Slack is down • Hacker News Board

Re: Slack is down

#711

Post by dheera » Mon, Jan 04, 2021, 9:33 PM UTC

welterde wrote: post #672

dheera wrote: post #661

Earlier quoted context omitted.

> self-hosted How often is Slack/Discord down? I mean it's not perfect, but I really honestly don't think I could match their uptime by self-hosting, as well as more on-call rotations for something that's not core product. I very much prefer that for something that isn't core product, if it goes down I need to do exactly nothing for it to come back up, and that the engineers at Slack will be starting to work on it li…

I would be surprised if you couldn't match or exceed slacks uptime running whatever alternative you want (IRC, mattermost, rocketchat, etc.) on a random dedicated server. Hardware is quite reliable these days. And updates can be scheduled to be at a convenient time for the team.

Yes, but what if you're taking a few days off to backpack in the wilderness with no signal while it goes down? Who deals with the downtime?

Re: Slack is down

#712

Post by megous » Mon, Jan 04, 2021, 9:40 PM UTC

thewebcount wrote: post #703

nix23 wrote: post #696

Earlier quoted context omitted.

>support is a better use of our short time, more often than not Not when it's down.

Which open source solutions never go down?

Everything goes down. But it looks like huge complicated distributed services shared by huge amounts of people, that are continuously updated and developed, and are constantly trying to attract more users/load, seem to go down more than a simple service on a simple server.

No hard data though. My mail server only ever went down when I upgraded the server and didn't check that everything was still working right away, or similar maintenance induced incidents. It never went down by itself.

Such systems only ever go down unpredictably on HW issues, or when overloaded/out of resources. Neither is very likely, because you're not trying to grow your service in any sense similar to VC backed enterprises. Most of the time it has constant very low load and resource use. And you can simply stop introducing changes to the system if you need more stability for some time. (stop updating, for example)

Re: Slack is down

#713

Post by feanaro » Mon, Jan 04, 2021, 9:41 PM UTC

zenexer wrote: post #649

Arathorn wrote: post #645

Earlier quoted context omitted.

Element (formerly Riot/Vector), has improved loads over the years, and the default matrix.org average send time is around 100ms these days rather than multiple seconds: https://matrix.org/blog/2020/11/03/how-we-fixed-synapses-sca... has details. I suspect you (and the parent) may be running off stale data. That said, Element could certainly use less RAM, irrespective of Electron - and http://hydrogen.element.io is ou…

It has, and I’ve been using it since its early days. I still use it. It’s still terrible, just slightly less terrible. And, no, messages don’t consistently send in 100ms on the default home server; there are regularly disruptions that cause significant delays, sometimes as much as 10-20sec. That’s a big problem for a federated chat platform. Edit 1: I want to love it; the design is everything I could ever hope for in…

It's weird that you're calling it Vector when it's now called Element and it was called Riot for years before that.

Re: Slack is down

#714

Post by vector_spaces » Mon, Jan 04, 2021, 9:42 PM UTC

thesuitonym wrote: post #634

I'd like to take this moment to mention self-hosted, open source, and federated alternatives like XMPP and Matrix. I'd like to, but unfortunately I don't feel like I can in good faith. Matrix is woefully immature, and suffers from a lot of issues, but I think is closer to being a functional Slack/Discord alternative. XMPP is much more mature, and works very well for chat, but doesn't have a nice package that does all…

Yep, I've stopped recommending Matrix because

1. There is virtually zero user-facing documentation. Need to know how to backup keys, verify another user, or what E2EE means? Ask your server operator. Basically the onus is on operators to document this stuff for their users. Except the stuff we're documenting is hard even for server operators, and especially challenging to document in a way that both nontechnical and technical users can understand.

2. Because this stuff is challenging even for more technically minded users to understand, it leads to a kind of burnout for interested non-technical users where they learn all they can about some feature and how it works at a high level from out of date random blogs, try to use the (complex, multi-step) feature, but then something won't work, and it isn't be clear whether it was because the user did something wrong or because the clients or server implementations are broken

3. Issues where core functionality is broken (e.g. two mutually verified users on my homeserver haven't been able to talk to each other in months -- see [1], [2], [3]) languish for months with zero response from maintainers.

4. While core functionality is both broken and undocumented, the maintainers announce rabbit hole features that no one asked for and seem very much like distractions, like their recently-announced microblogging view/client[4]

In short the Element maintainers have shown little interest in making the platform accessible to the people who need its differentiating features the most, and have prioritized the "mad science"/technical aspect of their platform at the expense of the human element (end-users and operators).

It'd be cool if Element used their resources to hire some UX folks and community advocates whose sole focus is addressing the horrid accessibility of their platform. I think most users would rather see that than further "mad science".

[1] https://github.com/vector-im/element-ios/issues/3762

[2] https://github.com/vector-im/element-ios/issues/3572

[3] https://github.com/vector-im/element-ios/issues/3393

[4] https://matrix.org/blog/2020/12/18/introducing-cerulean

Re: Slack is down

#715

Post by johnc1231 » Mon, Jan 04, 2021, 9:44 PM UTC

politelemon wrote: post #585

adrianpike wrote: post #310

Earlier quoted context omitted.

Same - and it's just... weird. The "everything must be in a thread" model feels really clunky.

Subjective... I've found Slack and co's interspersed conversations far too chaotic, and temporal; threading is a great way of organising many different concurrent topics. And to be clear I don't mean Slack's implementation of threads which is hiding it away in a separate panel and which doesn't get used by everyone either.

I find Zulip to be a nice balance. Threads are much more prominent than they are in Slack, but aren't clunky.

Re: Slack is down

#716

Post by wrycoder » Mon, Jan 04, 2021, 9:45 PM UTC

exabrial wrote: post #395

For a product that is so simple, there are no good self-hosted alternatives. Mattermost and RocketChat are written very poorly, reliability and getting your data out is impossible. Slack goes down so often we're thinking of writing a very boring clone that uses ActiveMQ and MySQL, just because chat should be boring and needs to "just work".

Zulip can be self-hosted - have you looked at that? I like the threads implementation.

Re: Slack is down

#717

Post by welterde » Mon, Jan 04, 2021, 9:52 PM UTC

dheera wrote: post #711

welterde wrote: post #672

Earlier quoted context omitted.

I would be surprised if you couldn't match or exceed slacks uptime running whatever alternative you want (IRC, mattermost, rocketchat, etc.) on a random dedicated server. Hardware is quite reliable these days. And updates can be scheduled to be at a convenient time for the team.

Yes, but what if you're taking a few days off to backpack in the wilderness with no signal while it goes down? Who deals with the downtime?

If you are the only technical person on your team then it's of course not ideal and would require some further thought into making things redundant. But even that is easy enough to do with IRC (setup two servers, link the irc servers together, single DNS record that points to both servers - job done).

If there are other people on the team that have _some_ technical skills then they can fix it..

IRC lacks quite a few features compared to other solutions, but the reduced complexity does bring very low operational complexity.

Re: Slack is down

#718

Post by mey » Mon, Jan 04, 2021, 9:53 PM UTC

jaywalk wrote: post #458

mey wrote: post #264

Why does the slack client not show connection issues instead of just hard locking up?

When it went down fully and I had the Windows client open, it went to a page that basically said "Slack is down, we don't know why, try restarting and see if that fixes it. Here's the status page." It would be nice if they could fix it so that a fresh start also goes to that page, at the very least.

How do you have the Slack app installed? I currently have it installed via the Windows/Microsoft Store, and I suspect that is a significant part of the problem.