Status (by brong at Tue May 29 11:31 UTC)
There appears to be a bug in our frontend redirection system which has caused users to be directed to servers which were being restarted for maintainence. This shouldn’t have happened.
The result of the bug is that some users will have seen 500 errors for about 10 minutes, depending which server they were on.
All servers are back up now, and we will fix this issue before shutting down any more web servers.
Status (by robm at Tue May 22 12:40 UTC)
iOS users trying to access their email via an SSL connection would have received errors trying to connect to the server for a few hours earlier today.
The problem was related to a configuration change that attempted to increase the security on our server. No other client seemed to have a problem with the change, but it caused iOS SSL connections to fail.
For those curious, the change was to increase the Diffie Hellman (DH) key exchange from 1024 to 2048 bits. We’re not sure why this causes iOS to fail, but we reverted the change when we found out.
We’re committed to moving all users to SSL/TLS connections over the next few weeks.
http://blog.fastmail.fm/2012/05/21/enforcing-ssltls-encryption-of-all-connections/
Status (by brong at Mon May 21 14:11 UTC)
There will be a short (1-3 minute) outage rolling across all servers for an urgent update.
Update (by brong at Mon May 21 14:23 UTC)
All finished, everything is working normally again.
Status (by brong at Wed Apr 25 10:15 UTC)
We’ve just had a very brief outage for all users as an urgent update was rolled out. Everything should be working fine now.
Status (by robm at Mon Apr 23 15:34 UTC)
A small number of users will have been experiencing intermittent problems using the web interface for the last hour. We’ve now tracked down this problem and fixed it. Appologies for any inconvenience.
Status (by robm at Tue Apr 17 05:51 UTC)
One IMAP server is down affecting email access for a small number of users. We’re investigating.
Update (by robm at Tue Apr 17 06:14 UTC)
All services have now been restored
Status (by brong at Wed Mar 7 15:04 UTC)
I’m just restarting Cyrus everywhere, so there will be a small (~ 1 minute max) outage for each server as the restarts are run.
Status (by brong at Thu Mar 1 13:30 UTC)
Each server will get a short outage (1-2 minutes) as we restart to roll out an urgent fix.
Status (by brong at Tue Feb 21 15:43 UTC)
Oldweb was down for a while today because it was still trying to connect to an old database for the ‘log’ DB when we changed DB engines.
It should be working fine again now.
Status (by brong at Fri Feb 17 14:02 UTC)
Looks like the same problem as yesterday – rebooting now.
Update (by brong at Fri Feb 17 14:12 UTC)
The server is restarted – services should be back to normal
