[Tostino]

For someone who has tried to avoid hearing as much as possible about this drama (and didn't notice this was about Twitter until looking at the comments), what are the condensed "lesson learned" takeaways if you don't mind elaborating?

[CameronNemo]

Don't fire or alienate hordes of engineers with firm-specific talent. Especially for a product/service as mature as Twitter.

[JeremyNT]

I honestly think the engineering/ops problems are the least of their failures. They bled revenue not because of outages, but because wild policy swings and chaotic management style alienated some of their biggest customers.

If they had frozen features and left the existing policies in tact, I suspect we would have a dramatically different narrative about the layoffs. If brief interruptions like this are the worst that happens when you cut engineering to the bone, it's a good argument that is Twitter was indeed wildly overstaffed.

Instead, though, we have a company in crisis due to its mismanagement of other areas, so we're primed to view stuff like this through the lens of that broader failure.

[raydev]

> If brief interruptions like this are the worst that happens when you cut engineering to the bone, it's a good argument that is Twitter was indeed wildly overstaffed.

Overstaffed in order to maintain Twitter as a static service that never ships new things, sure.

I guess they've been able to ship some things that the old Twitter had already implemented and/or a/b tested. But I'm not sure those count. Meanwhile people have been paying in advance for Twitter Blue features that were promised 3 months ago.

[norgie]

I mean the core issue is poor and capricious management either way.

[MrOwnPut]

Or else your site for logged out users will go down for an hour?

What was the reason when Twitter went down pre-Elon?

I really don't see the reason for this artificial drama. Do other "X service is down" threads go this way?

People say Elon is dramatic, but this thread is honestly ridiculous and way more dramatic than anything I've seen him post.

[silentsanctuary]

The frequency and types of outages and failures is significantly more frequent now than pre-Elon. This isn't a surprise to anyone, given Elon's strategy for maintaining Twitter (or not maintaining it, as the case may be).

I don't interpret a lot of "drama" as you put it, but interest. Many observers here are in this field, and follow the "chaos engineering" discipline. Some of them use tools like "chaos monkey" that simulates a metaphorical monkey running through your server room turning off random things, to see how well your resilience systems cope. It's a rare and greatly interesting sight to such practitioners to get to see what happens when the monkey "disconnecting the more sensitive server racks" is a more literal one.

[prenoob]

Well, the fail whale was a meme way before the musk era..

[favorited]

The Fail Whale hasn't been used since 2013.

[raydev]

Yes, almost a decade ago.

[somsak2]

is there data on this? doesn't feel like outages are more common to me.

[RC_ITR]

>What was the reason when Twitter went down pre-Elon?

To use a tennis metaphor, good players minimize unforced errors and recover quickly from forced errors.

This is a very clear unforced error that could likely have been prevented by just waiting to roll out the new feature.

To extend the tennis metaphor, it would be like Serena Williams losing a set on 50 double faults. Sure, she's lost other sets before, but it would be notable for her to lose in such a unique way, even if she still went on to win the match.

That's why people are talking about this, it's a very weird way for a site to fail and it's interesting how it happened.

[ISL]

Alternatively -- don't take on debt for an acquisition incommensurate with an acquisition-target's capacity to service that debt.

Musk's cost-cutting may yet pan out from a business perspective, but it seems to be a pretty risky move.

[MrOwnPut]

The lesson is if Elon runs Twitter, an hour long downtime takes up everyone's day so we can all speculate.

Before, it was "healthy for everyone" when Twitter went down (search hn.algolia.com for pre-Elon Twitter downtime threads)

Now it's an episode of Real Housewives just like any other Elon thread on HN.

[camjohnson26]

He created that drama when he tried to gaslight the entire tech industry into thinking 90% of engineering jobs were redundant, you can’t expect anything else than for him to get dunked on when the completely predictable consequences of that play out, namely that the service will get more and more unreliable over time.

[darthrupert]

It does hurt the proud IT worker in me, the knowledge that half of us can be thrown away and almost nobody notices.

Perhaps the takeaway is that we should get our shit together. The opposite of r/antiwork.

[MrOwnPut]

No. I think the people creating drama here should take responsibility and act like professionals.

No need to pull out terms like gaslighting to justify creating more drama.

A 1 hour downtime is nothing.

[ModernMech]

> A 1 hour downtime is nothing.

If it’s nothing, why is Musk on Twitter saying the whole codebase is brittle for no reason, and needs a complete rewrite? Sounds like a pretty big deal to me.

As far as the drama comment, you’re free to disengage from this thread at any time. No need to continually interject how against the drama you are. Just leave if you find it distasteful.

[MrOwnPut]

A 1 hour down time every infrequently is nothing, but at the same time a codebase can be brittle. Are you saying the two are mutually exclusive?

Brittle code means it's hard to change, and causes issues when changed, but other than that it's stable when running.

And I have every right to point out the drama here. People here do somewhat have the right to create drama, but it is distasteful and unprofessional and I will continue to call it out.

Glad for you to admit there is drama here though :)

[roboyoshi]

- When you fire a lot of engineers, legal, security experts... things will still somehow keep working - Also some people still want to keep working at this place (or must, because of VISA etc); It probably happens all the time, just this time everyone was watching it in full motion. - I was sure this whole takeover was the end of twitter, but somehow with how many users twitter has, it just won't die. Goes to show the advantages of the mass.. probably the same applies to microsoft/google.. - The whole 8$ for a checkmark story.. which destroyed the whole trust in the checkmark basically immediately. The whole verification process (how that worked) - many people shared some insights on that and how much effort it was to fine-tune all of that with the scale of twitter. - The free API just being shut down with basically zero communication to devs. and somehow getting away with it. I guess this is a reoccuring theme in tech that a small company has to be open for developers until it's big enough to turn on them. - With the reduced engineering and the new management, it seems more errors slip into production (and incidents like the one now) - or maybe it just feels like it with all the focus on twitter right now - but anyways, you just see things breaking you normally would not expect to see.. - How people are communicating, or in this case not communicating.

I'm probably not the ideal person to write this down, there was so much stuff going on, some people probably made a whole blog-article series on this. This is just a few of the things that I'm able to remember right now.. and with everything on hn here it gives you some ideas and things to think about. Hope that Helps.