I have recently upgraded to version 6.1.4.  The cache drive does not appear to be writing to the array, I think since I upgraded.  My cache has 435 GBytes of data on it. 

 

How can I get the cache drive to manually write to the array (and should I do this)?  How do I debug the lack of writing by the cache drive to the array?

 

Thank you.

How can I get the cache drive to manually write to the array (and should I do this)?

Main, Array Operation, Move Now

  How do I debug the lack of writing by the cache drive to the array?

By looking through the syslog and/or posting your diagnostics here

 

Thank you.

 

Thank you!  I started the move process. 

 

As for the syslog, is this what I'm supposed to see when I hit "log" on the webpage for the unraid server?  When I select "log", I get another window that says "Waiting for data", but never populates. 

Thank you!  I started the move process. 

 

As for the syslog, is this what I'm supposed to see when I hit "log" on the webpage for the unraid server?  When I select "log", I get another window that says "Waiting for data", but never populates.

Known issue in 6.1.4 / 5  Update to 6.1.6 and I believe its fixed.

 

Beyond that, hitting "Log" only brings up the last 42 lines.  You really want to go to tools / diagnostics to get the everything relevant

Thank you!  I started the move process. 

 

As for the syslog, is this what I'm supposed to see when I hit "log" on the webpage for the unraid server?  When I select "log", I get another window that says "Waiting for data", but never populates.

Known issue in 6.1.4 / 5  Update to 6.1.6 and I believe its fixed.

 

Beyond that, hitting "Log" only brings up the last 42 lines.  You really want to go to tools / diagnostics to get the everything relevant

 

Great, I will upgrade to 6.1.6 as soon as possible (probably over the holidays).  I think I've attached the data diagnostics file. 

data-diagnostics-20151206-1921.zip

Are those three shares set to use the cache drive (Use cache drive: yes)?

 

I'd be surprised if a manual invocation of mover is going to work

Are those three shares set to use the cache drive (Use cache drive: yes)?

 

I'd be surprised if a manual invocation of mover is going to work

 

I have six shares, and they are all set to use the cache drive.  I can't tell if the manual invocation of the mover is working, although the amount of data on the cache drive is ostensibly slightly smaller. 

 

Is there any other setting I should be looking for? 

Have you tried looking at the contents of the cache drive to determine exactly which user Shares seem to have content that is not being move?  That might help narrow down the possible causes.  I must admit that looking at the syslog mover does not seem to be finding anything to move.

 

Have you tried anything like stopping the array and going to Settings->Global Share Settings and disabling/renabling User Shares, That would cause the configuration file to be rewritten just in case there is anything in yours upsetting mover.

Have you tried looking at the contents of the cache drive to determine exactly which user Shares seem to have content that is not being move?  That might help narrow down the possible causes.  I must admit that looking at the syslog mover does not seem to be finding anything to move.

 

Have you tried anything like stopping the array and going to Settings->Global Share Settings and disabling/renabling User Shares, That would cause the configuration file to be rewritten just in case there is anything in yours upsetting mover.

 

Thank you again.  I'll try that.  Here are the shares that aren't being moved (this is from the log after I manually enabled the mover).

 

Dec  7 03:40:01 Data logger: mover started

Dec  7 03:40:01 Data logger: skipping "Family"

Dec  7 03:40:01 Data logger: skipping "Julies"

Dec  7 03:40:01 Data logger: skipping "Recorded TV"

Dec  7 03:40:01 Data logger: mover finished

 

The other shares are empty.  So, basically every single share I use is not being moved from the cache to the array. 

 

I will try your suggestions (thank you again), but I probably can't do that for a while.  I'll report back.

 

 

Have you tried looking at the contents of the cache drive to determine exactly which user Shares seem to have content that is not being move?  That might help narrow down the possible causes.  I must admit that looking at the syslog mover does not seem to be finding anything to move.

