how to make EVA Performance Report

[JuanCG]

Hey guys, a few weeks ago I'm having performance problems in two replicated with CA EVA4100. The replication is synchronous.

I'm getting much information EVAPerf

as-cont EVAPerf 1-dur 10000>> C: \ Reports \ as.txt
cs-cont EVAPerf 1-dur 10000>> C: \ Reports \ cs.txt
EVAPerf drt dur-cont 1-10000>> C: \ Reports \ drt.txt
hc-cont EVAPerf 1-dur 10000>> C: \ Reports \ hc.txt
hps-cont EVAPerf 1-dur 10000>> C: \ Reports \ hps.txt
pd-cont EVAPerf 1-dur 10000>> C: \ Reports \ pd.txt
pda-cont EVAPerf 1-dur 10000>> C: \ Reports \ pda.txt
PDG-cont EVAPerf 1-dur 10000>> C: \ Reports \ pdg.txt
vd-cont EVAPerf 1-dur 10000>> C: \ Reports \ vd.txt
VDG-cont EVAPerf 1-dur 10000>> C: \ Reports \ vdg.txt
EVAPerf vdts-cont 1-dur 10000>> C: \ Reports \ vdts.txt
VDRL-cont EVAPerf 1-dur 10000>> C: \ Reports \ vdrl.txt
EVAPerf vdwl-cont 1-dur 10000>> C: \ Reports \ vdwl.txt

EVAPerf all-csv-cont dur 1-10000-fo C: \ Reports \ all.txt


TLViz use to format the results and put them in a evadata.mdb, that I have generated a large number of parameters that can not interpret

How I can see if the CA is working well?
How I can see if a vDisk have performance problems?
Does watching for any problems in the cabin?

thank you very much

[JuanCG]

Good again, the two EVAs, since an error occurred in the past, records these events in the logs of each of the EVAs:


A Storage System Virtual Disk you changed controller mastership.
Corrective action code: 00

Indicated Virtual Disk've transitioned to ownership by the other HSV200 controller.
Corrective action code: 00

These two events repeat every 5 minutes for months

regards

[DeafFrog]

Hi Juan ,

This may not offer most of what you wanted to know , but may be helpful.

Regards,

[Johan Guldmyr]

JuanCG - can't tell how to interpret the command sorry. HP has an internal tool that they can run on a .csv file from evaperf that generates graphs and lots of details. If you have a support contract do consider asking for this, it's quite helpful.

About the ownerships / vdisk mastership changes - this sounds like you are using round-robin without ALUA or a preferred path.
On the EVA one controller owns the LUN, if more than 2/3 of the traffic comes in via the non-owning controller, the ownership will change. If you have round-robin (for example) then the ownership could flip back and forth like this. For better performance, do investigate setting a priority/preferred path.

I wonder if these vdisk ownership changes can come as a side-effect from a not properly configured CA as well?