Change disk in Predictive failure on Proliant 360G7

[Andrea811]

Hi, sorry for my bad english.
a physical disk of my Proliant 360G7 is in the state of "predictive failure" (honestly I do not know how long it is in this state).

The logical drive has two disks configured in raid 1 with P410i RAID controller.
The O.S. is Ubuntu server 16.04.2 LTS. In linux i don't have a graphic ACU, I discovered the problem using hpacucli utility with this command:

hpacucli ctrl all show config

I have read various theories about making a hot swap of a disk in the state of predictive failure state, but my problem is that I should replace the disk with a disk coming from another server where it was configured in another array.

if I pull out the broken disk and insert another one with foregin configuration, the raid controller automatically clear the foregin configuration and rebuild the disk? if to solve the problem I have to restart the server and enter the raid configuration utility, what are the steps to follow to cancel the foreign configuration and start the reconstruction? Thanks!

[Andrea811]

I forgot to say that the fault led of the disk is off. looking at the front of the disk, everything seems to work correctly.

[DANDKS]

Greetings.

It is not recommended to use a hard drive from another server which is part of an Array on the other server. Before replacing such foriegn drive, the Meta Data on that particular drive needs to be Erased.

While replacing a failed drive, make sure the replacement drive is same as the failed drive. Replacing with different size, part number or any other specification, will cause issues in the performance.

To Erase the Meta Data on the Hard Drive, please refer to the commands from the link shared below.

https://support.hpe.com/hpsc/doc/public/display?docId=emr_na-c03493210

You can also use the commands to verify the drive status or to confirm if the drive is failed or OK

Thank you

I work for HPE

[PDP-Fan]

From my own experience it is not a problem to take a replacement disk from another server as long as you insert it while the system is running. The disk will be overwritten with the existing configuration and starts rebuilding the RAID1.

The problem is that when you insert a used drive while the server is powered off, the Smart Array controller will find two different configurations on the two disks and might get confused which one to take. Sometimes it even mixes the two configs, resulting in some strange problems. But you should change disks on a running system only... anyway.

About the error LED on the Disk.... I have seen many disks with predictive failures and no yellow LED. You only see it in the Storage Administrator. I assume this has to do with firmware of the disk itself. Sometimes it takes a long time until the LED goes yellow. Sometimes it needs a reboot and sometimes it does not go on at all.

Something about different disks. You can always take a disk of bigger size to replace a smaller one. The additional storage will not be used because it is limited by the smaller drive, but it works. HP themself have shipped 146GB disk drives when they had no more 72GB in stock. Some were even relabeled to 72GB even with a 146GB inside. No worries about that.

Performance problems when mixing disks? None of our customers has ever complained after a disk change. If there are performance drops then only in a very small range. More important is that your cache (array accelerator) and battery (capacitor) is OK. THAT would make a huge difference if not working.