Hi,
We have a problem on one head in a 6070 fabric metro cluster where snapmirror is slow. The maximum speed is 9.6 MB/s no matter how high it is throttled. If throttle is set lower, then the lower value is the transfer speed. It was orginally diagnosed going over the network, but snapmirror from one volume to another on the same head is also slow. Vol copy and NFS writes to volumes in the same aggregate are ten times faster. Also, snapmirror on the partner in the FMC is fast.
The aggr has 84 15k RPM disks, raid size is 21, disks are 300 GB. I have tried destroying and recreating the aggr; also tried smaller aggr; same result. There is nothing special in /etc/messages or /etc/log/snapmirror.
For the incrowd, this is netapp case 2001426595, opened 21-Apr-2010, but there is no progress there, so I ask here.
Any ideas?
Grtnx,