Transport endpoint is not connected with /mnt/user

[jwmoss]

When attempting to browse to /mnt/user, I get the error here:

 

/bin/ls: cannot access 'user': Transport endpoint is not connected

I've seen this behavior before with rclone (which I was mounting a share inside of there) and I believe what may have caused this.

 

Here's what I've done to eliminate this:

 

1. Remove all rclone mount scripts that are trying to mount into /mnt/user

2. Remove rclone plugin(s).

 

I'm still having the issue. Help!

tower-diagnostics-20210210-1349.zip

[JorgeB]

Diags are after a reboot, if it happens again grab them before.

[jwmoss]

Here's a fresh copy. The error just happened once again.

tower-diagnostics-20210210-1402.zip

[JorgeB]

It's a known issue, though root cause has been somewhat elusive, usually something manipulating files directly on /mnt/user, for some user(s) it was a plugin, but most common are probably dockers, that unpack, move or delete files.

[mbc0]

I thought I had hopefully resolved this, I only get the issue if running tdarr and keep finding all my shares missing causing me to reboot.

 

I had tdarr's media folder set to /mnt/user but changed it to /mnt/user/tvshows etc I managed 36 to find the shares gone again.  

 

Can someone please take a look at my diags in case it is something else?  I really still do think it is tdarr but there are no paths set in the docker to use /mnt/user

 

 

Many Thanks

unraid1-diagnostics-20210216-2141.zip

[JorgeB]

7 hours ago, mbc0 said:

but changed it to /mnt/user/tvshows etc

That's still part of /mnt/user

[mbc0]

Ah, I see! 

 

That makes tdarr pretty much unusable then! 🥴

 

So from what I understand, until this issue is detected/fixed within unraid there is nothing I can do?

Edited February 17, 2021 by mbc0

[JorgeB]

It's an elusive issue, and not easy to reproduce, but those kind of dockers that directly manipulate files seem to be the ones that are more likely to cause this problem, but unless you can use disk shares for that (that would limit it to one disk or pool) don't known much else you can do.

[mbc0]

22 minutes ago, JorgeB said:

It's an elusive issue, and not easy to reproduce, but those kind of dockers that directly manipulate files seem to be the ones that are more likely to cause this problem, but unless you can use disk shares for that (that would limit it to one disk or pool) don't known much else you can do.

 

Thank you, I understand,

 

I am able to reproduce it several times a day at the moment if I can be of any use testing! 

 

I will install tdarr on my other unraid server to see if I get the same results as it seems I am the only tdarr user struggling with this problem.  

 

Just to make sure you are clear, the way tdarr works is it copies the file from the user share to the cache, it then encodes that copied file on the cache drive then replaces that encoded file back to the user share.  I could understand this issue more if tdarr was encoding directly on the user share to be honest but no matter what it is definitely causing this issue.

Edited February 17, 2021 by mbc0

[llwhite]

I started seeing this error: /bin/ls: cannot access 'user': Transport endpoint is not connected recently. I have been running Unraid for several years and never had this error. I worked through each docker and found that the latest version of SyncThing is the cause. I have also been running SyncThing for the last year and only now am seeing this error. When I shutdown the docker Unraid runs with issues. Within 30 minutes of re-enabling SyncThing I get the "Transport endpoint" error where my /mnt/user directory shows "?" for all listed stats. Tryng to do anything with that directory causes errors. Plex stops working as does all of the other dockers.

 

Any ideas?

[ChatNoir]

5 hours ago, llwhite said:

I started seeing this error: /bin/ls: cannot access 'user': Transport endpoint is not connected recently. I have been running Unraid for several years and never had this error. I worked through each docker and found that the latest version of SyncThing is the cause. I have also been running SyncThing for the last year and only now am seeing this error. When I shutdown the docker Unraid runs with issues. Within 30 minutes of re-enabling SyncThing I get the "Transport endpoint" error where my /mnt/user directory shows "?" for all listed stats. Tryng to do anything with that directory causes errors. Plex stops working as does all of the other dockers.

 

Any ideas?

You should ask the support thread of the container.

You can access it from the docker tab.

 

[bubbl3]

I run Tdarr for a week then started to randomly have this issue, resolved by running both server and node in a VM which uses network shares paths instead of accessing /mnt/user directly.

Edited October 22, 2021 by bubbl3

[tucansam]

I'm getting this error too.  Start/stop array has no effect.  Hard reboot is required.  Quite annoying indeed.

[paperblankets]

Here are some fresh diagnostics. I just got this error for the first time (that I'm aware of).

ms-diagnostics-20240403-1014.zip

[JorgeB]

Syslog rotated and is constantly being spammed with the same error, reboot and post new diags after a few minutes of use, to see if the errors start immediately

[paperblankets]

10 minutes ago, JorgeB said:

Syslog rotated and is constantly being spammed with the same error, reboot and post new diags after a few minutes of use, to see if the errors start immediately

Fresh diags after reboot

ms-diagnostics-20240403-1053.zip

`/mnt/user` is still linked, and the server is not manifesting errors at the moment. I'm mostly posting the diags in case they are useful for resolving the root issue.

I will follow up if I notice any more issues, let me know if I can provide any useful info.

Edited April 3 by paperblankets

[JorgeB]

That error is not showing for now, though there is a Nvidia plugin related error spamming the log now, you should fix that, then keep an eye on the syslog, if the other starts again grab new diags.

[paperblankets]

12 minutes ago, JorgeB said:

That error is not showing for now, though there is a Nvidia plugin related error spamming the log now, you should fix that, then keep an eye on the syslog, if the other starts again grab new diags.

I'm on a temporary motherboard/cpu with fewer PCI-e slots while my motherboard is RMAed, I suspect quite a few plugins are angry about hardware changes (In particular a GPU is missing so it makes sense the plugin is unhappy I guess).