G4 - RAID Rebuild Failure

Hey all
Just chasing some ideas. I've got a XServe G4 which was setup with a software RAID although before my time.
It is running 10.4.8 Server. We had one of the drives fail, reported failed also with SMART reporting a failure. We have now replaced the drive contained in the ADM with a Western Digital Caviar SE 320GB. The computer recognised the RAID and I have attempted to rebuild it. It attempts but Disk Utility reports that it fails.
Under DiskUtility.log it displays:
+Rebuilding RAID “WebServer”+
+Rebuilding RAID slice “disk0s3”+
+RAID Set rebuilt successfully+
Even though it lists it as successful Disk Utility reports it Failed and the RAID is still Degraded. Any input is greatly appreciated.

Hi David Moyle-
Was the drive that you used as a replacement an Apple-branded drive and the same size?
XServes have been known to have issues with drives that are not original ADMs as supplied by Apple. This could be one of those instances.
Luck-
-DP

Similar Messages

Raid rebuild times

To hold my libary material and keep a backup of current projects I have a Drobo Pro, 8 Hdd storage unit. It's not fast but very flexible in the way it uses the hdds. I have it set up to cope with two drive failures. I was backing up over night when one drive failed, this unfortunately made the unit go off line (iscsi connection). The only way to get the Drobo online again is to reboot it but it is now in the raid rebuilding mode and can't be reset. Talking to Drobo this rebuild can take between 24 and 48 hours per TB, I have 15 TB of material so we are looking at up to a month before I can reboot. Is this a normal amount of time for a raid rebuild to take. I know the Drobo is unusual in the way it uses Raid 5, they call it beyond raid, but a month without my storage seems ridiculous.

Richard,
Your situation sounds frustrating and the Drobo Pro sounds slooooowwww!
The RAID 5 build times using an Areca ARC-1880ix-12 card 10x3TB was 7 hours + 35 minutes, or about 15 minutes/TB; quite a bit faster than 24 to 48 hrs/TB. Areca and their competitors also allow for the RAID to continued to be used during operation (reads/writes) although at a reduced performance speed.
I know it would be really painful to cut over to a controller card backup system with hot-swap drives or an external cable connected box (Sans Digital, etc.), but you may want to start dreaming / planning in that direction.
Sorry for your pain on this one.
Regards,
Jim

B-Series Disk details RAID rebuild issues after disk removal

Hi,
after removing and reinserting one disk from a RAID-1 on our B-Series blades for testing, the disk details for the reinserted disk will not be fully shown.
Only the disk itself gets detected and "Presence" shows "equipped" (see attached screenshot).
It also doen not look like the RAID would be rebuilt.
So does anyone know how to:
* Get all disk details again.
* (Automatically) reinitialize the RAID rebuild.
* Monitor the RAID status.
* Be notified of a RAID failure.
Ralph

Hi,
not very happy with the fact that UCSM doesn't display a disk or RAID status.
Any chance that we might see that in an upcoming release?
We are using ESX 4.0 U2, RHEL 6 and W2K8 R2.
Hope that MegaRAID tools would work and can be queried by SNMP (or so).
I found out how to get the disk data again and get rid of the following warning that obviously comes with not getting the disk details:
Description:
* Server sys/chassis-1/blade-6 does not fullfill Service profile vmw66 due to incompatible-number-of-local-disks
* ID: 828447
* Cause: server-failed
* Code: F0337
Solution -> re-aknowledge the server...
Any chance that this might be a bug (in 1.4.1i)?
edit: Seems that has been around for a while:
-> http://terenceluk.blogspot.com/2010/12/fixing-cisco-ucs-b-series-raid-1-mirror.html
Ralph

No response from Oracle while RAID rebuilding

Hello,
At system startup the server started to rebuild the RAID ( RAID 10, hardware controller) on which the database is installed.
During this perriod the system responded to ping, shared directory where available but Oracle didn't responded.
Any attempt to connect to Oracle ( sqlplus or application) has generated "ORA-12203: TNS:unable to connect to destination".
After the RAID rebuilding was finished the Oracle become "alive".
Can you please tell me if this is a normal behaviour?
Thank you!

Thank you for your answer. I will chek the file alert.log
Server seems to be OK. I could transfer files from server ( by Remote Desktop or shared directory). I have a degraded speed but it worked. The database didn't respondend not even I tryed to use sqlplus & OEM from the server.

