Hello,

 

It turned out to be a routine now for me to restart the server whenever I need the shares.  The shares seem to be gone after a day or so after I restarted it.

 

I believed the issue started when I had a problem with my cache drive and now I don't see the cache drive in the array anymore, which I don't mind if I have it or not.  I just need my server to be up and running.

 

After I restarted the server, I can see all of my shares under /mnt/user/ and unraid web.  When the issue persists, unraid web will not load and /mnt/user directory could not be accessed.

 

My syslog.txt file is attached and hope the community could help to resolve my issue.  Thank you in advance for your help.

 

unRAID Server Plus

version: 4.7

 

Disk devices

parity device: pci-0000:00:11.0-scsi-1:0:0:0 host2 (sdb) SAMSUNG_HD204UI_S2H7JD2ZB06754

disk1 device: pci-0000:00:11.0-scsi-2:0:0:0 host3 (sdc) SAMSUNG_HD154UI_S1XWJ1KS704715

disk2 device: pci-0000:00:11.0-scsi-3:0:0:0 host4 (sdd) SAMSUNG_HD103SI_S1VSJ9CS624046

disk3 device: pci-0000:00:11.0-scsi-0:0:0:0 host1 (sda) SAMSUNG_HD204UI_S2H7JD2ZB06755

disk4 device: pci-0000:00:14.1-ide-0:1 ide0 (hdb) SAMSUNG_HD154UI_S1XWJ1KS704654

disk5 device: pci-0000:00:14.1-ide-0:0 ide0 (hda) SAMSUNG_HD103SI_S1VSJ9CS624048

cache device: pci-0000:03:05.0-scsi-1:0:0:0 (no device)

 

 

Cheers,

syslog.txt

I see a lot of "out of memory" issue in the log.

 

How much ram do you have in your unraid machine?

 

 

Hi jangjong,

 

Thanks for looking into this.  I have 1GB of RAM in my system.

 

Thanks,

Hi jangjong,

 

Thanks for looking into this.  I have 1GB of RAM in my system.

 

Thanks,

 

Are you running any plug-ins or installed extra packages?

 

--

 

Edit:

 

Also, I see many 'ACPI Error: No handler for Region [sACS] (f741d208) [PCI_Config] (20090903/evregion-319)'

 

Do you have AHCI mode enabled in BIOS?

Are you running any plug-ins or installed extra packages?

 

Yes, I have sabnzbd, unRAID-Web, proftpd and rsync.

 

Thanks,

That is way to much plugins running on 1gig... You can protect the shares by making an OOM adjustment but something else will get sacrificed because you are running out of memory:

 

http://lime-technology.com/forum/index.php?topic=22971.0

Yea, that may be too much for 1GB.. i feel like sabnzbd alone could take up to 1GB depending on the file size it's downloading/verifying/extracting.

 

I was about to recommend Helmonder's thread as well..

 

but you need to put more ram in your machine if you want to run those plugins correctly

...

 

Are you running any plug-ins or installed extra packages?

 

--

 

Edit:

 

Also, I see many 'ACPI Error: No handler for Region [sACS] (f741d208) [PCI_Config] (20090903/evregion-319)'

 

Do you have AHCI mode enabled in BIOS?

 

 

Jangjong,

 

I think we are heading the right direction here.  I looked at the syslog and saw those ACPI Errors too, but didn't know what does it mean.  I do have AHCI mode enabled in the BIOS.  I also slow down the CPU FAN to the lowest speed in the BIOS.  Do you think this is related to my issue that I am having?

 

Notes: I have been running my server for a couple years already with the same amount of RAM and plugins and I didn't see any problems then.  I don't believe my issue is related to memory, but it is still good to look into.

 

Thanks,

 

Well.. share going down is definitely the problem with memory though based on the log

Out of memory is killing off emhttp.

 

Dec 10 05:21:45 HomeServer kernel: Out of memory: kill process 1366 (emhttp) score 412 or a child
Dec 10 05:21:45 HomeServer kernel: Killed process 1366 (emhttp)

 

If you haven't had a problem for years though, first I would try memtest to make sure your ram is okay.

 

It could also be that newer version of those plugins and unraid may require more ram

and as you get more files in your server, i am guessing that it will require more ram to handle those data (just my guess, someone can tell me if i'm wrong)

 

I heard that ACPI errors can be ignored somewhere. but i dont know...

 

i would definitely run memtest and looking into getting more ram

Well.. share going down is definitely the problem with memory though based on the log

Out of memory is killing off emhttp.

 

Dec 10 05:21:45 HomeServer kernel: Out of memory: kill process 1366 (emhttp) score 412 or a child
Dec 10 05:21:45 HomeServer kernel: Killed process 1366 (emhttp)

 

If you haven't had a problem for years though, first I would try memtest to make sure your ram is okay.

 

It could also be that newer version of those plugins and unraid may require more ram

and as you get more files in your server, i am guessing that it will require more ram to handle those data (just my guess, someone can tell me if i'm wrong)

 

I heard that ACPI errors can be ignored somewhere. but i dont know...

 

i would definitely run memtest and looking into getting more ram

 

OK, I will do a memtest and post the result.

 

Thanks,

 

emhttp killing will make your unraid webinterface unreachable, it will not make your shares dissapear, for that to happen you would have to have smbd killed off (but emhttp and smbd are the most likely candidates to get killed so this is most likely also happening to smbd.

 

I would advice setting an adjusted OOM score (as in the other thread) for both emhttp and smbd. I would also advice to create an opposite OOM adjustment for SABNZBD, this will make sure that IT will get killed if needed (and you can restart it if needed witout rebooting the array). Info on this is in the other thread.

well.. emhttp was just an example.

what's killing the share is probably this line

 

Dec 10 05:21:48 HomeServer kernel: Out of memory: kill process 1481 (shfs) score 356 or a child
Dec 10 05:21:48 HomeServer kernel: Killed process 1481 (shfs)
Dec 10 05:21:48 HomeServer kernel: shfs invoked oom-killer: gfp_mask=0x201da, order=0, oom_adj=0
Dec 10 05:21:48 HomeServer kernel: Pid: 2976, comm: shfs Not tainted 2.6.32.9-unRAID #8

 

in your log file, if you search for 'invoked oom-killer', you will find out what triggered to run oom-killer.

 

it seems like shfs and rsync is invoking oom-killer a lot in your log file

 

so.. again my guess may be right where your file share is getting too big to handle..

 

Another possibility is..

You could take a look at this post for an example:

http://lime-technology.com/forum/index.php?topic=15596.0

 

This user changed some setting in sabnzbd accidently to use ram disk to unrar stuff, which will destroy free memory.

 

Just something to look at..

Jangjong, Helmonder

 

You guys are my heroes 

 

I dig into the rsync and realized I had setup rsync crontab job everyday at 5:20AM.  I will get rid of this rsync and hope this will clear up my issue.  Once again, thank you for your help.  I love this forum 

 

P.S. Did the memtest and the 1st pass had no issue.

 

Cheers,