Path Failed-over timne for SFW

[pgm10s]

Hello

     - I implement the SAN LUNs under the VXVM control (DSM DMP)on MS Winodws 2012 using 2 paths for each LUNs.

While I test copy the data, I unplug one FC ( one path), the copy progress is paused about 20 sec and then continue.

I am not sure this is the normal behavior of  DSM DMP. Are there any way to spedd up the fail-over process.

     - I also test mirror the volume across 2 SAN storage. When I shutdown 1 of storage , the copy process is paused about 30-40 sec and the continue.

     - About the DCO log (FAST resync) of mirror log. By default if I enable the Fast Resync , A DCO log is create on one of the DISK in DG.

Do we need to mirror the DCO to other LUN in DG?

Thank you.

[RiaanBadenhorst · Accepted solution]

Hi,

The failover between paths would depend on your array i.e. is it A/P A/A etc.

Dont understand how your mirror is set up so can't comment on that.

It would be good to have another plex added to the DCO volume.

[Douglas_Snyder]

Hello pgm10s,

Please start with this KB article:

https://support.symantec.com/en_US/article.TECH146325.html

You'll need to use the DMP admin guide for Windows but the basics of DMP tuning are the same.  There are polling interval settings you can change which will detect the outage quicker than the 30 to 40 seconds you are seeing.

As mentioned above you should add a volume for DCO - it should be separate from teh data volume.  You should also check to make sure your ASL is up to date.