IX4-300D - Raid rebuild stuck?

Hi,
When I arrived this morning at work I found that my NAS (IX4-300D) was doing a RAID Rebuild
I can ping the NAS but I can't access via web interface, storage manager or SSH.
I tried pressing the "Display control buttons" to try to see some info, but i'm stuck at the massage "Support access is enabled" since I add SSH enabled, the dipslay only shows that message.
The problem is that I don't now if the raid rebuid is working, if its going to take long, or if its simply stuck.
Can I hard reboot ths NAS? Should I wait longer?
The rebuild is going for more then 1h30min but I can't be shure because it may as well started during the night.
I did not received any drive error massage.
The NAS has 4 drives with RAID 5 setup.
Can you please give me some light of what to do next?
UPDATE: After +1h or so the Disk Activity led stopped (it was flashing before) but the System Status led is still flashing.
Thank you

Thanks for the response.
I've been exchaging email with someone of the suporte team.
I hade to turn off the NAS, after booting everything seams OK, although I havent check all the data.
The suport asked me to send them a dump files, but i'm not been able to get it, its takes a long time and i've lost the session a couple of times.
I tried to keep the session opened by using the web interface from time to time, and after 2hours it was still processing the file, had to left work so don't know if in the meanwhile the file was created, but i dont think so.
I've sent them another email, but since its the weekend I will probably only get an answer next monday.
Meawhile if I can get this questions answered it woul be great:
I notice that the system only has 42mb on "/" could it be a problem?
Where does the dump file goes? Can it be genarated by terminal?
Thanks

Working on raid while raid rebuild process ( Raid 5 configration)

I am using xserve raid (4 no.) with Xsan( for FCP-Editing ) and raid for data is configured for *raid 5* and for meta *raid 1*.
In one of my Raid (raid 5 configured) Disk 9 was showing offline and which I have swapped with new disk, after that raid rebuild process got started .
In this situation (while rebuilding ) can I do work ( access volume for Ingest and editing) or is there is any performance issue will come ?.
or we should start working after rebuild process got completed. Plz suggest ?

It's fine to work with the volume while the array is getting rebuilt. Just be aware that performance may be degraded - but it was already degraded when the first disk failed, so it shouldn't be too bad.

Rebuild a drive in RAID after failure

I work at the University of Kansas, and after Sunday's storm, the campus was completely thrashed, including one of my servers. It's a G4 with 10.3.9 running on two hard drives in a mirrored RAID. Somehow during the storm, one of the drives failed. How do I go about re-initializing the drive and getting it back to into the RAID? I'll try it using the existing drive, but I may have to swap out the bad drive. Thankfully, I haven't had to do this before, but I want to make sure I don't mess anything up when I do it.
Any words of advice on this? I'm hoping it will sort of take care of things as soon as I choose to initialize the replacement drive.

TommyReynolds wrote:
Before you use BTRFS, be sure you have found a file system checker fsck(8) for it...Oracle Linux 6 ships with a btrfsck that has the repair option. If you're not using Oracle Linux, the latest version of btrfs-progs also contains btrfsck with repair, so check your distro to ensure you're running the latest version.

Mirrorred RAID Rebuild Failing

Hello,
I had a mirrored raid setup on 10.6.7, everything was working great for the last month, then today I went into disk utility and it said one drive failed. After checking the drive, there didn't seem to be any problem with its operation, so I can only assume it was a software/communication failure.
Clicking the rebuild button resulted in this error:
Rebuilding RAID set failed with the error:
POSIX reports: The operation couldn't be completed. Operation not permitted.
So I tried removing a disk, that also failed with the same error.
I then erased the failed drive which automatically removes it from the RAID configuration. Now when I try to add it back to the configuration I get the exact same error.
Has any experienced this issue and have any suggestions on how to get RAID back online?
Thanks in advance!
Michael

