Problem with moving files on cachedrive

[b0mb]

Hi!

 

I´m running latest unRAID 6.4 with latest emby docker container. Filesystem used is btrfs.

 

With emby i use an so called autoorganize plugin to sort and organize new series and seasons.

 

I have an option in the plugin to copy or move the files beeing processed.

 

I prefer move because new series allways have been processed very fast because anything was done on my cachedrive (samsung ssd) within seconds.

 

A few weeks ago the files haven´t been moved anymore but copied instead. So organizing new files is taking ages now and there is a lot of traffic going on my cachedrive.

 

First i´ve thought that might be a bug of emby or unraid but the problem is still present. Today i´ve opened a thread in emby forum where an admin said that they did not change anything on the plugin and that it´s moving files fine on other systenms.

 

this is the code of the plugin - you might have a look

 

So can anybody help me with that problem? Maybe there has something changed in unRAID system especially handling files processed on cache drive.

 

Btw. when i move files within /mnt/user/ using Midnight Commander they wil be moved and not copied.

 

Thx for help in advance! [emoji6]

 

b0mb

 

 

 

Edited January 26, 2018 by b0mb

[b0mb]

Ok.... this might be an unRAID problem...

 

i´ve just tried filebot container and it acts exactly like emby when moving files within the cache disk....

 

it´s acting like moving the files from drive 1 to drive 5 instead of fast moving like the data is stored on the same disk.... so it´s more or less copying and not moving....

 

i would appreciate any kind of help

[Squid]

If you're moving from one mount point to another mount point (ie: from one host path to a second host path) then what you're seeing is both expected, and correct.

 

If your mappings on the container are something like

 

/downloads mapped to /mnt/user/downloads and

/movies mapped to /mnt/user/movies

 

And then you move a file from /downloads to /movies, the system (actually every OS ever made) will always do a copy / delete operation because the source and destination are not contained within the same mount point (regardless of the fact that they both exist within /mnt/user, and could in fact both reside on the cache drive.

 

MC wouldn't suffer from this, because you're moving from /mnt/user/downloads to /mnt/user/movies which are both within the same mount point (/mnt/user), hence the operation is basically a rename

 

If this is really annoying to you, your alternative is to pass /unRAID mapped to /mnt/user, and specify source and destination within filebot based upon paths contained within /unRAID

 

[b0mb]

Thanks for explaining.

 

What's confusing me it's that this kind of handling the files was working without a problem and then someone in December it suddenly stopped.

 

Would this be the right way to solve my problem?

 

Host = /mnt/user/Downloads

 

Container = /Downloads

 

It might be possible that I've set the mountpoints for the container before I've set up my unraid container from scratch in December....

 

 

Gesendet von meinem Redmi Note 3 mit Tapatalk

 

 

 

 

[Squid]

7 hours ago, b0mb said:

Would this be the right way to solve my problem?

 

Host = /mnt/user/Downloads

 

Container = /Downloads

So long as all moves happen within /Downloads, then yes.  If a move has to go from /Downloads to somewhere else, then no.

 

/unRAID -> /mnt/user  And then set the container up to reference everything within /unRAID

[b0mb]

i´ve tried several when not all kinds of mounting my shares into the container but none worked...

 

problems started december 2017, before i´ve never had that problem

 

crazy

[b0mb]

Meanwhile I've found out that this seems to be a problem that belongs to smb handling of the official docker container.

Switching to linuxserver.io container solved the problem

Gesendet von meinem Redmi Note 3 mit Tapatalk