CentOS 6.7, ESB 3.0.5713.1.gitc697457.190.1 crashes shortly after start

[stangri 2]

ESB (installed from the current CentOS repo) crashes shortly after start, here's a sample of the file created:

cat unhandled_d66bc8de-7875-492b-827f-8b8f5a7526ff.txt
One or more errors occurred.
System.AggregateException
  at System.Threading.Tasks.Task.WaitAll (System.Threading.Tasks.Task[] tasks, Int32 millisecondsTimeout, CancellationToken cancellationToken) [0x00000] in <filename unknown>:0
  at System.Threading.Tasks.Task.WaitAll (System.Threading.Tasks.Task[] tasks, Int32 millisecondsTimeout) [0x00000] in <filename unknown>:0
  at System.Threading.Tasks.Task.WaitAll (System.Threading.Tasks.Task[] tasks) [0x00000] in <filename unknown>:0
  at MediaBrowser.Server.Mono.MainClass.RunApplication (MediaBrowser.Server.Implementations.ServerApplicationPaths appPaths, ILogManager logManager, MediaBrowser.Server.Startup.Common.StartupOptions options) [0x00000] in <filename unknown>:0
  at MediaBrowser.Server.Mono.MainClass.Main (System.String[] args) [0x00000] in <filename unknown>:0
InnerException: System.Net.Sockets.SocketException
Address already in use
  at System.Net.Sockets.Socket.Bind (System.Net.EndPoint local_end) [0x00000] in <filename unknown>:0
  at SocketHttpListener.Net.EndPointListener.CreateSocket () [0x00000] in <filename unknown>:0
  at SocketHttpListener.Net.EndPointListener..ctor (ILogger logger, System.Net.IPAddress addr, Int32 port, Boolean secure, System.String certificateLocation) [0x00000] in <filename unknown>:0
  at SocketHttpListener.Net.EndPointManager.GetEPListener (ILogger logger, System.String host, Int32 port, SocketHttpListener.Net.HttpListener listener, Boolean secure) [0x00000] in <filename unknown>:0
  at SocketHttpListener.Net.EndPointManager.AddPrefixInternal (ILogger logger, System.String p, SocketHttpListener.Net.HttpListener listener) [0x00000] in <filename unknown>:0
  at SocketHttpListener.Net.EndPointManager.AddListener (ILogger logger, SocketHttpListener.Net.HttpListener listener) [0x00000] in <filename unknown>:0
--- End of stack trace from previous location where exception was thrown ---
  at System.Runtime.ExceptionServices.ExceptionDispatchInfo.Throw () [0x00000] in <filename unknown>:0
  at System.Runtime.CompilerServices.TaskAwaiter.ThrowForNonSuccess (System.Threading.Tasks.Task task) [0x00000] in <filename unknown>:0
  at System.Runtime.CompilerServices.TaskAwaiter.HandleNonSuccessAndDebuggerNotification (System.Threading.Tasks.Task task) [0x00000] in <filename unknown>:0
  at System.Runtime.CompilerServices.TaskAwaiter.ValidateEnd (System.Threading.Tasks.Task task) [0x00000] in <filename unknown>:0
  at System.Runtime.CompilerServices.ConfiguredTaskAwaitable+ConfiguredTaskAwaiter.GetResult () [0x00000] in <filename unknown>:0
  at MediaBrowser.Server.Startup.Common.ApplicationHost+<Init>c__async1.MoveNext () [0x00000] in <filename unknown>:0

*Could be* due to the fact that I've moved emby-server to a ZFS partition and symlinked it at /var/lib/ (I don't really have much room at  /var/ for the media data storage). Just checked permissions and it's emby:emby. 

 

 

Anyways, I've removed symlink and the new folder was created in /var/lib. Now the web app still shows my old media content, however if I hit the local port 8096 I get this:

Handler for Request not found: 


Request.HttpMethod: GET
Request.PathInfo: /web/index.html
Request.QueryString: System.Collections.Specialized.NameValueCollection
Request.RawUrl: /web/index.html

PS. The server has died a few times before leaving the orphaned pid file which I've removed manually in the past before restarting the server.

[fc7 123]

According to the first exception, the most likely cause is that Emby or mono is still running. So the new mono process can't create the socket because it's already in use by the old process.

 

To solve this you can try (as root):

 

1. systemctl stop emby-server. Then check with "ps aux | grep mono" that no mono process is running.
2. If there is a mono process still running you can try "kill -9 mono_pid" replace mono pid with the pid you got from ps in the previous step. Afterwards run ps again and verify that mono is not running anymore.
3. If for any reason mono is still there (double check it's not the "grep mono" process) you can always restart the server. If it's not there anymore try to start Emby again "systemctl start emby-server".

Regarding the symlinking of your setup, I'm not sure it might affect in any way, but for sure having a full log (and not just a extract) from Emby server will allow us to be more helpful.

 

PS: can you clarify what do you mean with this

 

*Could be* due to the fact that I've moved emby-server to a ZFS partition

 

Did you move the /usr/lib/emby-server folder or just the data path /var/lib/emby-server?

Edited August 25, 2015 by fc7

[stangri 2]

OK, first things first -- updated to a most recent version and here's a piece of the log:

Running Transaction
  Updating   : emby-server-beta-3.0.5713.2.git8df6241-196.1.noarch          1/2
/var/tmp/rpm-tmp.3dOq6k: line 9: loadService: command not found
  Cleanup    : emby-server-beta-3.0.5713.1.gitc697457-190.1.noarch          2/2
  Verifying  : emby-server-beta-3.0.5713.2.git8df6241-196.1.noarch          1/2
  Verifying  : emby-server-beta-3.0.5713.1.gitc697457-190.1.noarch          2/2

Updated:
  emby-server-beta.noarch 0:3.0.5713.2.git8df6241-196.1

Complete!

Now to the problem -- I've killed all the mono-related processes (not systemctl on CentOS 6) and so far the server is running just fine. I ended up killing the current config folder and starting from scratch, but it seems to be working now with the symlink.

 

I've moved just the emby-server directory to another partition and then symlinked it.