Thanks for your reply.
Yes, I am sure that the drives are working. I do not know what caused one of the drives to 'fail' as far as disk utility was concerned - but it was online when I checked it again.
I had issues while originally creating the RAID configuration, similar POSIX issues.
I have managed to get the configuration back online using the SoftRAID, but I am unable to create any new RAID configurations or modify existing ones using Disk Utility. Is it possible its an issue with POSIX and the drive USB drivers? Though I have tried with 3 or 4 different brand drives.
SoftRAID allows the drive set to be converted to a Disk Utility set after creating/repairing it - so thats what I have done for now.
I will try to setup/repair this all again once I have Lion installed and see if they fixed it in that version.
Any suggestions on how to get Disk Utility's RAID configuration on a Intel Core 2 Duo Mac Mini would be appreciated. I have filed a bug report with Apple as well. SoftRAID was a suggested alternative to a few other posts on the discussions for similar issues.
Michael

NSS4000 RAID rebuild time

Yesterday I setup our new NSS4000 - I updated the firmware to 1.16-3 (Mon, 08 Jun 2009 14:12:52 +0000) - I created a RAID 5 array with four Western Digital WD15EADS 1.5TB drives - after the array completed one of the drives was marked as failed - I removed and reinserted it and added it back to the array and it failed again a few min later - I replaced the drive and added the new drive to the array. This was 14:10 yesterday and the rebuild is only at 55% over 24 hours later. I don't see the WD15EADS drive on the approved list specifically but the WDS10 and WDS20 are both on there.
Is this a normal rebuild time for 1.5TB drives? I am going to delete the array and recreate it since there isn't any data on the system yet but I would like to know if there is a problem with these drives in the NSS4000 before I put it into production.
Thanks,
Tom

I have seen this behavior a couple of times (actually twice) on TB drives and creating a RAID 5. I don't think the RAID level matters much, just the fact that we are addressing a lot of space. One word of caution, drives that are not on the approved drive list should really be avoided. That was not a sales plug, or preaching Cisco literature. Granted drives that are not on the list will work and will spin and all that fun stuff, but most of the RAID failures I see are using non-approved drives. Another thing to keep in mind; is that the NSS devices can only be configured for up to 4TB. That is not to say you can not add a larger RAID. You can install 4, 2TB drives, but it does not mean you can safely create a 6TB RAID5 structure.
My guess would be that one drive is not liking or just not playing nice with the NSS and is having a hard time being accessed and addressed. Now add the fact that we are over our 4TB limit, not really surprised it is taking a long time.
My personal rule of thunb; I always create RAIDs with same drives, models and size and never mix manufacturers. That is me talking no one else, its just what I have learned over the years via bad experiences.
Hope that helps.

Has anyone experience Mac Pro Server RAID card failure?

RAID card of our company's Mac Pro server broke twice in less than 12 months, resulting in our server being out of office for more than 11 weeks for repair. I really would like to know if we are just unlucky or if this kind of hardware failure happened also to other people?
Thanks for your help.

Hi
I know it's no consolation to you but judging from my experience I think you've been unlucky.
I've installed Apple RAID Cards in numerous XServes and MacPros at many sites and apart from the rare battery failure they've all gone OK and are still working as intended AFAIK.
HTH?
Tony

RAID 1 failure of both drives

Hi
My HP m9520f was configured with 3x1TB drives, two in a RAID1 mirror and one as a spare. This had been running fine and was the boot volume for Windows Vista.
Today, the computer would not boot from power-on, and the Intel Matrix Storage Manager ROM (ctrl-I) indicated "Error Occurred(0)" for both of the RAID1 drives. It asked me if I wanted to recover and use the spare, to which I answered YES. Unfortunately, it then goes on to say that the rebuild will occur in the operating system, which is will not boot.
The following is a picture of the configuration screen as it is today... Drive 4 is the spare which was promoted to "Member Disk(0)" but contains no data as the system will not boot.
Can anyone suggest how I can recover my RAID volume and boot the computer?
Alternatively, are there any further diagnostics to indicate why the RAID volume failed completely?

