Jump to content

Drive 'red dots' but has nothing wrong with it?


Recommended Posts

Okay this is updated from here,

 

http://lime-technology.com/forum/index.php?action=post;topic=32544.0;last_msg=299067

 

since I did not want to continue on a thread that was a different topic.

I have run a smart test, short test and extended test and the drive has passed all.

However, the unRAID array sets it as a disabled drive.

From the other thread (Ref link above) I checked all connections and cables, as well as using a different cage to test and make sure it was not a backplane issue.

I would call the drive capable and useable but unRAID still does not like it.

 

Are there any ideas, I am going to put the drive back in and test it via the unMenu  (smart, etc) and I will post all logs shortly.

 

Any help I can get is greatly appreciated,

 

Thanks,

 

Ice

Link to comment

The drive looks fine in the SMART data.

 

I'd put it back in the system, double-check that it's firmly seated in the cage and that the cables to the cage are firmly seated (unplug and replug them at both ends) ... and then let UnRAID rebuild the drive on itself.  You'll need to unassign it;  Start the array so UnRAID shows a "missing drive"; then Stop the array; assign the drive; and restart it to force that rebuild.

 

If that doesn't work, do it again -- but put the drive in a different slot in your drive cages (I know you have several unused ones).

 

Link to comment

That's what I had already done (referenced in the referred thread above), I should probably quote that in this thread huh?...lol.

 

But I will try again later tonight and report back

 

Referenced thread from above:

I had thought about doing that, but for now I just had one of the drives go red last night, this morning I checked all connections, and even moved the drive to check the backplane itself, I am now testing the drive on another system with L=WD DLGDIAG to see if it is actually the drive.

1. Smart test - OK

2 - Quick Test - OK

2 Extended test - Running (3 hours to go)

 

I am hoping this will be an isolated incident and not a 'shape of things to come' since I 'upset the apple cart' as it were.

 

I have just never heard of changes causing an issue such as this (of course unless there is a spike or something of that nature) this as I am fairly well versed in the anti magnetic and careful handling of these devices and nothing was jostled or overheated.

 

Any ideas off-hand would be greatly appreciated

 

I will post back when it is finished, I will also pop it back in, reboot and post log (i know I forgot to save the log when it happened, shame on me).

Link to comment

The 'problem child' drive is rebuilding right now, in order for unraid to allow me to unassign it and then reassign it, took 3 attempts in different slots.

I am just assuming that this, for some odd reason, was just very stubborn since I placed two other random precleared drives in those other slots and they were assignable right away.

Once this is done, I am going to attempt to move this drive back to it's original position since the backplane (cage) and breakout cables all tested okay.

I am hoping that I will not be able to reproduce the errors  and then I will chalk it up to a random error, but I will record and keep all logs and post here just in case.

 

Thanks,

Ice

Link to comment

Okay, It got done and then I slipped it back into its old slot and it error'd again and this time I grabbed the log, (posted below), but if I had to guess it is a bad slot on the backplane. I will be calling them to try to replace it at this point. Here is the log to confirm (I shortened parts of the log to make it under 175k)

ZIP it all, if you've still got the syslog, there's still a chance someone will spot something.

 

RMA the backplane is the step I'd advise...but that's just because about now, I'd be screaming at it and pounding on it with a sledge hammer. If you do that, I doubt they'd accept the return.  ::)

Link to comment

Actually that is all I have and it is rebuilding (again, sigh) now. It wouldn't allow me to upload the zip either, Next time if need be I will just load everything to my site and just link to it from there  :).

However i did make sure that even though it seemed like it was seeking different commands to retry accessing the disc, most if not all were exact duplicates.

 

I'm thinking after this that I may move everything off the drive, remove the drive from the array, preclear it and assign it as a new drive. Just to be safe and sure.

 

Oh, and btw Dale, I serious had already thought about the hammer  :)

 

Link to comment

As I had previously tried swapping out all cables as well as a new slot in another cage, I will eliminate that route.

Once this is done, I am going to attempt to move this drive back to it's original position since the backplane (cage) and breakout cables all tested okay.

