I have below issue for quite sometime.
Most the clone operation will take place in night time for backup purpose. that time max WP counts exceeding more than 50%.whenever WP cpunt reaches 50% or more, all clone is becoming very very slow..how to resolve this issue?
but last few days we noticed however Max WP counts are at 10 to 15%, but still clone is taking very long time..
can anyone throw some light on this?
in SYM/VMAX there is a stage called priority destage when frame starts to destage more aggressively and the end result is higher response time for host writes. ideally there are 2 situations which triggers this , one being system reaches 50% system WP limit or at device level reaching 100% device WP limit.
is this a single engine vmax?how much cache at this time?Do you ahve the load distributed properly across DA and physical drives?
what values see across `system prefetched tracks per second` versus `system tracks not used per second`?
what is the %hit you for the array mostly?what is the % sequential IO do you see?what %RD versus %WR?
what value do you see for `DA average LRU fall through time`?
You can activate a clone before the precopy completes if having the clone target available is more important than the potential host impact on the source devices.
Also the clone copy is usually gated by the speed of the target devices. Do you have enough disks? Are they SATA? RAID6?