Hi,
Unplug the failing hard drive and reboot. This has worked for me in the past when one hard drive of a RAID 1 volume set has failed. You better hope that both hard drives have not failed (it happens). The rebuild process should be transparent but it will take a while to complete.
HP DV9700, t9300, Nvidia 8600, 4GB, Crucial C300 128GB SSD
HP Photosmart Premium C309G, HP Photosmart 6520
HP Touchpad, HP Chromebook 11
Custom i7-4770k,Z-87, 8GB, Vertex 3 SSD, Samsung EVO SSD, Corsair HX650,GTX 760
Custom i7-4790k,Z-97, 16GB, Vertex 3 SSD, Plextor M.2 SSD, Samsung EVO SSD, Corsair HX650, GTX 660TI
Windows 7/8 UEFI/Legacy mode, MBR/GPT

RAID rebuilds whenever I restart

I have two new LaCie 500GB disks connected directly to my iMacG5 via USB (no hub). I've set them up as a mirrored RAID pair using the Apple RAID built into Disk Utility. This seems to work fine, but for some reason one of the disks gets rebuilt every time I restart my computer (I'm not sure if it's always the same one -- I've just started keeping statistics). I've tried ejecting the RAID pair after the rebuild completes and before restarting, but this doesn't seem to help. Since rebuild takes about 30 hours, this doesn't seem like a great situation. Any suggestions? Have people had better luck with SoftRAID? Thanks,
Norm
iMac G5 Mac OS X (10.4.4)

Hello
I had the same problem with my LaCie 250's...I ended up getting SoftRAID and have had no problems since. I wonder how robust the built-in Apple RAID functionality actually is... The biggest single advantage of SoftRAID (aside from it's reliability) is the ability to rebuild the RAID set in the background, allowing you to continue working.
I also think the LaCie drives are a bit sensitive when powering off. I wrote a short AppleScript, which I can't now recall, because I saved it as run only ;-(
But it was something like "Tell Finder eject RAID. Tell Finder shutdown." I keep this in the dock, and click on it whenever I want to shutdown. You can see whether this will work for you by manually ejecting/unmounting your RAID set before you restart.
Hope this helps
Paul
PS. Eagerly awaiting Intel version of SoftRAID now my new iMac is on it's way...

Raid rebuild need to relink all folders to LR 5

Since Adobe makes it impossible to talk to them and I am forced to post this question here. I had to rebuilda raid from my back drives now all the files are no longer linked and have the little question mark how can I relink all the folder at one time? I don't want to do each folder individually as I have way to many to do that. I think there is a merge function however not sure how to do it and I don't want to screw this up and have to rebuild my raid again once is enough. Thank you for your help!

The ? means something is different in the path to the images than when they were first imported. What is different: a device name or drive letter or top-level path folder name? If so, then the easiest thing is to make the drive-letter or device name or top-level folder name the same as before.
If this is not possible, then you want to relink things (via right-click Locate Missing Folder) as high as possible in the folder tree--the top level folders. If you do that than all the folders underneath will also change without having to go to them individually.

RAID Rebuild very long time / freeze?

We have a Mac Server with the following setup.
SATA 500G HDD x 1 (for Mac OS X 10.6.8)
SATA 2TB HDD x 2 (Raid Mirror)
I found one of the 2TB hdd was shown failed but it didn't show errow when I clicked it individually. So just thought it might be the writing error on RAID.
So I clicked Rebuild and then it showed message "Adding booter for RAID partition disk1s2" and the status bar indicates that it is running till around 40%. After 2 days, it is still running around 40% with the same message.
It's adnormal. What should I do now? stop it or the HDD is really failed? Many thanks.

I turn off the Server lastly and resart. It's back to original state the says one HDD failed.

Does RAID Rebuild Fix Bad Blocks on RAID Slice?

I have RAID 1 enabled (the mirroring option) between two drives on my XServe G4. Yesterday, Server Monitor reported that the RAID was degraded. I booted off the install CD. I launched Disk Utility. Disk Utility immediately rebuilt the RAID. By immediately I mean that it didn't ask if I wanted to, but it took over two hours.
Today, I ran TechTool Pro. It says that one of the disks in the RAID has 3 bad blocks. It says that the only way to fix that is to erase the disk and zero all data.
When TechTool scanned the RAID itself as a disk, however, it did not report any bad blocks. This makes me wonder if rebuilding the RAID somehow took care of the bad blocks already.
Please let me know if my theory is correct!

Oh...if you have data you are trying to save then its best to try cloning the drive to another one, which should save its data as-is to another drive. Hopefully you can do so without there being hangups around the bad blocks and keep at least most of your data intact. I'd recommend you try using Carbon Copy Cloner to do this, but you will need a secondary drive that is at least the size of the data on the problematic drive.

G4 - RAID Rebuild Failure

Similar Messages

Maybe you are looking for