 

The drive has now been rebuilt and is running fine in one of the other cages, the parity-check is now running (though slower than the last time) and I have removed the backplane that was giving me the issue in lieu of submitting it for rma.

 

I am just keeping my fingers crossed as this is the second time getting to this point, if I am right, this will be the last time I will have to do this

Link to comment

Whoa.  What a difference.  I'd remove the troublesome drive; do a New Config without it; and then do a parity sync (followed by a parity check) and see what kind of speeds you get.    If the slow speeds are only with this drive, then it may in fact have issues that have been the cause of your problems all along.

 

Link to comment

I have stopped the parity check and I am dumping the files off it as I post this.

 

But if it is indeed the drive, wouldn't it show what the issue is in the log? Because if it shows nothing then I am screwed out of a drive since it passed all of the WD diag tests with flying colors.

The last time I returned a drive to WD there was a poor performance reason.  So I don't think you would have a problem RMA'ing it if the drive is still under warranty.
Link to comment

WD is fairly good at accepting returns for any valid reason -- and poor performance is one of those.    When you create an RMA "fails diagnostics" is only ONE of the choices ... clearly not the one that applies here; but poor performance is one of the choices (I don't recall exactly how it's worded, but it won't be a problem).

 

First, however, do what I suggested so you can confirm the issue is indeed this drive (circumstantially it seems likely ... but it's best to confirm that).

 

 

Link to comment

I actually had to completely rebuild the parity which ended up running at an average of 52mb/s, which is well below what I expected it to run at but it seems to be acceptable for the most part.

 

I am preclearing the problem drive at this time to see if it makes a difference in performance, and I will use it just to confirm whether it was the backplane or the drive or maybe both that was causing the issues.

 

Link to comment

I have no idea what is going one here but transfer speeds, since installing the supermicro card and changing backplanes have dropped from an average of 30-45mb/s down to 16-24mb/s

 

So to be honest almost everything I do with this now has been cut in half. I am posting another log so hopefully it will show something.

 

My configuration is now this:

 

SUPERMICRO 8-Port Controller Card (new)

Rosewill RC-211 2-port Sata card

1 x WD30EURS-63SPKYO 3TB (PARITY)

2 X WD20EURX-63T0FY0 2TB (array) (new)

4 X WD20EURS-63S48Y0 2TB (ARRAY)

1 X WD20EVDS-63STB0 2TB (ARRAY)

1 X CRUZER FIT 8GB USB (OS)

2 X iStarUSA BPN-DE350SS Cages (new)

 

Thanks again for any help

 

Ice

syslog-2014-04-02.txt

Link to comment

Update

 

The problem hard drive precleared with no issues, however it took about 3 hours longer to perform this than the other 2tb I just installed. Could this be sata 3gb/s vs 6gb/s  times, even though it is on a 3gb/s motherboard controler?

 

I just placed the drive in the array, formatted and it is now in the array and I am running a parity check, however it is also running slow (30-40mb/s) is there a chance that this is only one of two slow performing drives and how would I tell which ones? Can I shut down the array and pull the drives one at a time and test them on one of my windows machines without a chance at destroying the data?

 

Thanks,Ice

Link to comment

I don't know if anyone else has done this and I would love some feedback.

I am planning on using a utility called 'Parkdale' to take each drive out of the array (yes, while it is offline..lol) and using it to test each drive on one of my other systems (windows 7).

Has anyone any experience with something like this or know if this will even work, considering the RFS vs ntfs?

Or even if there is another utility I can use this way to test the read/write speeds on the drives in this array, the utility in unRAID will not do it for me as it does not post that type of info. I would like to get this done so I can fix this performance issue problem.

 

Thanks for all the help,

 

Ice (Rick H)

Link to comment
  • 1 month later...

Well I am back and I hate to necro a thread (especially my own) but I would like to know how to test drive performance on my unRaid box drives while they are in the server.

Is there a specific utility I can use to get accurate numbers on them for read/write so I can compare them to the numbers I get when i put them in my other system (this way I can test the breakout cable performance in case that is an issue here. As of right now I am getting abysmal writes (4.5mb/s) and am at wits end.

 

I am posting a new thread with the transfer speed issue as the title, please mark this as solved

 

 

Link to comment

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...