RAID Degradation

[LuserKid]

I have a RAID 1 on a new X-Serve G5. We recieved the system sometime in April, and already the RAID has degraded. I’m not sure what caused it, and because the company I work for didn’t feel like dropping cash on the large support contract (only the small one), Apple won’t help with the diskutil command line application, nor do they feel like explaing how to find out which drive is the active drive, beyond suggesting transferring a “Really Big File, and see which lights come on.” The next steps involved yanking the good drive, booting from the server CD, reformatting the bad drive, plugging in the good drive again, and then rebuilding the RAID Array. I really guess this would work, but I’m more wondering what the people here have to say, as apple tech support has never provided any real help. Has anyone used the apple RAID? Any reason why the status would degrade? Would I be better trying to figure out diskutil on my own, or should I get Apple to help with the GUI version?

[Anonymous]

First, I’ve seen RAID 1 get degraded when I had external drives that were set to power down when idle. Then, if the first thing that happened was a write, the driver would time-out and think the drive wasn’t available and fail the drive that spun-up last.

OK, how to figure this out. Let’s assume you only have two disks (it will make the example easier to understand). All of this is from the command line

diskutil list

You will see three disks, probably “/dev/disk0”, “/dev/disk1”, and “/dev/disk2”. One of them is not a physical device, but is made up from the other two being mirrored. That one will have the name of your disk. If that was “/dev/disk2”, run the following command:

diskutil checkRAID disk2

In “status”, you will see “degraded” and one of the two ‘slices’ (#0 or #1) will have “Failed” (or something like that) for a status code. Now, let’s say that disk0 had a status of “OK” and disk1 was the bad guy; run the following command:

diskutil repairMirror disk2 1 disk0 disk1

This says you want to repair a mirrored disk. The “virtual” disk is disk2, the drive that failed was slice #1 and you’re going to copy from disk0 (the good one) to disk1 (the failed one). This process can take hours depending on size of disk and speed and what else is going on on the system. To see how far along it is, do another “diskutil checkRaid disk2” command.

[Anonymous]

thanks for the cmd line tips…

in my case, after going through the checkRAID command, I have a degraded raid, but both disks have a status of OK. ???? anyone else seen this one?

hard to choose one to rebuild from if they both look ok.

[sylvain]

I have a similar problem but if checkRAID give me disk2 as damaged, Server Control/Check (I don’t know the english name of the app), gives me disk1…

I’m a bit confused, and the last one I repaired it with repairMirror, I lost everything and needed to format…

I’ve got a remote xserve…

[pixpixpix]

THIS IS COPIED FROM ANOTHER AREA OF THIS SITE

I am running OS 10.3.5 on a plain G5 with external FW800 LaCie twin 305GB drives as RAID 1. After a freeze I had to shut down the machine and on restart got an unrecognixzable disk in the RAID array, so shut the machine down and went to bed.

This morning it boots fine and DiskUtility shows the RAID array with both disks degraded.

But…I checked the Raid status with terminal as suggested in

http://docs.info.apple.com/article.html?artnum=106987

and it says both disks have status OK

Name: Hydra
Unique ID: Hydrab2e9e566214a11d99823000a95afef40
Type: Mirror
Status: Degraded
Device Node: disk6
————————————————————-
# Device Node Status
————————————————————-
0 disk5 OK
1 disk2 OK
————————————————————-

I also ran Disk Warrior to rebuild the directory and ran repair in Disk Utilities and they checked out OK

Is it possible there isn’t a problem? or if not how would I find a problem?

Apple notes that with the OS X Server it is possible to ignore the degraded status in some cases, implying it might be mistakenly reported

http://docs.info.apple.com/article.html?artnum=107406

anyone ever heard of this? any suggestions

thanks

ps I also note that the disk is journalled and diskutil implies the journal is overflowed

disktuil info disk6

G5:~ user$ diskutil info disk6
Device Node: /dev/disk6
Device Identifier: disk6
Mount Point: /Volumes/Hydra
Volume Name: Hydra

File System: Journaled HFS+
Journal size 32768 k at offset 0x98d000
Permissions: Disabled
Partition Type:
Bootable: Not bootable
Media Type: Generic
Protocol: FireWire

Total Size: 305.3 GB
Free Space: 112.3 GB

Read Only: No
Ejectable: No
OS 9 Drivers: No
Low Level Format: Not Supported

[Anonymous]

this line exists in an above post:

diskutil repairMirror disk2 1 disk0 disk1

Is that an extra “1” in there, between the disk2 and the disk0.

I suspect so. Freaked me out a bit, but I’m going to remove it and try it anyways.
Andrew

[Anonymous]

[QUOTE BY= small typo in diskutil line?] this line exists in an above post:

diskutil repairMirror disk2 1 disk0 disk1

Is that an extra “1” in there, between the disk2 and the disk0.

I suspect so. Freaked me out a bit, but I’m going to remove it and try it anyways.
Andrew[/QUOTE]

No, the number refers to the RAID slice that is being repaired. The first slice will be “1” as given above. If you have sliced your raid, then you may need to repair them individually.

Sean

[Anonymous]

Why don’t you all just face it, Apple is no good at the server stuff. I don’t know how they have taken such a rich and robust OS as Darwin and messed it up so bad. I have used at least 2 versions of every OS out there, and for server applications, Mac OSX Server has got to be the worst. Most things do not work the way they should and many of the ones that do will just…well…stop after a period of time, fo no rhyme or reason. And everytime I bring up one of these problems to the Apple techs (which I have paid so much extra to be able to use) They have me go through everything I tell them I have already done and then there answer is either one of 3 things
1) I am not sure, I will have to come to your site to see it. How is 2 months from now?
2) Hmmm, that’s odd. It shouldn’t do that.
3) We have a team working on that problem, check back in a few months.
Do yourselves a favor and spend some time learning Linux. From the sounds of it, some of you are halfway there to being a Linux guru just because of what you have had to do to work with these peices of Mac.

[Anonymous]

My mirroring RAID was degraded, disk utility was not abel to rebuilt the RAID, I was wondering if someone knew how to use diskutil (the terminal command, the GUI utility is limited to simple functions) to repair one of two drives in a RAID array.
After many trying I’ve delete the RAID 🙁 now I have the two slices and I’m not able to mount it

In order to repair the raid without reformatting (data is probably all still there) I’m thinking I need help in three things;
Repair a disk slice individually.
Flag disk2 as a raid slice–I don’t know how to set this property, maybe “mount”?
Create a mountpoint in /Volumes that is raid and then attach the two repaired slices to that so it appears as one logical volume.

Can any terminal master tell me how to do this?
Otherwise, is there any software that can repair a raid without reformat?

[Anonymous]

https://www.afp548.com/article.php?story=20040827122302975

[Author]

Posts