Twitter – what is going on?
Update: I got an email from Biz Stone. Basically, they are asking for our patience and they are not announcing anything (yet). Biz stressed that they are working hard on making Twitter more reliable, but that it won’t happen overnight.
After a rough night, Twitter has broken up with the host they loved so very, very much (or the other way round). Apparently sometime last night around 10pm, they made the switch to somebody else (can’t figure out who yet – Netcraft hasn’t updated its stats for them since the 24th. Update: Looks like it is Verio).
Based on this comment on a post about Twitter by Steven Hodson on WinExtra, it would seem that the issue had something to do with the Twitter infrastructure:
All I would like to say is applications built on Rails CAN and DO scale – whether on Joyent or another provider. We see it every.single.day.
At the same time, there are rumors flying around that this switch is a possible preamble to a soon to be announced acquisition of Twitter.
I doubt that is true (why would they switch hosts before the acquisition, after all), but we will see, I guess.
My theory: maybe Joyent didn’t want to take the blame for Twitter’s failures anymore and yesterday’s happy post about them was just Twitter trying to smooth things over with them, but didn’t have that effect. But then the outage was just an extended maintenance window, so that might leave my theory in the dark, too.
I have contacted Twitter about this and will let you know if I get any reply.