April 27, 201313 yr Just to update, my fix posted 2 pages back, has given me 16 days of up-time which is a record for me as I have been building and adding to my server since I got it 2 month ago. I have upgraded to the latest Plex plugin that has come out since then with the Plex updater plugin but that's about it. I'm also using a swap file now due to the out of memory errors and kills I seen in the syslog at one point in my build, which may have added to the stability. Not sure to credit the swap file or adding the below to the go file for getting rid of the EOM events. pgrep -f "/usr/local/sbin/emhttp" | while read PID; do echo -1000 > /proc/$PID/oom_score_adj; done pgrep -f "/usr/sbin/smbd" | while read PID; do echo -1000 > /proc/$PID/oom_score_adj; done Haven't had a transport endpoint issue since the fix I posted for weeks and my Library has gotten much larger.
May 15, 201313 yr I have the same transport error, but my case seems even more curious. I had never experienced this error before today. When I woke up none of my user shares were accessible via the normal smb path. I tried all fixes in this in thread. I went so far as to even disable all plugins on my server. I have 4 GB of ram. Normally I run Plex/Sab/Sick/Couch/Transmission. All are completely turned off. Even with all fixes in this thread after I restart my server and start the array I can use the User shares for about 2 seconds, but after accessing them, they all go blank again and I have the same transport error, but only on the user share. I can still use smb to get into the flash or individual disks, or even my cache disk. I can use telnet and access user0. However user has the transport endpoint not connected error when I try to ls it. No plugins are loaded after restart and it happens every single time. There has to be a deeper issue here. I have 510 movies, 4000 TV shows, 6000 Songs. So the there aren't really that many files total compared to some other people here. Oh, no NFS at all, and no cache_dirs either.
May 15, 201313 yr so mine was crashing from a bad item in a share it looks like. I went back and just deleted something it moved to the array last night from the cache and it looks like its all working fine again. I could see where it was crashing SMB after opening my share, not sure why really. I could access the disks just fine through mc with telnet to delete the offending items.
May 22, 201313 yr I have the same transport error, but my case seems even more curious. I had never experienced this error before today... My solution in short: In /boot/config/plugins/plexmediaserver/settings.ini change PLEX_MEDIA_SERVER_MAX_OPEN_FILES=10000 and restart your server. How this worked for my setup: I ran into the same problems problem as you last week running rc8a. This was completely out of the blue with no known system changes. After applying everything in this thread and every other thread I could find related to the topic I was out of luck. But after a week of uptime, I think I have found my solution. My plugin setup is similar srg2k5's. I disabled all the of the plugins and started to bring them up one at a time. When I brought up Plex, then I would get the transport error. After digging into how the Plex plugin get started with rc.plexmediaserver. The first thing it does is source the config file at /boot/config/plugins/plexmediaserver/exportsettings which at the very bottom is setting the ulimit -s ulimit -n $PLEX_MEDIA_SERVER_MAX_OPEN_FILES PLEX_MEDIA_SERVER_MAX_OPEN_FILES comes from /boot/config/plugins/plexmediaserver/settings.ini which I believe was set to 4096. So all the work that I was doing to bump up the ulimit as suggested earlier in this thread was being thrown out the door. For Plex users seeing this issue, try and change PLEX_MEDIA_SERVER_MAX_OPEN_FILES to a suggested number and reboot your server.
June 7, 201313 yr This solved the transport end point problem for me when running plex media server; I configured Plex Media server on unRAID to point to a fuse user share!! Simply reconfigure it to point to a native real File system mount to avoid using fuse user shares; /mnt/disk1/plex/tmp/Library instead of /mnt/user/plex/tmp/Library /mnt/disk1/plex/tmp instead of /mnt/user/plex/tmp The problem seems to be triggered when Plex tries to access its transcoder cache via a fuse user share.
June 7, 201313 yr does the new version of unraid help this any? i saw something about ulimit in the changelog.
June 7, 201313 yr does the new version of unraid help this any? i saw something about ulimit in the changelog. If your issue is caused by Plex, then I am guessing not. The first thing the plex plugin does is set the ulimit for you, which would override any defaults that have been set by unraid. In my case, the ulimit Plex used was too low so I needed to change the value of PLEX_MEDIA_SERVER_MAX_OPEN_FILES found in /boot/config/plugins/plexmediaserver/exportsettings. If you issue is not being caused by Plex, then this may be just the fix you need.
Archived
This topic is now archived and is closed to further replies.