Pro/Forums - View Single Post

admin69 · 09-17-2008, 01:34 AM

Thanks Douglas - this is excellent news!!!

For those keeping up with my adventure here is the scenario:

My 4000(s) likely had been powered down incorrectly due to recent storms from hurricane activity. (I don't live near the coast but the remnants usually generate strong thunderstorms in my area.)

At work my 4000 (300GB Seagates in RAID5, cable select, 256MB RAM) disk check would fail at 5% and drive 2 was orphaned (it would not show the remaining three drives being in a "degraded" state or I would have replaced / rebuilt drive 2). My problems were twofold on this server, SpinRite stopped running on drive 4 (could never read the drive again) and SpinRite found numerous errors on drive 2 at the MBR (was able to fix those). This Snap was not allowing me to configure the failure notification section - it kept telling me "address could not be deleted" when I was trying to add an e-mail address [when I back-revved the OS from 4.0.855 to 4.0.830, the addresses I had set a couple years ago appeared and could be modified - very strange]. More importantly, there was an incorrect IP address set for the APC UPS used to signal the Snap server (graceful shutdown not enabled). I had an old backup of this 4000 on another (third) 4000 so I replaced the drives with new 250GB Seagates (after checking them with SpinRite), fired up S2S and waited 3+ days for the data to copy.

My home 4000 (250GB Western Digital in RAID5, cable select, 256MB RAM) has the drives Douglas performed data recovery on - it had two orphans (drive 1 and 2 or 0 and 1; depending on how you prefer to number them). The controller for those drives is not bad. SpinRite showed no problems with any of the drives in this Snap. I had no signaling UPS set up, no failure notification configured and most importantly, no backup (shame on me).

Lessons learned:

1. I purchased a couple more 4000s. After reading through numerous posts here, I will follow the lead of others and run S2S (or RSYNC, etc.) between a pair of 4000s both at work and home.

2. I will make certain the failure notifications work on all Snap servers.

3. A signaling UPS will be installed at home.

4. I think that using SpinRite prior to installing drives and as a yearly maintenance routine is a very good practice.

5. As stated earlier in this thread, Douglas is very fair and started working on this as soon as the drives arrived at his location. I urge anyone needing data recovery or hardware repair (he helped me get my 12000 working as well) to seek him out.

09-17-2008, 01:34 AM	#7
admin69 Cooling Neophyte Join Date: Jul 2005 Location: GA Posts: 29	Re: Raid Recovery Thanks Douglas - this is excellent news!!! For those keeping up with my adventure here is the scenario: My 4000(s) likely had been powered down incorrectly due to recent storms from hurricane activity. (I don't live near the coast but the remnants usually generate strong thunderstorms in my area.) At work my 4000 (300GB Seagates in RAID5, cable select, 256MB RAM) disk check would fail at 5% and drive 2 was orphaned (it would not show the remaining three drives being in a "degraded" state or I would have replaced / rebuilt drive 2). My problems were twofold on this server, SpinRite stopped running on drive 4 (could never read the drive again) and SpinRite found numerous errors on drive 2 at the MBR (was able to fix those). This Snap was not allowing me to configure the failure notification section - it kept telling me "address could not be deleted" when I was trying to add an e-mail address [when I back-revved the OS from 4.0.855 to 4.0.830, the addresses I had set a couple years ago appeared and could be modified - very strange]. More importantly, there was an incorrect IP address set for the APC UPS used to signal the Snap server (graceful shutdown not enabled). I had an old backup of this 4000 on another (third) 4000 so I replaced the drives with new 250GB Seagates (after checking them with SpinRite), fired up S2S and waited 3+ days for the data to copy. My home 4000 (250GB Western Digital in RAID5, cable select, 256MB RAM) has the drives Douglas performed data recovery on - it had two orphans (drive 1 and 2 or 0 and 1; depending on how you prefer to number them). The controller for those drives is not bad. SpinRite showed no problems with any of the drives in this Snap. I had no signaling UPS set up, no failure notification configured and most importantly, no backup (shame on me). Lessons learned: 1. I purchased a couple more 4000s. After reading through numerous posts here, I will follow the lead of others and run S2S (or RSYNC, etc.) between a pair of 4000s both at work and home. 2. I will make certain the failure notifications work on all Snap servers. 3. A signaling UPS will be installed at home. 4. I think that using SpinRite prior to installing drives and as a yearly maintenance routine is a very good practice. 5. As stated earlier in this thread, Douglas is very fair and started working on this as soon as the drives arrived at his location. I urge anyone needing data recovery or hardware repair (he helped me get my 12000 working as well) to seek him out. Last edited by admin69; 09-17-2008 at 01:46 AM.