7 ms·

GitHub Down Again?

60 points by sudmishra 4y ago

Probably the second time in 12 hrs.

lrae 4y ago
Seemed like it had some short hiccups but seems fine again. Edit: Shouldn't have posted that, seems like they still have some issues.
Maxion 4y ago
Seems to be very slow, periodically works.
dinom 4y ago
Yeah, definitely having some problems.
d3mcfadden 4y ago
unicorns here
topicseed 4y ago
Definitely...
JOnAgain 4y ago
Yes
cr3ative 4y ago
It's having a moment.
- nathanaldensr 4y ago
  One of many.
KptMarchewa 4y ago
Yes. Lately there has been a lot of slowness and 500s.
WreckItRob 4y ago
Seems to be up for me
- xyst 4y ago
  same here
blueflow 4y ago
Repost: https://news.ycombinator.com/item?id=32735734 https://news.ycombinator.com/item?id=32735734
- intunderflow 4y ago
  That post is a day old
  - blueflow 4y ago
    Then its about time that Github fixes the issue.
- irajdeep 4y ago
  > Repost: https://news.ycombinator.com/item?id=32735734 https://news.ycombinator.com/item?id=32735734 LOL, this clearly shows how grave the situation is. Guess it's time to listen again to: https://youtu.be/XimyEvDJWzA https://youtu.be/XimyEvDJWzA
MontyCarloHall 4y ago
Yes, still having hiccups for me. I started having issues about 24 hours ago. What is the reason for this spate of outages? GitHub is a mature, stable product. What core features could its engineers possibly need to be working on that whatever they did broke core backend functionality so badly?
- garciasn 4y ago
  Once Tableau was acquired by Salesforce, shit that had been stable as can be for YEARS started to break as Tableau (I assume) began to interface with SF systems or requirements. Last week, CLI functionality was down for 3-4 days, depending on the cluster you're on. Nothing like crushing production reporting for a week at the end of the month because Salesforce wants X, Y, and Z added to a mature product that had seen no issues for the 3 years prior. My guess is that is the same sort of thing happening here w/Microsoft.
- oneplane 4y ago
  Might simply be infrastructure consolidation since Microsoft bought them. Same reason behind the Azure pipe / GitHub Actions mess.
  - ta988 4y ago
    GitHub actions is still a mess. I have outages once a week (variable time from minutes to hours).
- tpoacher 4y ago
  Probably just Windows Update /s
JOnAgain 4y ago
What is even the point? https://www.githubstatus.com/ https://www.githubstatus.com/ Save money by just not having the page or the people pretending they do something useful by maintaining it.
- gtirloni 4y ago
  "operational" comes with strings attached. We don't know what internal SLAs specify but most likely the service is meeting those agreements. Should we get a refund? :)
  - darkwater 4y ago
    Yes, my company is paying for my GH user seat.
  - pavlov 4y ago
    Maybe there's a business model here? Sign up with Objective Uptime Inc. and tell them which vendors you have SLAs with. OU will monitor them on your behalf, compare the actual global availability to what's displayed on the vendor's status page, and if the outage exceeds agreed limits, OU can even automatically file suits on your behalf for breach of contract and misrepresentation (assuming the vendor's status page was showing all peachy).
- Karunamon 4y ago
  I know over on AWS they are loath to update the status page since this is tied to SLA compensation. Does github have the same problem?
- srveale 4y ago
  "When a metric becomes a target, it ceases to be a good metric" I regret that uptime became a clause in SLAs, or a reputation/marketing thing. I don't care about how many 9s are after your decimal, I just want to know if your service is down or if something is wrong on my end.
- petercooper 4y ago
  One of the many status page/"is it down" services could probably get some good PR out of having a GitHub status page that's based upon live results from several geographical locations..
amelius 4y ago
Decentralize them.
tambourine_man 4y ago
Remember when git came out, one of its features was being a decentralized, distributed version control system? Yeah.
- hk__2 4y ago
  It still is one, but it’s easier to work with it if there’s a single source of truth. You can still work on your Git tree, you just can’t sync with this SSOT.
- ModernMech 4y ago
  I mean... it still is? I can work on my codebase and commit on my local machine. Push remote to Gitlab if Github is down. When Github is back up I can push to Github. What's the problem?
  - darkwater 4y ago
    Do you have the same build and deploy pipeline configured in GitLab as well? If so, congrats, you are nailing it.
- paranoidrobot 4y ago
  Every time there's an outage with $SCMProvider, there's always a comment like this. Git is not GitHub. GitHub is not git. You can absolutely do git in a distributed way. That doesn't help when your business workflow relies on GitHub for access control, artifact hosting, issue management, PR coordination & approval, and all the thousand other things that people use GitHub for.
- gtirloni 4y ago
  Git is decrentralized! You win GitHub down bingo!
  - tambourine_man 4y ago
    I don’t know. If it keeps happening again and again and people keep bringing it up, maybe it’s a remake worth considering.
niffydroid 4y ago
Still better than bitbucket
bluedino 4y ago
It's up, it's just really slow
msdrigg 4y ago
Downdetector shows both the outage last night and one this morning. https://downdetector.com/status/github/ https://downdetector.com/status/github/
rozenmd 4y ago
Good to see it's not just me - seems to be flip-flopping: https://github.onlineornot.com/ https://github.onlineornot.com/
remram 4y ago
Actions haven't run for me since two official outages ago. Multiple jobs stuck as "Waiting for a runner to pick up this job" for a day.
rvz 4y ago
Again? Last time this happened was 12 - 13 hours ago [0] If you are self-hosting like GNOME, wireguard, Redox OS, Wine, etc it seems it is business as usual. But for those who went all in on GitHub, it's pretty much a recurring disaster. At this point, those who are self-hosting might as well say they have better up-time than GitHub had since they are still unreliable even after two years of warning against the 'centralize everything' [1] nonsense. [0] https://news.ycombinator.com/item?id=32744717 https://news.ycombinator.com/item?id=32744717 [1] https://news.ycombinator.com/item?id=22867803 https://news.ycombinator.com/item?id=22867803

