DL360 G6 Raid 5 Drive issue

[robedj]

Hi there

I recently had a drive flag up that is has a predictive fail and showed the amber light.

I have ordered another drive ready to Hot swap. In the small amount of time since I ordered the replacement drive I have a second drive in the array also showing the amber light.

I havent replaced any drives yet.

Now, what will happen if i replace just the one drive of the 4 in the array, if there are now two with amber lights on??

Will I loose data as its a Level 5 array?

Any help would be appriciated

Regards

DR

[xmate]

Hello.

It is a really good question. First of all if it is still working - it is very good.

It is a good practice for all vendors (HP, IBM, DELL, etc.) to keep the last prefail drive working if marking it as failed bring the RAID to unusable state.

Everything is gonna be OK if you replace the first of the bad drives.

But it is strongly recomended to have a full backup of data.

As an alternative you can place a new drive as a hot-spare (if you have some free slots). So you do not have to remove any drives. And it lets the controller make appropriate steps to recover the RAID.

And the best solution is to ask HP service for an action plan.

[TVVJ]

Fault tolerance for RAID 5 is only one hard drive failure. A failure of a second drive will result in loss of array and data.

If the first hard drive is still in predictive failure mode and the second drive failed, then replace the second drive first, meaning to say replace that has been marked as failed first so that a rebuild completes. Once done, replace the first drive.

Here, it appears that both drives are in predictive failure state and have not been marked failed. Hence the first drive may be replaced.

It is recommended to run the Array Diagnostics Utility to determine if the firmware levels of the Smart Array Controller and the hard drives are up to date as older firmware version could, in some instances, flag a drive as predictive failure when they are actually not so.

Regards,

Vijayasarathy

[robedj]

Hi there

Thanks for your replies.

They are both in predictive fail mode and everything is working ok at the momement.

A assume that with the server running and in Windows, I can simply take out the first of the drives that failed, wait 30 secs and then insert the new one??

Always nervous with this sort of thing!!

[robedj]

Checked Raid controller firmware and it is at v2.74

Checked HP website and there doesnt seem to be any updates to the firmware for storage controllers.

I have already installed the new controller driver for Windows

[xmate]

Yes, just replace the first of the pre-failed drives. Then Wait until the rebuild completes. And then replace the second drive.

But be aware of the leds status. Read about them in http://bizsupport1.austin.hp.com/bc/docs/support/SupportManual/c01704762/c01704762.pdf on page 74. If the online/activity led is flashing regulary (1 Hz - 1 time per second), then do not remove a drive.

[robedj]

Thanks for that response.

New drive is installed and rebuilding as I type!!

Will let it complete and see about replacing the second.

Thanks for everyones response