Slow cluster resource failover W23k EVA 4000 Full MPIO

[Brandon M Estes]

We have a 2 node cluster that has the following configuration:
Fabric=4 Brocade200E (A and B fabric with ISL A<>A B<>B)

EVA4000 With current VCS

Servers
DL380G5
Emulex lp1050-ex (pci-e) x2 each server
Storport Driver 5-1.30a9
Firmware 1.92a1
MPIO SQST version 2.01


Sometimes it takes 5min to fail a drive over from one node to the other

[Brandon M Estes]

In addition the windows SA's were reporting very slow format times for some of the LUNS.

I suspect that the MPIO is to blame for no good reason

[Stephen Kebbell · Accepted solution]

Hi,

there is an updated fix for storport.sys from Microsoft. Near the end of the article is mentions something about slow failover of cluster resources. Have a look:
http://support.microsoft.com/kb/932755/en-us

Regards,
Stephen

[Brandon M Estes]

Good catch but we have the required storport.sys ect...

Anyone have a link to HP's web site for Emulex driver settings?

I have

Quedepth and Quetarget set per HP's rec