Tuesday, 21 February 2012 9:37 PM
I have a cluster set up with 1 head node and a bunch of compute nodes. I'm thinking that the job queue that HPC uses sits on the head node somewhere. Correct me if I'm wrong here.
My question is what happens to all the jobs that are sitting in the queue if the head node goes down? If the queue gets lost on a server breakdown are there any other methods available during real time that I can use to backup the job queue other than the obvious server backups at specific intervals to another media?
Wednesday, 22 February 2012 12:24 AM
The job queue is actually stored on the SQL Server. By default it is stored on the head node, but you have the option of setting up a remote database for it to use.
You can even go even further by using Microsoft's Failover Clustering service to create multiple instances of SQL and/or the Headnode to provide more durability.
Finally, if worst comes to worst, we do also have a backup and restore feature, which you can read more about here: http://technet.microsoft.com/en-us/library/hh332930(v=ws.10).aspx
Let me know if I can be of more help to you!
Wednesday, 22 February 2012 3:05 AMThank you very much. This is exactly the answer I was looking for.