The power supply in one of our perimeter servers has been intermittently failing overnight causing access issues for both webmail and client access along with website access for some users. The supply is currently being replaced and should be back online shortly.

10am - The fan had failed causing the power supply to fail as it slowly overheated. Entirely new power supply fitted, all back online.

There was partial connectivity for the service this morning between 06:00 and 08:30 after a combined issue of a datacenter routing issue affecting 4 of our servers outside of our control, followed by another mail server crash, identical to last nights problem although this time on a different server. Currently investigating why the 2 mail servers have developed an idential issue within hours of each other.

We have been investigating a new mail storage system for the last 6 months to improve speed and reliability of the entire service. This new system finished preliminary testing last week, and will be starting user testing later this week followed by a migration of all existing members in the coming month or two. Schedules are liable to change if we encounter any issues during testing but we hope to start moving your accounts to the new system as soon as possible. Naturally we can't rush a process like this!

One our mail servers is currently offline and the automatic failover to other available servers hasn't functioned correctly. We're currently working with the datacenter to get the failed server back online - we hope to have the service restored by 17:25. Further updates to follow.

Update 17:20 - Manually removed the failed server from the cluster, service restored for affected users. Apologies for the inconvenience.