapy. This also happens with things like alcohol.

Using medication to deal with occasional severe anxiety or panic attacks is far safer than substituting it with alcohol.

jmprspret• 2 years ago

Xanax is a benzodiazepine. You get "high" when taking just barely more than instructed. It's addictive. It's abused frequently. I've abused it before. It feels nice.

I would be surprised if you asked your GP "I have mental breakdowns now and then, I'd like some Xanax", and they just handed it out to you. Then again, I don't live in the US, so maybe the doctors there do just dish it out.

Der_Einzige• 2 years ago

One thing I did not expect about working in enterprise tech companies was the amount of coworkers who popped Xanax regularly.

I've been an anxiety basket case my whole life and the idea that an addictive pill could take it all away terrifies me - I'd be hooked for life.

digging• 2 years ago

I'd be hooked for life.

I mean, that's how a lot of prescriptions work anyway. I take daily medication to prevent anxiety and depression. It's not addictive, but I try pretty hard to never miss a day, because anxiety and depression will get in the way of me living my life. Assuming Xanax is safe to take daily (I have no idea), the only difference is that a supply disruption would have side-effects. But if I run out of my medication, I'll get depressed within weeks, almost guaranteed, withdrawals or no.

supportengineer2• 2 years ago

It's hard to compete against the people who are "enhanced" with drugs.

kayodelycaon3• 2 years ago

It depends on the type of anxiety and how you're using the medication. If you it routinely to handle everyday stress, you're heading down a very bad road. If you have events a few times a year that are extremely stressful, medication can really make a difference.

I had the same fears and I'm not gonna say you should try medication, but I will share it my psychiatrist did. She wouldn't let me have more fifteen low-dose pills at any one time.

Consider your relationship with alcohol. If you already have a problem with it or completely abstain from it, I would strongly recommend avoiding medication.

dotnet00• 2 years ago

I'm the same, and similarly, the idea of taking a pill to make it go away is terrifying.

I have to keep myself very disciplined on the "helpful" medications I do allow into my life because I know how easy it'd be for me to get hooked and start depending on the meds to avoid my problems. Instead I've been working on training myself to see through my nerves.

Alcohol at least has the drawback that the few times I've had it, I've had such an awful experience that I don't associate it with avoiding my problems.

zer8k• 2 years ago

I shouldn't have to drug myself to cope with the stress of constantly half-assed released, features getting pushed in without thought, and 3AM pagerduty alarms because of these things.

We will likely see a large cohort of people who got into tech in the mid 2000s dying off from stress related disease.

xxr• 2 years ago

My wife is a critical care nurse, and she used to work at a hospital in an area with a lot of tech workers about five years ago. She saw a lot of heart attacks in men in their late thirties and early forties that she rarely saw elsewhere.

dangoodmanUT• 2 years ago

write my own everything gang rise up

dbacar• 2 years ago

Since you were able to think and act, I would not call this mental breakdown. That kind of thing is very, very different.

jppope• 2 years ago

Love the story. The question for the author of course... what did you learn and how can you keep this from happening again

robinhouston• 2 years ago

As others have mentioned, the bug that led to this late-night stress was a one-line change to the PostHog library[0].

I take this as a reminder of the importance of giving precise names to variables. The code

    res = await originalFetch(url, init)

looks harmless enough. But in fact the `url` parameter is not necessarily a URL, as the TypeScript declaration makes clear:

    url: URL | RequestInfo

The problem arises in the case where it is not a URL, but a RequestInfo object, which has been “used up” by the construction of the Request object earlier in the function implementation and cannot be used again here.

It would have been more difficult to overlook the problem with this change if the parameter were named something more precise such as `urlOrRequestInfo`.

(A much more speculative idea is the thought that it is possible to formalise the idea of a value being “used up” using linear types, derived from linear logic, so conceivably a suitable type system could prevent this class of bug.)

