The BOINC manager could not report the completed tasks, .


cio_redulla

Ok - this problem has been resolved. There was a query that was updated during the last sync with BOINC that had less then optimal performance. I have changed it to run faster and that has reduced the load on the database considerably. I am also continuing to look for other queries that could be optimized.

We are also close to moving our database from the 2 processor/4 GB server it is on now to a 4 processor/8GB server. Once a couple more parts arrive we should be able to complete this move (we will announce more when it is time because we will have a outage associated with this switch). This will ensure that we can handle considerably more users then we have now.

I know we started a little "push" May 1, didn't realise we'd rock your servers that bad!
Plenty of work here but the morning stats update looks like it didn't go through.

This problem caused a bunch of work to be delayed in getting validated until after the 12:00 UTC stats update. It has been validated now so those stats will get picked up in the next stats update.

The problem is actually starting to look like it wasn't the db server at all but instead it may have been the monitoring tool that we use. It apparently went out of control and put a real high load on one of the servers. Once it was restarted everything went back to normal so we have something that we are going to have to investigate.

I must be sitting on the good 50% again ;P..... just uploaded file and received the post "ready to report" acknowledgment at 10:35 UTC. Knreed wrote that the validation cycle is once per minute.... validation applied and credit granted at 10:36 UTC.

Web responsiveness is good here in the boot of Europe!