MSA500 Disk LEDs + Read/Write%

[Malc_2]

Hi there, I have an MSA500 (version1) with 2x 18GB disks(Raid1-Quorum), 5x 36GB(Raid5 - Data) Disks connected to 2x DL380G3 in a Cluster configuration. The DLs have 2 internal 18GB disks (Raid1 - OS partition).
I have used the HP Maintenance CD to update the Firmwares to the latest versions. And added all supplemental patches to All Devices.

I am used to Proliants, but not so familier with the MSA or Proliants configured in a Cluster config.

We are experiencing major disk problems with data corrupting, I have placed a call with HP helpdesk who told us to upgrade firmware and we are still running tests.
But in the meantime I have a couple of questions.

-

1 - When you turn the MSA on the Disk LEDS seem to be on constantly. Once the MSA displays "Startup Complete" the disks are still active. The LEDs eventually go off but as soon as First Node is brought up the Disks LEDs are on constantly for about 5mins, including the DL internal disks. Then all goes quite again. But as soon as you run any commands all the disks go bonkers again and the LEDs are on constantly.
Is this usual for the disk activity to be so busy on an MSA500?

2 - There is a setting on the controller called the Accelerator Cache Ratio which is attached to the MSA that allows you to change the Read / Write percentage.
For an Microsoft Exchange server what are the reccomended settings for this?

Many thanks for your patience reading through all this,

Mal

[Jefferson Humber]

Malc,

Are you refering to the HDD activity LED ?

With smart array based controllers there is a background 'bad block' check that happens. This helps to predict hardware failures on HDD's upto 72hrs before they happen.

I'm not saying this is what you are witnessing, but it's worth bearing in mind that this takes place automatically by the controller.

As for your cache ratio setting I'd go for 50/50.

Jeff

[Malc_2]

Yes, thats exactly what I'm refering to. It just seems very busy.

Thanks for that info,
M