Just wondering if anyone has seen a similar open replicator issue to the one i encountered yesterday.
OR hot pull VMAX 10k 987 pulling from a VMAX 20. ~25 devices in 2 sessions.
Everything looked good, the sessions created ok and copies progressed, several devices went to a copied state but noticed a few devices did not complete, just stuck on 480 tracks, thought they'd finish eventually but no, just stuck there. 9 devices in total all with 240, 480 or 720 tracks outstanding, no errors as far as i could tell.
Support dialled in and checked the box, no obvious issues, we were running out of time on our change window so we did a terminate - symforce and created a new session with the stuck devices, they ran though clean,
Just curious, why did you break the migration into 2 sessions for only 25 devices? Is there a best practice recommendation for number of devices in a session?
This was an SRDF/CE cluster and the sessions were just a logical split of devices, local devices for that node in one session and cluster devices in another.
I'm not aware of a best practice in terms of dev numbers, i have a hot push -differential session at the moment from a VMAX source that contains 91 devices (24TB).
When using a DMX as source we used to try to stick to about 30 devices per session max, it just seemed to work better when doing recreates of differential sessions, though we have no conclusive proof on this.
sauravrohilla
859 Posts
506
1
Posted February 11th, 2014 04:00
This is a very common issue with ORS. PSE should have told you if the copy has completed in the backend and its safe to terminate the ORS.
regards,
Saurabh