Cache Drive full, Mover doesn't free up space

[daemian]

Good morning, Hoping someone can help me figure out what i am missing.

 

I have been running on unraid 6 (6.3.0 specifically at the moment) happily for a few months now. I added a cache drive when i jumped fomr 5 to 6 and it all seemed to work as expected at first. However, my cache drive is now full and running the mover is not freeing up space.

 

Here is what I am see over and over in the system log:

 

Feb 15 08:54:44 ur01 root: .d..t...... isos/
Feb 15 08:54:44 ur01 root: .d..t...... isos/cisco/
Feb 15 08:54:44 ur01 root: >f+++++++++ isos/cisco/virl.1.2.83.esxi.mf
Feb 15 08:54:44 ur01 root: rsync: rsync_xal_set: lsetxattr(""/mnt/user0/."","user.org.netatalk.supports-eas.2ZTeJA") failed: Operation not supported (95)
Feb 15 08:54:44 ur01 root: rsync: rsync_xal_set: lsetxattr(""/mnt/user0/."","user.org.netatalk.supports-eas.AqXzlB") failed: Operation not supported (95)
Feb 15 08:54:44 ur01 root: rsync error: some files/attrs were not transferred (see previous errors) (code 23) at main.c(1178) [sender=3.1.2]
Feb 15 08:54:44 ur01 move: file ./isos/cisco/virl.1.2.83.esxi.mf rsync error: 23

 

I have searched around and found a few different posts with similar issues:

 

https://lime-technology.com/forum/index.php?topic=27450.0

https://lime-technology.com/forum/index.php?topic=40510.0

 

but I really haven't been able to to translate them to a solution for me. My diagnostics are attached if that helps. and for what its worth my cache drive is an SSD formatted with btrfs, while all of my array drives are [spinning disks] formatted reiserfs. Any idea whats causing these errors?

 

Thanks!

ur01.marx.local-diagnostics-20170215-0811.zip

[JorgeB]

I believe the problem is related to this:

 

http://lime-technology.com/forum/index.php?topic=52362.msg503483#msg503483

 

But this was supposed to be fixed on the kernel included with v6.3, I would just convert all disks to xfs to avoid these and the other reiserfs issues.

[th3rdparty]

Hello, I'm also having what I believe is the same exact problem and are getting the same errors.  It all started after I upgraded to 6.3.1.  I have started another topic yesterday but got no responses so if you don't mind I will also post my diagnostics in the event our issues are related and additional information can shed light on the cause.

 

I guess in the interim, I can disable the use of cache for my media directories and read up more on the threads you already found.

 

Thanks!

 

noahsarc-diagnostics-20170215-0829.zip

[daemian]

Thanks for the reply. That does seem to be the same issues. Like johnnie said it should not be affecting us on 6.3.x according to that post. Is there any workaround?

 

I tried upgrading to 6.3.1 just to see if it made a difference. no luck - which wasn't very surprising when th3rdparty reported the same issue on 6.3.1

 

Obviously like johnnie stated converting to xfs on my array drives would solve my problems - and I may try to do that, but its going to take some time... I don't have a spare drive large enough at the moment...

[th3rdparty]

Yea, I'm in the same boat re; converting to XFS...going to take some time.

 

Right now I'm going through and deleting any files created by have AFP shares in the past.  I'm adding this to my SMB extra config

I block a few more items than this, but similar:

veto files = /._*/.DS_Store/.AppleDouble/.Trashes/.TemporaryItems/.Spotlight-V100/
delete veto files = yes

 

then using this to remove any existing files...

You could use something like

 

find /mnt/user/Sharename -name ._* -exec rm '{}' \;
find /mnt/user/Sharename -name .DS_Store -exec rm '{}' \;

 

etc.

 

Once you've converted to XFS you'll want to re-enable those files and folders in order to get the complete macOS experience!!

 

I don't know if your using a Mac but it might be related to the way Apple implements AFP.

 

Good luck and report back if you have any success!

 

Thanks!

[daemian]

Just to follow up on this. I managed to delete enough old junk I was able to shuffle my data around and free up and drive. Then I went through the conversion to XFS disk by disk. 

 

See here:

Just finished a few minutes ago. I kicked off the mover and no more errors. So fwiw the migration to XFS does solve it.

Edited March 10, 2017 by daemian

[squirrellydw]

I think I have the same issue.  How do I convert to XFS, do I need an extra drive or can I convert what I have without losing data?

 

Mar 9 20:13:38 Tower root: rsync: rsync_xal_set: lsetxattr(""/mnt/user0/."","user.org.netatalk.Metadata") failed: Operation not supported (95)
Mar 9 20:13:38 Tower root: rsync: rsync_xal_set: lsetxattr(""/mnt/user0/."","user.org.netatalk.has-Extended-Attributes") failed: Operation not supported (95)
Mar 9 20:14:08 Tower root: rsync error: some files/attrs were not transferred (see previous errors) (code 23) at main.c(1178) [sender=3.1.2]

[trurl]

1 hour ago, squirrellydw said:

How do I convert to XFS, do I need an extra drive or can I convert what I have without losing data?

See sticky at top of this subforum. TL;DR: Any drive you change the filesystem on will be formatted so you need to put its data somewhere else.

[sirhotness]

i have this same issue currently and have found that a i have 2 syslogs and one of them is 300g+ in size taking up all of the 500g cache pool space.  do i need this file? what is it, can i move it, delete it or replace it w/ the smaller one etc?  i'm new and just fixed a few other things from the other days hdds being taken offline out of nowhere and i'm now trying to make sure the cache has enough space to function.  or do  i need a larger cache pool?

 

thank you for the help in advance.

[trurl]

52 minutes ago, sirhotness said:

same issue

Often the beginning of a discussion that is very different in the details.

52 minutes ago, sirhotness said:

2 syslogs and one of them is 300g+ in size taking up all of the 500g cache pool space.

You must have setup syslog server and told it to log there since otherwise syslog is in RAM and doesn't survive reboot.

 

Attach diagnostics to your NEXT post in this thread.

[sirhotness]

sorry about that i forget the diags. just dled them and attached them for you. can i delete it?

blacksheep2-diagnostics-20220310-1333.zip

[sirhotness]

4 minutes ago, trurl said:

Often the beginning of a discussion that is very different in the details.

You must have setup syslog server and told it to log there since otherwise syslog is in RAM and doesn't survive reboot.

 

Attach diagnostics to your NEXT post in this thread.

 

and thank you for the help, i appreciate it.

[trurl]

So did you?

15 minutes ago, trurl said:

setup syslog server and told it to log there

 

10 minutes ago, sirhotness said:

can i delete it?

Yes. Is it still enabled?

 

 

[sirhotness]

10 minutes ago, trurl said:

So did you?

 

Yes. Is it still enabled?

 

 

yes its still enabled on my end.  attached a screenshot for you.