[0] https://github.com/PostHog/posthog-js/pull/1351/commits/2497...

AlotOfReading• 2 years ago

The problem with linear/affine type systems is that they have an incredibly high barrier to entry. Just look at ownership semantics in something like Rust. They're not impenetrable (especially with experience), but they're severe enough to be the number one complaint for learners.

adamredwoods2• 2 years ago

So Zarar needs to keep the localdev as close to prod as possible, or have a separate pre-prod environment that can run integration tests to catch vital function disruptions.

sltkr• 2 years ago

Your advice is useful for detecting bugs in the code that you release, before your push it to production, but it would not have helped here, because the bug was in Posthog code that was pushed to production asynchronously.

In fact, it would have made debugging this particular issue harder, because the difference in Posthog configuration between dev and prod is what clued the author in on Posthog causing the problem.

To avoid this kind of problem, the solution is to avoid “live” dependencies which can change in production without your testing. Instead, pin all dependencies to fixed versions and host them yourself.

adamredwoods2• 2 years ago

Ah, thanks, I didn't know it was a live dependency.

> Instead, pin all dependencies to fixed versions and host them yourself.

Indeed.

coolhand2120• 2 years ago

Side loading 3rd party scripts in a critical path is asking for problems. Try https://partytown.builder.io/ runs 3rd party scripts like this in a web worker. I'm not sure it would help in this case. Maybe? Probably couldn't hurt to try.

lifeisstillgood2• 2 years ago

Weirdly I think this is heavily related to social anxiety / shame - as in “everyone will knowingness me and point”. This is buried so deep in our brains it’s almost certain to do with herd behaviour.

And it’s (IMO) why anonymity online is usually a bad idea - we need to learn, deep in our bones, that what is said online is the same as standing up in front of the church congregation and reading out our tweets - if you would not in front the vicar, don’t in front of the planet.

gosub100• 2 years ago

you would not in front the vicar, don’t in front of the planet.

Every tyrannical government in history would love this maxim.

Apocryphon• 2 years ago

Not really. One can be civil without self-censorsing oneself. Candidness is not the same as rank rudeness.

codexb• 2 years ago

Relevant [XKCD](https://xkcd.com/2347/)

languagehacker2• 2 years ago

I know that it's not a real post-mortem, but this is the opposite of what a good post-mortem looks like.

It includes:

* Blaming the tools (and the author)

* Not focusing on facts in the timeline

* Not considering improvements

But that doesn't make for engaging content, right?

* At $TIME we observed HTTP POST calls failing
* At $TIME customers reported inability to make changes to ticket prices and promo codes
* $PERSON took the following steps to debug...
* Root cause: an update to a vendor library resulted in cascading failures to the site
* 5 whys (which might include lack of defensive programming, the use of a CDN without a fixed version, etc. etc.)
* Next steps: pin the CDN version or pull the dependency into the build, etc.

Actually, that still looks like a pretty good story to me without any of the associated mania.

photonthug3• 2 years ago

The trouble with post-mortems is always the things that you can't say.

Are we going to talk about how someone wanted analytics everywhere even though it's expensive and they don't use it, that they chose the vendor and didn't provide time to evaluate it, that they wanted it added now-now-now to production even though it hadn't gone through dev, even though it was unplanned work in a busy sprint with other risky work, and that they wanted it on the critical path despite specific objections from some in engineering?

Not saying any of this was the case here, but this kind of thing happens frequently. While engineering is saying "let's use blameless post-mortems to highlight problems in the system" a lot of other people are thinking "now we get to see who could or couldn't fix a problem we caused, focus on that, and downplay our own role in causing trouble". Devs talking about CDNs and HTTP-POST and engineering processes are super helpful for steering the conversation away from broken business processes.