 

Have you tried anything like stopping the array and going to Settings->Global Share Settings and disabling/renabling User Shares, That would cause the configuration file to be rewritten just in case there is anything in yours upsetting mover.

 

Thank you again.  I'll try that.  Here are the shares that aren't being moved (this is from the log after I manually enabled the mover).

 

Dec  7 03:40:01 Data logger: mover started

Dec  7 03:40:01 Data logger: skipping "Family"

Dec  7 03:40:01 Data logger: skipping "Julies"

Dec  7 03:40:01 Data logger: skipping "Recorded TV"

Dec  7 03:40:01 Data logger: mover finished

 

The other shares are empty.  So, basically every single share I use is not being moved from the cache to the array. 

 

I will try your suggestions (thank you again), but I probably can't do that for a while.  I'll report back.

I noticed that as well?

 

Do the folders corresponding to those shares occur on at least one disk in the array?  I quick way to find out is from a command line do the command

ls -l /mnt/user0

as that corresponds to the user share ignoring anything that is on the cache disk.  Note that the folder names need to exactly match the share names (including case as Linux filenames are case significant).  That will also show if permissions are correct - they should show up as user:nobody and group:users

Have you tried looking at the contents of the cache drive to determine exactly which user Shares seem to have content that is not being move?  That might help narrow down the possible causes.  I must admit that looking at the syslog mover does not seem to be finding anything to move.

 

Have you tried anything like stopping the array and going to Settings->Global Share Settings and disabling/renabling User Shares, That would cause the configuration file to be rewritten just in case there is anything in yours upsetting mover.

 

OK, I tried this, stopping the array and disabling/reenabling user shares.  While there I also toggled the cache drive.  Restarted the array, manual mover still doesn't work.  I'll try your next idea.

Do the folders corresponding to those shares occur on at least one disk in the array?  I quick way to find out is from a command line do the command

ls -l /mnt/user0

as that corresponds to the user share ignoring anything that is on the cache disk.  Note that the folder names need to exactly match the share names (including case as Linux filenames are case significant).  That will also show if permissions are correct - they should show up as user:nobody and group:users

 

This is what I get:

 

drwxrwxrwx 1 nobody users          72 Feb 12  2012 Bobs/

drwxrwxrwx 1 nobody users        784 Nov 28 03:40 Family/

drwxrwxrwx 1 nobody users          48 Jul  3 21:19 Jaclyns/

drwxrwxrwx 1 nobody users        136 Nov 27 03:40 Julies/

drwxrwxrwx 1 nobody users      12896 Dec  7 21:00 Recorded\ TV/

drwxrwxrwx 1 nobody users          48 Nov 22  2011 Stellas/

drwxrwxrwx 1 nobody users          80 Dec  1 09:19 appdata/

-rw-rw-rw- 1 nobody users 10737418240 Dec 10 06:28 docker.img

drwxrwxrwx 1 nobody users        944 Jul 18 12:46 lms/

 

For me, these titles look exactly the same as what's on the cache drive.  That is, the titles on the cache drive are the same as what's on the array. 

 

As an aside, since I did not enable disk shares when doing the toggling of the user shares, now Windows Explorer no longer sees the individual drives in the array.  I've always wanted that done but didn't know how to do that.  (Although it would be helpful to go through Windows Explorer to see the contents of the cache drive to debug this.)

 

I've already copied the files from the cache to a temporary location and will remove the cache drive and then copy them over to the array. I have to have them backed up.

 

I'm thinking that maybe I reformat the cache drive once this is done and then add it back in?  Or will that not help? 

 

I like having the cache drive, which is also my hot spare.  However, I guess it's not technically necessary.

