BE2010R2: "Bad block" error on HD or Tape?
Hi all. Dell server (4 years old), with internal DAT72 single drive, Server 2003 R2 and BE2010R2.
Lately been getting a few things happening, one being the cleaning LED lit (which I posted about in my "library robot" thread on here a day or so ago), and the backups themselves are reporting errors. This seems to occur during the Verify operation of a full backup job. Here's the pop-up alert:
=======================================
The following event was reported on 5/13/2011 10:48:32 AM:
The device, \Device\Harddisk0, has a bad block.
--------------------
This error signifies bad blocks. This could be the result of a outdated tape device driver, faulty tapes, or dirty read/write heads on the tape drive. Update to the latest Backup Exec tape device drivers, and run a cleaning job. If possible, replace the tapes with new tapes. If the issue continues, please contact the hardware manufacturer.
=======================================
The 1-liner about Harddisk0 leads me to think BE sees a problem with the disk or array, yet the paragraph below that talks about tape-specific issues. Anyone have any idea which BE is really talking about?
Being a Dell server, I ran their DSET and diiags utils and it is not reporting any SMART or other errors on the hard drives. The array controller has a Degraded state but this is due to the battery on it.
The DAT drive is connected to SCSI (39160 I believe), not the array controller.
Comments
You are correct, that is
You are correct, that is pointing to an issue with the HDD, not the tapedrive.
Check the Eventlog (System) for errors/warnings which might give further indications where exactly the problem lies.
Furthermore perform a Full Backup of this hard disk and after that a chkdsk to check for filesystem errors.
You can find out which HDD is affected by following this article: http://support.microsoft.com/kb/159865/en-us
PMCS GmbH & Co. KG - Consulting und Support für BE/NBU/EV und andere Symantec Produkte.
Please take the time and mark this post as solution if it solved your problem - thanks!
The classic catch 22 :)
The classic catch 22. I can't perform a backup (or rather, it does run, but gives errors about corrupted data) since the HD is having problems, but I should't do any disk-risking activities without a backup. :)
ACtually just something to mention - I looked at that article and yes the info is valid, but in situations where you have a hardware-based array (RAID), the OS doesn't know how to differentiate between the hard drives. To the OS, any given raid set is prsented as a single disk so determining which specific drive is really having the problem can only be done via vendor-specific utilities unless there's a 3rd party product that's certified to do the same. HardDisk0 is correct beause this is a 2-disk RAID set. And that's where the real problem lies - windows thinks it sees bad blocks but in truth this could be problems with the array controller in it's presentation to the OS - an I/O error on the controller can be seen to the OS as a "hard disk error".
Do any of the Options or Agents in BE 2010 R2 allow for a image of the system to be craeted? I'm just wondering if there's a simple way I can just say, boot to a CD and create an image backup as a disaster recovery optionb efore I do anything to troubleshoot this. Of course this boot disc would need to be able to load server array controller drivers.
RE: The classic catch 22 :)
I'm just wondering if there's a simple way I can just say, boot to a CD and create an image backup as a disaster recovery optionb efore I do anything to troubleshoot this. Of course this boot disc would need to be able to load server array controller drivers.
Symantec would be more than happy to sell you SSR Symantec System Recovery (Formerly BackupExec System Recovery) that will do just that
The closest Backup Exec comes is IDR, which basically just automates a system restore (to like hardware) using BE data (files) and a boot disk
If this response answers your concern, please mark it as a "solution"
Symantec puzzles me sometimes
Yeah this is a topic I've thought on and never came to an understanding of. The industry is moving towards image-based backup solutions as the ideal method for fast, accurate disaster recovery and for migrations involving moving one server to another newer machine (bare metal restore). Or so the marketing material reads though I've yet to try one of these.
Yet BESR doesn't support tape drives, so as a sollution for taking your data offsite to keep it safe, BESR doesn't do the job - unless perhaps can afford the kind of high-speed upload link that would allow 100's of GB per night to some offsite location. None of my customers could come close to affording something like that. Cloud-based storage is poitnless if your upload speed is capped as is the case with any DSL or Cable (consumer or business account) that I've ever seen. So T3 it is then - how much does that cost? :) Anyway, not sure if SSR 2011 supports tape.
So to me it's like the two products should be combined, or just give the bigger product, BE, the functionality of SSR and get rid of SSR. Neither of these two products are ideal in my mind. BESR/SSR can't be used with a rotating media strategy and BE can't do optimal disaster recovery the way BESR can.
Again, I never came to an understanding why things are the way they are so I had concluded that it must be me missing something.
So can IDR effectively be a complete solution for DR on a single machine? Where I pop in the boot disc, tell it to restore from my tape, and boom my entire system is back to the identical state it was during the backup? I have yet to test IDR. Problem is I can't take a live server, down it, and hope the restore works and since it seems IDR is not desgined for restore to disimilar hardware.
seriously what am I missing? :) I'm re-puzzled after thinking about this all over again.
Would you like to reply?
Login or Register to post your comment.