Jump to content
XPEnology Community

Sparsebundle & Reliability on NAS (False Errors)


X505
 Share

Recommended Posts

Hi ~ I'm getting strange errors after upgrading my BootLoader to 5.2-5644.5 and DSM 5644-5 on my ASROCK box.

 

It was running fantastically since November (5 months uptime, not a hiccup). After upgrading from Celeron processor to i3, I decided to do the whole "bootstick & DSM" update, since I had the whole box apart anyway. After doing so, I began getting "disk unplugged" on Disks 3 & 4. These showed up as alerts saying "Volume Degraded / Crashed," which had me:

 

* Attempting to repair (Disks were visible when I first attempted repair)

* After 3 or 4 days of volume degrading, I noticed by accident, while viewing HDDs in DSM, that 3 & 4 would "disappear" and "reappear."

* Changed SATA cables

* Seemed to fix, but 2-3 days in, drives 3 & 4 popped out again (totally disappearing, causing Volume Panic)

* Turned off "Hibernation" because I read that non-enterprise drives might cause this problem.

* Changed SATA cables again, AND:

* Removing 3 & 4 and running EXTENDED S.M.A.R.T. tests on both. Reported NO Reallocated Sectors or "Awaiting Reallocation". In other words, both drives 3 & 4 appeared TOTALLY healthy.

* Zeroed both drives, reinstalled and rebuilt array , reloaded data.

Everything seemed stable, ran for 2 or 3 days, then, 2 days ago,

* "Read error at internal disk [3] sector 98649952" and "Bad sector at internal disk [3] was remapped."

 

So, when I went to HDDs in storage manager, 3 & 4 were completely gone again.

 

Scouring Xpenology in Google got me almost no clear answer, except one post where a user said they got their drives back by using an older Xpenoboot, with their theory being [might have drivers necessary for the SATA ports that are missing in the newest boot loader.]

 

Now, if all this isn't weird enough, hold on.

 

The ASROCK box is on a shelf, requiring unplugging ethernet & power. So I took it down, and put it on the workbench. When I powered it, ALL THE DRIVES appeared! For grins, I went ahead and repaired the array in the Storage Manager for the 10th time this week. And then got THIS message:

 

"The file /volume1/.../iMovies Folder.sparsebundle/bands/46d is damaged because of bad sectors."

 

Yes - I have automatic backups from the ASROCK going to the QNAP, so I wasn't too concerned. Weird part?

 

The sparse bundle OPENED JUST FINE! All movies in iMovie exported without a problem. Disk Utility found no errors on the s.bundle. I ran EXTENDED S.M.A.R.T. test on BOTH disks 3 & 4 in the Storage Manager, and, just like I expected,

 

* NO Bad Sectors reported

* NO "Pending" Sector errors reported

* DISKS reporting in the Health Info as 100% HEALTHY.

 

So, THIS time, I went ahead and made a new boot loader, with the 5565.1 Xpenoboot USB loader (Oldest one I could find). That's what the machine is running on now - This is day 2 that it's on. I'm waiting for a few more days to see if this "fixes" the problem.

 

Other than the complete pain in the *@#)@)$ from having to repair the array every 2-5 days, I am only concerned about

 

* SparseBundles getting corrupted by this P.I.T.A., and, then - Getting "backed up" to the QNAP server. The reason I have the network configured this way is because the ASROCK system gives everyone in the house very fast access to concurrently work on their projects, and the QNAP, being a bit slower, a great daily rsync backup. Before this started happening, everything was terrific.

 

The QNAP hasn't had a problem running the most recent boot loader and DSM, there have been absolutely no problems. I was thinking of reversing the order-of-operations across the network, but that would slow everything down bcz of the processors & RAM in each.

 

Other than just "waiting to see" if this problem repeats, even with the downgraded Xpenoboot test, do you have ANY idea whatsoever of what's causing this or how to correct it?

 

(I know the temptation to talk about "replacing drives" is great - But keep in mind that I've had drives go bad on the ASROCK, and the drives didn't "plug out" or "disappear" from the Storage Manager. This problem is not looking like it's drive-issues. I mean, it could be, but please, if you actually think it is, please support this theory with actual evidence / experience so posts aren't spent distracting from other software / driver related solutions).

 

Thank you!

Link to comment
Share on other sites

 Share

×
×
  • Create New...