I just noticed this error in the Monitor Tool "Drive Error Analysis needed".  Under the Drives tab a single error is indicated for Drive 10 of my Unity system (I have 15, 4TB configuration).  All my clients are editing OK.  What should my next course of action be?  One of my clients is currently inputting a lot of P2 media, so I can't stop the File Manager or disconnect anyone just yet.

Unity MediaNet 5.0.1, 3x XP clients on fibre channel

You need to figure out the cause of that error ASAP. it could be nothing, but it could be a failling drive.

Client importing P2 media is important but not as important compaired to lossing media due to a failing Drive ( Unless you have Morroring on "All" Parition )

Run Analyser ASAP.

You should run ASM, and do a non-destructive test.  If the error persists, replace the drive with the spare.  You may want to do the drive replacement prior to shutting down the File Manager, in case the drive can't spin back up.

Also, as a side note, you say you have 15 drives in your Drive Set.  Do you have any mirrored Workspaces?  Mirroring requires an "even" number of drives.  An "odd" number of drives can yield bogus "Drive Full" errors.

CorneScheepers:

rive 10 of my Unity system (I have 15, 4TB configuration)

What?  You mean you have 15 drives in the allocation group and only one spare in the chassis?  WRONG! Allocation groups SHOULD BE even numbers of drives for mirroring to work correctly.  This is an avid guideline.  Anyway, can't fix that now, lets get to the drive issue.

ok, to find out exactly what is going on with the drive, the first thing you should do is run disk error analizer.

go to start, programs, avid unity, disk error analyzer.  A dialog box willl open looking in the root unity folder for the unityclientslog.txt file.  Load that.  The file will load and you will see what exactly happened with the drive or drives. 

To make the DEAN lights (Disk Error Anaylsis Needed) go back to green, in the monitor tool click on reset events. 

To start a new unityclientslog.txt file, in the monitor tool, go to advacnced settings and click on start new unityclient log.  I would do this, so when you get intouch with support, you can send them the log with the errors in it.  It is saved in the root unity folder with the date of the save. 

Next, run Avid Storage Manager.  Sequential Read / Random Read + 60 minnute surface scan.  It sould take 5 to 6 hours to run.  Run it overnight.  When you get the log from it, it will say REPLACE next to a failing drive or drives.  Its probally reallocated blocks or Long Command Times.  LCT threshold for avid is 750 miliseconds.

Call avid and get a replacement.  If you have not preformed an OFFLINE drive recovery, call support and have them walk you through it when you get the replacment.  I say OFFLINE.  this is the one you want to preform, it requires all editing clients to be logged off, and the file manager stopped.  ONLINE is slow, avoid it.

Hope this helps.

Jay

Edit:   As soon as i post, Randy beats me AGAIN!  HAHAHAHAHA!!!!

"Edit:   As soon as i post, Randy beats me AGAIN!  HAHAHAHAHA!!!!"

On the bright side, it proves we're all on the same page, and the advice is consistent & reliable ... :)

 Thanks for the replies.  I mistyped 15...it's 16 drives.  My system is fully Avid compliant.

"I mistyped 15...it's 16 drives."

16 drives with 2 "hot" spares?

 Yes, I suppose so.  I've been operating the system as an ACSR set it up for me.  All my workspaces are currently mirrored.

I think I  know what caused the error now - my aging MCA system running on an xw8200 box.  I have found that this system takes a long time to access workspaces thought Explorer (I've posted about this on the MC forums somewhere) 

The error states "Long Read Command Time" and the source seems to be Edit 1 (my MCA system).  Do I need to worry and how can I remedy the situation? See my specs for details on the MCA system

 Found the post about the slow 8200 http://community.avid.com/forums/p/58747/329241.aspx#329241

"Long Read Command Time"

You really should run ASM for an extended test.  A common error is "Long Command Time" or LCT.  This indicates a failing drive.  Your xw8200 (Edit1) may also be contributing, but I'd definately want to test those drives.

Thanks Randall, I'll do so this weekend and post back if I find anything.

Wondering if you found anything and if a drive recovery was needed?

Get this, as soon as we were all talking about this stuff, Thursday of last week i had the same thing happend to me here on my lanshare ex.  DEAN light activated, so i ran ASM.  It failed one drive in the Lanshare chassis.  250gig sled, 3 RB's, 7 ERR's, 8 LCT's.  Max LCT was 1438721.  ASM stated replace.  I preformed an OFFLINE drive recovery, all files transfered, and the spare was activated.  Got a new sled from Avid yesterday, took the bad one out of the chassis, and slapped in the new one.  Bam, made it a spare and all is well. 

Just wondering how you made out, Have to love the timing on these things!

Jay

 Thanks for taking such an interest in this Jay.  Fortunately my system checked out OK.  I ran ASM as advised and all my drives were OK - no replace messages next to any of the drives.  I still suspect that the error might have been caused by my aging MCA system somehow.  Hopefully we'll be able to retire that box next year and put in something else like Nitris DX on a faster Hp box...fingers crossed. But to date I've had no further issues on the server or any of the clients. I plan to run ASM once a month now and keep monitoring the system.