overworkedops
asked on
Cluster name taking a long time to become online....
I am using the "change group" function on the cluster administrator... it takes quite a long time for "Cluster Name" to get back online.
Any reason this might be?
I can wipe the boxes if need be, but I'd like to understand the reasoning for the problem first! :)
Thanks!
Any reason this might be?
I can wipe the boxes if need be, but I'd like to understand the reasoning for the problem first! :)
Thanks!
How many instances in the cluster? Is it active/passive?
change group will make the node failover to the other one, swaping the resources over.
If the required memory is high or if long transactions needs to be rolled forward or backwards, it can increase the time taken.
The SQL service should restart in less than a minute most times (if failed by hand), asside from transactions.
change group will make the node failover to the other one, swaping the resources over.
If the required memory is high or if long transactions needs to be rolled forward or backwards, it can increase the time taken.
The SQL service should restart in less than a minute most times (if failed by hand), asside from transactions.
ASKER
Single instance, active/passive. I'm going to wipe and see what happens... I'm sure it will fix it.
ASKER
Okay, I've wiped... no fix. The cluster name is still taking a long time to fail over. I wiped EVERYTHING -- this is a cluster that's set up from scratch. Any ideas what it might be? I am thinking something on the domain controller, but I'm not sure.
What is the load like? If any?
ASKER
None... there's nothing on the box except SP1 and the cluster services.
I think it's a DNS issue, but I only find 1 entry for the cluster in DNS and I deleted that before I rebuilt... just weird :\
I think it's a DNS issue, but I only find 1 entry for the cluster in DNS and I deleted that before I rebuilt... just weird :\
Available memory on each node?
Any load on the SAN?
Any load on the SAN?
ASKER
8 GB of RAM, plenty free... it's not a SAN, just DAS, an HP msa 500, using the shared bus.
Mind you, the disk group fails over perfectly and fast, so does the cluster IP... just the cluster name takes a long time.
Mind you, the disk group fails over perfectly and fast, so does the cluster IP... just the cluster name takes a long time.
Have a look in the SQL server logs to see the sequence of events - perhaps they will indicate a particular process that is taking time.
ASKER
It was reverse DNS... works fine now :)
Cool :)
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
ASKER
Thanks again!