cancel
Showing results for 
Search instead for 
Did you mean: 

SF hangs and Cluster reboot

Leoben
Not applicable
Hi,
 
For my first post here, I'm beginning with a tough one.
 
I have a MSCS cluster build on 2 EVA 8000 Arrays. We use SF5 to replicate volumes accross both arrays.
The volumes are concatened and mirrored, each mirror belongs to one different array.
during our most recent tests we tried to mask one mirror from the volumes. SF dealed pretty well with the mirror loss and no side effect been observed on the clustered volumes.
But, one of the test consisted in shuting down (power off) one array. Doing this lead to a SF freeze. Both nodes lose the visibility on the volumes and restart on their own. After the reboot done (obviously Windows was to busy to log anything in the event logs) SF starts and see half of his disks disconnected (which is normal at this point) and restore connections to the degraded volumes correctly. The cluster become available again.
When we restart the powered off array, the rebuild starts and everything return to normal.
 
Did anyone ever face this kind of problem? or anything similar?
Is the Maintenance Pack 1 available for SF5 already? if not when can we expect a release?
 
Thank you in advance for your support.
0 REPLIES 0