Easily 80% of PMs I've been involved in probably come down to "we had rules for moving tested code through dev/qa/prod, but we were forced to ignore the process we had agreed to because $PERSON / $DEPARTMENT said so". No one can actually talk about the elephant in the room though, because it's career suicide after you're branded as not a team-player. For all the kids out there.. it's really important to learn to be able to recognize the difference between a PM that's going to be an earnest effort to understand things, vs a PM that is going to be an unhelpful but necessary ritual of humiliation.

fsociety2• 2 years ago

Let’s not pretend it’s just PMs. I’ve worked with PMs who fought against shipping fast but ultimately an engineering manager or engineer comes in and forces a fast ship.

languagehacker2• 2 years ago

Good post-mortems should provide ammo and action items to review existing business processes. If you don't have an advocate in your engineering organization who is working across departments to improve these processes, that may be a sign of organizational immaturity or dysfunction. No organizational function happens in a vacuum, and in the right company, collaboration can help influence the things that engineering doesn't directly control.

ziddoap2• 2 years ago

It's also not a good fantasy novel or stage play.

It seems pretty silly to look at a blog post that is very obviously not meant to be a technical post-mortem and criticize how it isn't a good one.

begueradj• 2 years ago

That's a display of endeavour and persistence. Congratulations.

flumpcakes• 2 years ago

I've been on my end of plenty of operational outages. I don't want to be harsh but this could have been written by one of my colleagues, the type of colleagues that I really wish I didn't work with. Console logging for hours? Randomly disabling things? Sometimes when you feel "imposter syndrome" you shouldn't ignore it and maybe up your game a bit.

In fact, I have dealt with an extremely similar situation where a bunch of calls for one of our APIs were failing silently but only after they had taken card payment transactions. Dealing with the developers of this system was like pulling teeth, after we got them to stop stammering and stop chipping in with their ideas (after half a day with this issue ongoing) it took 10 minutes to find the culprit by simply going through the system task by task until we got to the failing task (confirmation emails were unable to send so the API server failed for the entire order despite payments being taken etc.).

This only required 2 things: knowledge of the system, and systematic process to fault finding. You would think that developers who have at least the first, being the ones who wrote it, but sometimes even that is a big ask.

Maybe I'm just burnt out from this industry and incompetent people but... come on... no excuses really.

sleazebreeze• 2 years ago

Agree with everything you said.

And then I’d add: Start with reading the error message. In his panic state, he seems to have thought it was a red herring. Error messages are gold. It gives you a concrete thing to work backwards from.

vvoruganti• 2 years ago

A mantra I heard recently that has been helping me with my own 3AM panics was "None of this matters and we're all gonna die". A bit Nihilist maybe, but has been helpful and just kind of removing the weight of the situation

dennis_jeeves2• 2 years ago

Good thinking. Sometimes I have to remind my nerdy colleagues about what a true emergency is. A true emergency is when your granny is having a heart attack and has to be rushed to the hospital. When a big corporate website, that does millions of dollars of transactions is down at night it is NOT an emergency. They have to hire staff specifically for night shifts if they are so concerned about the money that they may lose.

gcommer• 2 years ago

I know solo projects always have an infinite list of "nice to haves". But personally I never skimp on vendoring dependencies.

In my experience, not vendoring has _always_ led to breakages that are hard to debug and fix.

Meanwhile, vendoring is quite easy nowadays. Every reasonable package manager, and even npm, can do this near-trivially.

sethammons2• 2 years ago

the argument is always "the pr that pulls in the dependency is gross to review with dependency updates" -- and there are ways to mitigate that. I vendor dependencies. My customers want stability and that means a bit more process in managing dependencies. Easy win.

kayo_20211030• 2 years ago

I feel your pain. Someone else shot me in the foot. No fun whatsoever.

ThinkBeat• 2 years ago

Whenever I am dealing with a 3rd party services, I like to write a small adapter for it that bridges the connection point and can keep an eye on a few things.

Primary I use a code generator to write most of it.

