Hi all hope someone can help, having an issue where SMB stops responding. This can easily be seen on a VM which uses network drives as a seed location. It will be seeding at full speed then simply halt seeding. This also co-insides with the network shares not responding on other machines and files playing in kodi stopping. This lasts a few seconds to a few minutes then everything resumes.

Ive read about some issue's being file system and XFS not supposedly having this issue however all my disks are using XFS.

 

Also attached the diagnostics thanks.

 

i5-3570K

32 GB Ram

14 3TB drives (2 Parity)

500GB SSD cache

Couple of unassigned drives 1TB x2. 

 

pluto-diagnostics-20180521-1702.zip

I see no answers to this but I experience the same problem. If copying files from my windows desktop to the array and I simultaneous want to open a text file or something, samba becomes unresponsive. The copy task just grinds to a halt and the network share freezes with "not responding". After a while, the file opens and the copy task starts doing its thing again. This has happened to me several times. The web gui also responds very slowly when this happens. Like it can't cope with two requests at the same time. CPU usage is OK when it happens. 

Edited January 18, 20197 yr by Niklas

I wonder if this could be due to some disks needing to be spun up, and using a controller which freezes access to other disks managed by the controller until the spin up completes?

I only have two disks with data. The third disk is sleeping all the time because not needed. Spin up takes like 1-2s. The tasks freezes for at least 1 minute or more. Even when the disks are spinning. 

Hi,

 

For me I found the issue seemed to be the plugin Dynamix Cache Dirs once i removed that i was ok.

I had that plug in installed but not activated. Removed it now. Will check to see if it made any difference. 

 

My server stops responding when transferring big files using samba. Made a test 10gb file and transferred it while streaming. The stream stops until the transfer is done. The stream was 720p so not much data. Strange.. 

Is the share that you were saving the test file to being cached?  The fact that the file is being encrypted (the read file being de-encrypted) may be a factor.  What else was going (Docker wise) on at the time(s) when the stalling occurs?  Have you checked to see what the CPU loads look like on the Dashboard?

12 minutes ago, Frank1940 said:

Is the share that you were saving the test file to being cached?  The fact that the file is being encrypted (the read file being de-encrypted) may be a factor.  What else was going (Docker wise) on at the time(s) when the stalling occurs?  Have you checked to see what the CPU loads look like on the Dashboard?

No cache. The dockers did nothing special. Streaming from the Emby docker. I checked CPU load using the dashboard and htop in terminal. No excessive use. I don't think it is bandwidth problem with the hdds but maybe? 

I have the array and parity connected to LSI controller and the cache and one spinner for temp data connected to the motherboard. I'm at 6.7 rc1 now and need to test it again. New samba version. 

Edited January 23, 20197 yr by Niklas

Try writing the Test file to a share that is cached.  Basically, what you need to do is to find a set of specific conditions where the problem always happens and is repeatable.  Then make a post in the 'Bug Report' section.  Remember, this does not appear to be a wide spread problem (a lot of  folks are using Unraid in a streamer environment...) and may well be a related to a specific combination of factors and circumstances.