The issue is with a type of database process which is dying for some reasons. We are in the process of figuring out why so we can fix and deploy a patch and get things working. We did some changes earlier but obviously it didnt work.
We have some good people working on this issue right now and hope it wont be too long.
We dont want to deploy a fix only to mess people about with another crash. The idea is to now make sure we get it working properly.