Do the folders corresponding to those shares occur on at least one disk in the array?  I quick way to find out is from a command line do the command

ls -l /mnt/user0

as that corresponds to the user share ignoring anything that is on the cache disk.  Note that the folder names need to exactly match the share names (including case as Linux filenames are case significant).  That will also show if permissions are correct - they should show up as user:nobody and group:users

 

This is what I get:

 

drwxrwxrwx 1 nobody users          72 Feb 12  2012 Bobs/

drwxrwxrwx 1 nobody users        784 Nov 28 03:40 Family/

drwxrwxrwx 1 nobody users          48 Jul  3 21:19 Jaclyns/

drwxrwxrwx 1 nobody users        136 Nov 27 03:40 Julies/

drwxrwxrwx 1 nobody users      12896 Dec  7 21:00 Recorded\ TV/

drwxrwxrwx 1 nobody users          48 Nov 22  2011 Stellas/

drwxrwxrwx 1 nobody users          80 Dec  1 09:19 appdata/

-rw-rw-rw- 1 nobody users 10737418240 Dec 10 06:28 docker.img

drwxrwxrwx 1 nobody users        944 Jul 18 12:46 lms/

 

For me, these titles look exactly the same as what's on the cache drive.  That is, the titles on the cache drive are the same as what's on the array. 

 

As an aside, since I did not enable disk shares when doing the toggling of the user shares, now Windows Explorer no longer sees the individual drives in the array.  I've always wanted that done but didn't know how to do that.  (Although it would be helpful to go through Windows Explorer to see the contents of the cache drive to debug this.)

 

I've already copied the files from the cache to a temporary location and will remove the cache drive and then copy them over to the array. I have to have them backed up.

 

I'm thinking that maybe I reformat the cache drive once this is done and then add it back in?  Or will that not help? 

 

I like having the cache drive, which is also my hot spare.  However, I guess it's not technically necessary.

Why do you have appdata and docker.img in user0? Normally people configure things so those stay on cache.

 

*added*

 

And, docker.img could not have been put in user0 by mover because mover doesn't touch files at the top level of cache, only folders and their contents.

Why do you have appdata and docker.img in user0? Normally people configure things so those stay on cache.

 

*added*

 

And, docker.img could not have been put in user0 by mover because mover doesn't touch files at the top level of cache, only folders and their contents.

 

That's a good question, and I don't know.  I just added them in recently and followed the "normal" procedures.  Perhaps I made a mistake?

 

My problem is that I have very limited time to work on these issues.  I started my unraid server years (8+ ago), and it ran perfectly until I upgraded to version 6, a few weeks ago.  I never did anything other than upgrade to version 5 in all that time, and replace a drive or two.  Even though I have an EE degree, I don't remember how to telnet or use the command line, so I do the least I can to get everything working, and then don't touch it once it is. 

 

I took the cache offline and will continue to trouble shoot.  Perhaps the best thing is to simply install version 6 anew?  I'll have some time over the holidays. 

Why do you have appdata and docker.img in user0? Normally people configure things so those stay on cache.

 

*added*

 

And, docker.img could not have been put in user0 by mover because mover doesn't touch files at the top level of cache, only folders and their contents.

 

That's a good question, and I don't know.  I just added them in recently and followed the "normal" procedures.  Perhaps I made a mistake?

 

My problem is that I have very limited time to work on these issues.  I started my unraid server years (8+ ago), and it ran perfectly until I upgraded to version 6, a few weeks ago.  I never did anything other than upgrade to version 5 in all that time, and replace a drive or two.  Even though I have an EE degree, I don't remember how to telnet or use the command line, so I do the least I can to get everything working, and then don't touch it once it is. 

 

I took the cache offline and will continue to trouble shoot.  Perhaps the best thing is to simply install version 6 anew?  I'll have some time over the holidays.

Though many would prefer to avoid it, I actually consider starting from scratch a good idea. If you start over, that gives you a chance to remember or relearn how things work and cleans up your installation of any crud that may have accumulated.

 

The main thing you have to do correctly when starting over is assigning the parity drive correctly. You don't want unRAID writing parity to one of your data drives. Everything else can be worked through one step at a time.

I second that, and the instruction videos of Limetech can help in initial set up.