EVA 5000 Disk by passed

[Hoodlight]

We have had a number of these errors on disks this year.
A Physical Disk Drive was bypassed, rendering it unusable

Once I re-insert the disk it is fine and re-levels, has anyone else has this problem. We are on firmware 3.28

regards

[TTr · Accepted solution]

I get the same errors on an IBM disk array. The explanation is that physically the array knows the drive is there (electrical sensing) but when accessing the drive for data, the drive was not there.

Do these repeat on the same drive(s)? If so you should replace the drive(s).

Otherwise look into firmware upgrading and check if the newer firmware versions specifically mention this issue or anything regarding drive timeouts.

[IBaltay]

Hi,
1. it is possible that the EVA controller firmware 3110 has solved it, because there was plan to include switch to cycle HDD power for 15 seconds to reduce the unneccessary replacement.
2. the second possibility is that the disk firmware is not compatible with the rest of the disks on the loop and therefore it is being bypassed during the loop communication.

The general approach to this problem is the following:
1. power the disk down - you are doing it via the disk unplug
2. power it back again - you are doing it via the disk replug
3. replace the drive (or try to ungroup it and then remove it physicaly into another disk bay if its free to check if there is not a problem outside the disk drive

[McCready]

For this issue, if it was under warranty or contract, replace the disk. "Something" is wrong, and I would not risk keeping the disk.

However, if you are operating on the cheap, I would be sure I had double protection level (not the cheapest way of doing things) or plenty of extra space along with a not terribly busy EVA.

Your main risk is retaining faulty drives is that more than one drive in an RSS group could fail before the (parity) data can be reconstructed on other drives (can take an hour or two, or even more). Keeping faulty drives around increases the risk, especially after a power failure (unplanned or otherwise) of multiple drives failing at or near the same time.

[Timothy Cusson]

The disk drives are getting old and worn out. I never try to "revive" a failed disk.

Always check the firmware version of replacement disks. You should upgrade the firmware of the replacement disk while its in the ungrouped disk group.

If you upgrade the EVA from 3.028 to 3.110, you can do online disk firmware upgrades.

Regards,

Tim