One of my favorite pastimes over the last year or so has been to tear Twitter apart over uptime issues. See, for example, Twitter Suffers Minor Period Of Uptime (my personal favorite), Twitter: Something Is Technically Wrong and Amateur Hour Over At Twitter?.
Twitter has been mostly stable for the last month or so, though. And today cofounder Biz Stone wrote an email to Twitter users that says they’ve had 99.9% uptime over the last couple of months:
Reliability Update
Twitter has been making great progress in terms of uptime and reliability. Fail Whale sightings are far less frequent these days thanks to our efforts but we still have a long journey ahead. Last month we saw 99.88% uptime and so far this month we are at 99.96%. Our engineering and operations teams have been taking a very methodical approach to improving Twitter. We’re using the word “craftsmanship” to characterize our work here at the office. Reliability and dependability continue to be top on or list of key goals.
I’d say the email is tempting fate a little, but I’ve called Twitter on that same issue before and they failed to fail. Perhaps the worst of Twitter’s scalability issues are behind them (knocks on wood).
It’s been wonderful not seeing the big whale lately. I think they finally got the issue under control and the best is yet to come for Twitter.
They hire 3 guys to watch the server 24×7 instead of using only one guy. One person might get sleepy. So, whenever he is sleepy, the server will be down.
Yep got the mail..
Good to know that it was online almost always this month! But i never cared about that. It was on when i was tweeting! That’s enough for me.
AMEN !
Congrats to the Twitter team! Probably sweeter for all the criticism they went through.
they are up most of the time(after past lessons) and hope they will continue to stay so
$15mil in funding and they can’t keep the site up? Maybe they should find some 13yo’s who are running warez sites off HostGator hosting accounts and consult them…
Gregor,
“Last month we saw 99.88% uptime and so far this month we are at 99.96%.”
Is that not good?
Hardly any Fail Whale sittings for me. When there were, sightings lasted less than 10 seconds.
David
If I click “read post” one more time and all I get is a small block of text that could have been left on the front page i’m going to freak.
Long post should get the read post link, other post should be on the home page in there entirety. And don’t cap a limit on the number of characters for what stays on the homepage or gets a the read post link….use your good judgement.
I understand the whole faster page loading and other concerns, but reading the site has become more cumbersome on a browser and on my iphone
Cheers
I picked a perfect time to get into Twitter, once they fixed the problems
I think the funniest Twitter Jab came from Owen Thomas at The Wag:
“Twitter, a famous provider of Failure as a Service, is down again.”
I’m still in using my twitter, maybe because in the right time.
Twitter has seemed more stable lately. Though, the second half of August my RSS feed wasn’t publishing.
This is some great news for Twitter.
there’s an apple event coming up, isn’t there? we’ll see then if this is true or not.
They have definitely had more up time in the past 4 months from what I have noticed.
Hi,
I know why they resolved the problem…
It is not the Downtime they are concerned about. It is the fear/pleasure of being the first POST on TechCrunch( Home Page )
Now, if they feel they want to be on the homepage of TechCrunch, All they need to do is
$./apachectl stop
(or)
sqlplus > shutdown immediate
Cheers, Nag
http://uptime.n...ite=twitter.com
They’re no longer on Apache, they’re on the “hi” server now. They probably aren’t foolish enough to use Varnish like Techcrunch does. That thing is so buggy.
I have no idea what they renamed to hi, so I won’t speculate. You can rename Apache in httpd.conf, but why would you want to?
I still can’t find a use for Twitter.
Really? Twitter is akin to crack, start once and you can never stop.
I’m still in using my twitter, maybe because in the right time.
Well my favorite feature for Twitter is still not back after many months of being unavailable. This of course is tracking.
Basically Twitter says the underlying issues are being addressed but it will require a significant amount of time to re-architect and re-write the various pieces needing attention.
john adams to the rescue
I’ve enjoyed following @netik as he and the team work through the issues.
Fail whale finally disappeared ! She asked for it, damn whale !
Now if plurk can become stable too, it would be nice .
Offtopic but…
Mike,
Please stop self-linking so much. I love TC but I am sure I am not alone when I cringe at the links to back articles. There are just too many and most of them are gratuitous. They add so little and give no real value to the article.
Have they redone their site in PHP, maybe?
uptime
7:39am up 365 days 3:00, 10043 users, load average: 0.26, 0.27, 0.33
Hmmm, perhaps Twitter needs new a new fail mascot cos clearly the whale is just not cutting the mustard.
Please, give us IM feature back now
It has become “stable” on the cost of the IM feature so id say it isnt that stable……
How does the load compare with that back during Amateur Hour?
from the start it was very impressive until the more traffic came in and it just don’t survive anymore.. worse than a pice of crap!! wishing it would get better though..
“…so far this month we are at 99.96%.”
If they had only gathered the stats few minutes past midnight on 1st of Sep instead of 3rd of Sep they could have claimed 100% availability so far in Sep.
looking at things like alexa they are just about back to the same ranking level of mid-May. So if the trouble started in June, lets see how they cope when all their lost traffic is recovered.
I’ve been monitoring Twitter for a while now, things are better but not perfect… http://www.just...29ab669d9df9944
Luke
Twitter should be more efficient and i hope that they reach 100% uptime soon.