2014/04/08 - staff IMAP server slow/unavailable
One of the backend mailbox stores behind is seeing an abnormally high load at the moment, and data is being read/written faster than the disks can cope with. This backend store hosts primarily staff and postgraduate students; the problem does not affect the other backend (which contains mostly undergraduate students).

This will make accessing your mailbox very slow. Repeatedly polling your mailbox will compound the problem.

As a result of this, a large amount of mail has backed up, and attempts to deliver this are also compounding the problem.
In an effort to reduce the total load being experienced by this backend, we are trying to clear the backlog of mail deliveries. We've done this by temporarily suspending IMAP and POP3 access to mailboxes -- at present you will be unable to poll for mail. Once the backlog has cleared, we will re-enable normal service.
It is looking increasingly likely that the root cause of this is the same physical host server that caused the outage last Tuesday. We're still investigating.
The underlying physical host is almost certainly now involved in the problem. We have opened a severity 1 support request with VMware.
With help from VMWare, we've managed to confirm that the physical host was in fact responsible for the slowness we've been seeing with mail. We've also, with their assistance, managed to migrate the affected IMAP backend to another physical host.

Access to mailboxes has been restored, and there should be a noticeable improvement over the performance we saw earlier this morning.

However, the problem on the underlying physical host has not yet been resolved. This may have implications for the scheduled maintenance later today.