For huge services it may not be practical, but for most it usually provides a heads up if something stops working. with an integration.

Octabrain• 2 years ago

I can relate to this. I used to be on call for many years and honestly, it destroyed my mental health. In the last company I did it, it felt like falling in a meat grinder for a week. I remember once spending a whole weekend giving support on an bug that was introduced by a recent release. 72 hours of working non stop. Because of that among others, I got a severe burnt out that took me to the deepest dark place I've ever been.

To this day, I simply refuse to do on call. There's no enough money you can pay me that would make me to suffer that again.

PS: Fuck you, Rackspace.

tristor• 2 years ago

During the mid-early part of my Ops/SRE career I had a senior who was a mentor to me. I noticed as we dealt with outage after outage he was always calm, cool, and collected, so one day I asked him "<name redacted>, how do you always stay so calm when everything is down?" That's when I found that before he'd been in tech, he'd be a Force Recon Marine and had been deployed. His answer was "Nothing we do here is life or death, and nobody is shooting at me. It'll be alright, whatever it is."

While I have never experienced anything similar myself, it really helped me to put things in perspective. Since then, I've worked on some critical systems that were actually life or death, but I no longer do. For the /vast/ majority of technology systems, nobody will die if you let the outage last just a few hours longer. The worst case scenario is a financial cost to the company who employs you, which might be your own company. Smart companies de-risk this by getting various forms of business insurance, and that should include you if it's your own company.

So, do everything you can to fix the outage, but approach it with some perspective. It's not life or death, nobody is shooting at you.

wmichelin• 2 years ago

Pin your dependency versions! The rollback should've fixed things :(

wonderwonder• 2 years ago

Nothing good comes from working at 2am. Company I worked at pushed a major waterfall upgrade one day. Completely tanked the database while attempting to update the schema.

I spent hours on a call with the clients sr. engineer and we eventually came up with a script to fix it. It was after midnight, my director said, good job, you are tired, I'll run the script, call it a night.

An hour later director ran the wrong script... and then called me.

Clients sr. engineer was legitimately flabbergasted, only time I have ever seen that word apply in real life.

Was a not good, very bad day.

recursive3• 2 years ago

B&H Photo Video shuts down intentionally for a whole day every single week, and as far as I can tell, they're one of the top retailers of pro/prosumer A/V electronics.

seanthemon• 2 years ago

We once planned an international trip with my dad to new york around visiting b&h to buy electronics about a decade ago.

Best memory was walking in there and seeing all the conveyer belts above me shooting packages across.

merek• 2 years ago

During these situations, a prompt message can be very reassuring for users, for example:

"We're investigating an issue affecting $X".

As a user, I can rule out that the issue is at my end. I can focus on other things and I won't add to the stack of emails.

This is one of my biggest frustrations with AWS being slow to update their status page during interruptions. I can spend much of the intervening time frantically debugging my software, only to discover the issue is at their end.

darepublic• 2 years ago

I remember as a junior dev getting a call at around 3am from Indian tech support about a failed deploy I had been heading. I stressed myself about it so much and only later reflecting back do I realize nobody but me cared.

Also funny that the culprit was posthog since I have some past experience with it.

KennyBlanken• 2 years ago

The thing that struck me the most was how wildly unprofessional Paul D'Ambra's comments are in response to the bug report.

And then he rolled out a fix that was broken, too - showing incompetence in development, understanding the problem, and a total failure to do proper QA on the fix.

Royally fucked the pooch twice and he's all "gee golly whillikers!"

gsora• 2 years ago

This post resembles how I’ve been feeling lately at work, too bad it’s been months like this now!

hoseja• 2 years ago

I'm so glad I don't have to work with these byzantine JS monstrosities.

nurettin• 2 years ago

More like "The anatomy of a calm and collected response while facing a dire situation thanks to years of expertise, eminem and my sweet wife."

The anatomy of a 2AM mental breakdown

Related Stories