We're currently experiencing a database problem that is affecting some services, eg sending/receiving of email. This is being looked in to at the moment.
Email services are running. We are needing to repair our database cluster. No email would have been lost as part of this problem, but some may have been delayed.
Whilst the database was fixed overnight, the problem has reoccurred. We're investigating this again now.
We have restated our database cluster on the previous kernel. The database is now operational, there was a period of a few minutes where some emails would have been delayed.
Services have been running fine, but as we have had identical database problems twice within the space of a few hours we'll keep this incident open for now.