MSA2000: how to get status of vdisks and physical disks (CLI or WEB)

[igor-k]

Hi all.
The situation is like this: I've got message from the device (MSA2000 FC) that one of physical disks is down. Hot spare had been used successfully and the process of array rebuild began. Next all the meta data from malfunction HDD was erased and it was set as a global spare. Now the questions:
1. How should I know about status of array rebuilding process? I mean its current state: completed percentage, time elapsed and so on. I haven't found any appropriate CLI command or WEB interface page.
2. I supose that HDD, which has been marked as a failed, has hardware problem. I want to get its SMART status to see reallocated units counter. How can I do it?
Thanks.

[igor-k]

I have found information about array rebuilding process. It's on the Virtual Disk's main page in Web interface. But question about SMART is still actual.

[SUBHAJIT KHANBARMAN_1]

You can verify any disks SMART status from the SMU. Please try the following,

1> Select the MSA

2> Then from Physical-> Enclosure 1 (or the particular enclosure for which you want to verify)

3> Then from Enclosure Overview, Front Graphical tab select the particular disk.

4> You can find the SMART value as 1 if it is enabled.

In order to get more information pleas refer MSA SMU reference guide.

[igor-k]

May be I have missed something, but all I can see is SMART:Enable.  Did you tell about this string? That is not enought ofcourse. I need more detailed information.

And one more question. How can I get current IOPS for particular LUN (with HP SIM, SMNP or any other way). I have installed and set up HP SIM, and it can get some information from the device, but this information is quite static: LUNs info, health status, interfaces identifiers and so on. I didn't find any performance info. Did I missed (again!!??) something? 

[igor-k]

Any help please?

[SUBHAJIT KHANBARMAN_1]

I have heard about IO Meter.............this may help you

[igor-k]

It is supposed, I should install IO Meter (or any other software of this kind) directly to the server, which has been provided with a LUN? I think this is wrong way, because it is quite possible to have dozens of LUNs, connected to different servers. I think, it becomes quite difficult to get the full picture of IO load to SAN storage in this kind of scenario.
No, really, there is no STANDART/TYPICAL/BUILT IN way to get current IO load?
I have checked supported SNMP keys and found nothing really usefull. Just a static information, as I said before.
Guys, please, share the way you have solved the performance monitoring problem. It's really actual.
Thank you in advance.