I have a cluster consisting of two DL585 G1 Quad processor dual core and two DL585 G2 Quad Processor quad core systems. The cluster is configured with DRS and EVC no HA. From the attached images, the system appears to be servicing happily all requests.
The issue is that since the upgrade, the whole environment is noticeably slower. Our monitoring system is showing a large amount of ballooning occuring on systems and many of the systems that are not ballooning have high memory usage. On checking cpu ready, I have noticed that the average cpu ready is noticably higher than prior to the upgrade.
The only configuration change made after the upgrade was to increase the port count on one of the virtual switches to 120 ports.
I have gone through all the resource allocations in the resource groups and across those issued to individual machines to check for anomalies but it all appears to be correct.
Any suggestions would be appreciated.
Cluster Summary
how can I upload images?
what kind of shared storage you are using?
We have 10k and 15k fibre channel disk in DS4800 and 2*6800 IBM arrays which are virtualised behind a HDS USP-VM. The configuration of the storage is 500GB LUNs for most systems, with DB servers having RDM. This configuration has not changed during the upgrade process. We have a NFS mounted share from HDS HNAS for iso shares.