GitHub Down Again?

60 points by sudmishra 4y ago

Probably the second time in 12 hrs.

lrae 4y ago

Seemed like it had some short hiccups but seems fine again. Edit: Shouldn't have posted that, seems like they still have some issues.

Maxion 4y ago

Seems to be very slow, periodically works.

dinom 4y ago

Yeah, definitely having some problems.

d3mcfadden 4y ago

unicorns here

topicseed 4y ago

Definitely...

JOnAgain 4y ago

Yes

cr3ative 4y ago

It's having a moment.

nathanaldensr 4y ago
One of many.

KptMarchewa 4y ago

Yes. Lately there has been a lot of slowness and 500s.

WreckItRob 4y ago

Seems to be up for me

xyst 4y ago
same here

blueflow 4y ago

Repost: https://news.ycombinator.com/item?id=32735734 https://news.ycombinator.com/item?id=32735734

intunderflow 4y ago
That post is a day old
- blueflow 4y ago
  Then its about time that Github fixes the issue.
irajdeep 4y ago
> Repost: https://news.ycombinator.com/item?id=32735734 https://news.ycombinator.com/item?id=32735734 LOL, this clearly shows how grave the situation is. Guess it's time to listen again to: https://youtu.be/XimyEvDJWzA https://youtu.be/XimyEvDJWzA

MontyCarloHall 4y ago

Yes, still having hiccups for me. I started having issues about 24 hours ago. What is the reason for this spate of outages? GitHub is a mature, stable product. What core features could its engineers possibly need to be working on that whatever they did broke core backend functionality so badly?

garciasn 4y ago
Once Tableau was acquired by Salesforce, shit that had been stable as can be for YEARS started to break as Tableau (I assume) began to interface with SF systems or requirements. Last week, CLI functionality was down for 3-4 days, depending on the cluster you're on. Nothing like crushing production reporting for a week at the end of the month because Salesforce wants X, Y, and Z added to a mature product that had seen no issues for the 3 years prior. My guess is that is the same sort of thing happening here w/Microsoft.
oneplane 4y ago
Might simply be infrastructure consolidation since Microsoft bought them. Same reason behind the Azure pipe / GitHub Actions mess.
- ta988 4y ago
  GitHub actions is still a mess. I have outages once a week (variable time from minutes to hours).
tpoacher 4y ago
Probably just Windows Update /s

