Slow SnapMirror sync... is there something that can be done?

[rancid cradle]

We have a few SQL servers running backups from inside SnapCenter... we are having the known errors of "snapshot copy not found on the SnapVault destination..." which you typically fix by increasing the timeout and retry values inside snapcenter... but it's a bit strange that it has to take over 8 hours to transfer 153GB of data between an A300 and a FAS8300 (with FlexCache in front of a lot of spinning disks)... most other snapmirror relations transfer more in less time... we have tried to move the backups to a time where no other backups are running, but it just seems to take forever... We only have 32 snapshots on this destination the destination volume has 16TB used... so not sure what to do next, other than recreating the relationship on a fresh volume? Not ideal of cause, but I bet it will work... (for awhile anyway...)

Last Transfer Size: 153.5GB Last Transfer Network Compression Ratio: 1:1 Last Transfer Duration: 8:34:6 Last Transfer From: na-iscsi:sqlclu_f02_usrdb01 Last Transfer End Timestamp: 10/14 01:09:16 Progress Last Updated: - Relationship Capability: 8.2 and above Lag Time: 16:51:13

Suggestions are welcome... 🙂

[bronze vigil]

check if the volume compression options are the same on source + destination, if not, SnapMirror has to decompress and recompress everything which slows it down

[rancid cradle]

but doesn't the snapmirror destionation inherit the source options?

hmmm aparently not? It seems that the source has "everything enabled" as it is an A300, but the destination has "sis disabled"....

so if I enable sis on the destination... do I have to run a "sis start.... -scan-old-data" ? I fear we may have some data to "troll" though.. 🙂