Hello Joe,

I hope you can provide some advice/comments, relative to unmenu report.

I haven't turned unmenu on for quite some time, and following a parity disk upgrade in the last couple of days, I just thought to launch it and check what it says.

 

Prior to the upgrade, in the standard main unraid screen there were no parity errors. I have not checked any smart reports.

I took the parity drive out (1.5TB), and replaced it with a slightly larger (2TB).

I restarted the array and started building parity on the new drive.  Now is finished, and the main screen also shows no errors. The new parity drive is green. After I launched unmenu, on its main screen I see this message: 

 

Parity is Valid:.  Last parity check  16076  days ago .  Parity updated 16 times to address sync errors.

 

I interpret that unmenu is somehow relating the parity check to the new drive (which was never checked), while I have checked the parity no later than a month ago on the old drive. So the 2 numbers listed at the top do not make much sense to me - while I'm trying to correlate them with the stock main screen which shows no error. Can you please comment, and perhaps explain how I should interpret them.

 

Issue #2

 

Switching to myMain, where the report can be set to represent performance, smart, etc. I chose to see smart report and I see this picture:

my first reaction was to think that I somehow disturbed the cables (for those udma errors), but I'm not sure if that is the case - since there is no time stamp attached to them.

But there are also other things, and some related to rather new drives. disk 3 is an antique, disk 8 follows in age, and then the other two are much newer. my server is up 24/7, and the power cycle is an indication of how many times I've powered it off over the life of those hard drives.

 

So my predicament is that i don't know if I have lost / through corruption/ any data...  can you comment on some checks I could do to establish that? I still have the old parity drive intact - and I hope I can use it somehow to compare things - but i'm not so sure.

 

 

Finally I see the overall health stamp passed - while the other messages alarmed me. I'm not sure I can replace the newer drives while in warranty (if they still are), based on this report.. Again, please comment...

 

Thanks much for your time,

 

 

Continuation:

 

I just started a parity check with the new drive (correct flag unchecked - in unraid main) and the report became even more colored!

I'm keeping my fingers crossed!

 

 

Same story... health status is green while the Issues/Failures column abounds in  messages.

 

Not sure what to do with all these. Before seeing this report I was planning to use my old parity drive as a replacement for the oldest data drives, and consolidate 3 drives in one. I also have another spare 1.5TB in my drawer - used sporadically for backup on my personal pc - which I keep as an emergency replacement, should any drive fail. But I'm not sure I have a failure scenario yet.

 

Kindly please let me know what you think/what would you do in this situation.

 

 

This is the final state, with no errors encountered during the parity verification.

 

Btw: (as you may have also noticed) I realized that WD hard drive (with most errors)  is a refurbished/replacement I received from WD, late 2011, after having RMA-ed one which failed during the warranty period. Looking at the ages of my hard drives, and their state of health, it is quite obvious that the older ones have fared much better than the new ones.

 

Thank you.

See here: http://lime-technology.com/wiki/index.php/Troubleshooting#Resolving_a_Pending_Sector

Hello Joe,

I hope you can provide some advice/comments, relative to unmenu report.

I haven't turned unmenu on for quite some time, and following a parity disk upgrade in the last couple of days, I just thought to launch it and check what it says.

 

Prior to the upgrade, in the standard main unraid screen there were no parity errors. I have not checked any smart reports.

I took the parity drive out (1.5TB), and replaced it with a slightly larger (2TB).

I restarted the array and started building parity on the new drive.  Now is finished, and the main screen also shows no errors. The new parity drive is green. After I launched unmenu, on its main screen I see this message: 

 

Parity is Valid:.  Last parity check  16076  days ago .  Parity updated 16 times to address sync errors.

 

I interpret that unmenu is somehow relating the parity check to the new drive (which was never checked), while I have checked the parity no later than a month ago on the old drive. So the 2 numbers listed at the top do not make much sense to me - while I'm trying to correlate them with the stock main screen which shows no error. Can you please comment, and perhaps explain how I should interpret them.

you've NEVER checked parity on the new parity disk.  (You've written parity to it, but never read it back (by checking parity) to see if you can read it.

 

In older versions of unRAID, the last date of a parity check in the status returned by /proc/mdadm status was"blank" until a parity check was made.  The newer versions of unRAID return a "0" in that field.  Now, that field will contain a number > 0 once you first check parity.

The number returned will be the time of the last check represented as the number of seconds since Jan 1st, 1970.

Doing the math, subtracting the current number of seconds since 1/1/1970, the number zero will calculate to be 16076 days ago.

Issue #2

 

Switching to myMain, where the report can be set to represent performance, smart, etc. I chose to see smart report and I see this picture:

my first reaction was to think that I somehow disturbed the cables (for those udma errors), but I'm not sure if that is the case - since there is no time stamp attached to them.

The parameter in the smart report does not tell when the errors occurred.    If you see UDMA errors, then your cables are probably picking up noise from adjacent cables.

But there are also other things, and some related to rather new drives. disk 3 is an antique, disk 8 follows in age, and then the other two are much newer. my server is up 24/7, and the power cycle is an indication of how many times I've powered it off over the life of those hard drives.

 

So my predicament is that i don't know if I have lost / through corruption/ any data...  can you comment on some checks I could do to establish that?

Compare your data with your backups of that data.  There simply is no other way unless you have previously created checksumfiles so you can again run a checksum program and see if they have changed.

I still have the old parity drive intact - and I hope I can use it somehow to compare things - but i'm not so sure.

Completely useless for that purpose once you wrote anything to the array.  It does not store anything that could be used to compare to.

Finally I see the overall health stamp passed - while the other messages alarmed me. I'm not sure I can replace the newer drives while in warranty (if they still are), based on this report.. Again, please comment...

The only error that really needs addressing is disk4 with 70 sectors pending re-allocation..  Those are un-readable sectors.  they may be parts of a file, or files, or just empty space, but would prevent you from properly re-constructing ANY other drive, so they must be fixed. (array stopped, drive un-assigned, array started, array stopped, drive re-assigned, array re-started)

Thanks much for your time,

You are welcome. 

Thank you gentlemen..

I'll focus then on disk 4. I'll stop the array and will replace it with my previous parity drive, let it rebuild, and when the array is stable I'll run preclear on the trouble maker.