JOnAgain 4y ago

What is even the point? https://www.githubstatus.com/ https://www.githubstatus.com/ Save money by just not having the page or the people pretending they do something useful by maintaining it.

gtirloni 4y ago
"operational" comes with strings attached. We don't know what internal SLAs specify but most likely the service is meeting those agreements. Should we get a refund? :)
- darkwater 4y ago
  Yes, my company is paying for my GH user seat.
- pavlov 4y ago
  Maybe there's a business model here? Sign up with Objective Uptime Inc. and tell them which vendors you have SLAs with. OU will monitor them on your behalf, compare the actual global availability to what's displayed on the vendor's status page, and if the outage exceeds agreed limits, OU can even automatically file suits on your behalf for breach of contract and misrepresentation (assuming the vendor's status page was showing all peachy).
Karunamon 4y ago
I know over on AWS they are loath to update the status page since this is tied to SLA compensation. Does github have the same problem?
srveale 4y ago
"When a metric becomes a target, it ceases to be a good metric" I regret that uptime became a clause in SLAs, or a reputation/marketing thing. I don't care about how many 9s are after your decimal, I just want to know if your service is down or if something is wrong on my end.
petercooper 4y ago
One of the many status page/"is it down" services could probably get some good PR out of having a GitHub status page that's based upon live results from several geographical locations..

amelius 4y ago

Decentralize them.

tambourine_man 4y ago

Remember when git came out, one of its features was being a decentralized, distributed version control system? Yeah.

hk__2 4y ago
It still is one, but it’s easier to work with it if there’s a single source of truth. You can still work on your Git tree, you just can’t sync with this SSOT.
ModernMech 4y ago
I mean... it still is? I can work on my codebase and commit on my local machine. Push remote to Gitlab if Github is down. When Github is back up I can push to Github. What's the problem?
- darkwater 4y ago
  Do you have the same build and deploy pipeline configured in GitLab as well? If so, congrats, you are nailing it.
paranoidrobot 4y ago
Every time there's an outage with $SCMProvider, there's always a comment like this. Git is not GitHub. GitHub is not git. You can absolutely do git in a distributed way. That doesn't help when your business workflow relies on GitHub for access control, artifact hosting, issue management, PR coordination & approval, and all the thousand other things that people use GitHub for.
gtirloni 4y ago
Git is decrentralized! You win GitHub down bingo!
- tambourine_man 4y ago
  I don’t know. If it keeps happening again and again and people keep bringing it up, maybe it’s a remake worth considering.

niffydroid 4y ago

Still better than bitbucket

bluedino 4y ago

It's up, it's just really slow

msdrigg 4y ago

Downdetector shows both the outage last night and one this morning. https://downdetector.com/status/github/ https://downdetector.com/status/github/

rozenmd 4y ago

Good to see it's not just me - seems to be flip-flopping: https://github.onlineornot.com/ https://github.onlineornot.com/

remram 4y ago

Actions haven't run for me since two official outages ago. Multiple jobs stuck as "Waiting for a runner to pick up this job" for a day.

rvz 4y ago

Again? Last time this happened was 12 - 13 hours ago [0] If you are self-hosting like GNOME, wireguard, Redox OS, Wine, etc it seems it is business as usual. But for those who went all in on GitHub, it's pretty much a recurring disaster. At this point, those who are self-hosting might as well say they have better up-time than GitHub had since they are still unreliable even after two years of warning against the 'centralize everything' [1] nonsense. [0] https://news.ycombinator.com/item?id=32744717 https://news.ycombinator.com/item?id=32744717 [1] https://news.ycombinator.com/item?id=22867803 https://news.ycombinator.com/item?id=22867803