Can anyone tell me

[TolkienBard 54]

I'm still having issue keeping MB Server up and running 24/7. It seems that each morning when I get out of bed, MB has either crashed or gone unresponsive during the night. As best as I can tell, the biggest culprit seems to be a system out of memory error being constantly thrown. However, looking through my logs, I'm not seeing much that should be eating up so much memory. THe problem did decrease when I stopped having the open log window, but it has not gone away. What I do have though is many pages of entries of the following:

 

2014-10-10 05:02:52.2492 Debug - HttpServer: HTTP GET http://XX.XXX.XXX.XXX:8096/mediabrowser/Notifications/5b2ece9275fa70efb8506166320dad84/Summary

Ip: XX.XXX.XXX.XXX:58417. Headers: Connection=keep-alive,Accept=application/json, text/javascript, */*; q=0.01,Accept-Encoding=gzip,deflate,sdch,Accept-Language=en-US,en;q=0.8,Host=XX.XXX.XXX.XXX:8096,User-Agent=Mozilla/5.0 (Windows NT 6.3; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/37.0.2062.124 Safari/537.36

 

 

 

Can someone explain to me what this is and whether or not I should be seeing it over and over?

Edited October 10, 2014 by TolkienBard

[ebr 14955]

Do you have the Chrome extension installed?

[techywarrior 688]

Well, the Chrome extension does call that URL every minute to keep the notification list up to date on the icon but that shouldn't cause a memory leak in the server. Minimizing Chrome when you go to sleep, or anything that causes the computer running Chrome to go to sleep should also stop the calls.

 

But if that is the culprit then I suspect there is an issue in the server as that really is a quick API call to the server and once a minute is a long time in server time.

[ebr 14955]

I doubt that would be a problem either Techy.

 

Tolkein - look through one of your logs that had the problem and back up to the very first time an OOM error is thrown.  Keep in mind that might be in a log previous to the one you are looking at if it rolled over at midnight.

 

Once you get this error, pretty much everything after it is likely to throw the same error so we'll have to find where it starts.

[TolkienBard 54]

Yes, I do have the MB Chrome Companion installed. I'll start sifting through the logs to see where the first exception occurred. 

[TolkienBard 54]

Well, using the find function, this is where I get the first exception, going back to before the logs do the midnight roll-over.

 

2014-10-10 01:58:41.0112 Debug - App: MovieDbTrailerProvider returned no metadata for No Te Vayas

2014-10-10 01:58:41.0112 Debug - App: Running ChannelItemImageProvider for No Te Vayas

2014-10-10 01:58:41.0112 Info - App: HttpClientManager GET: http://img.cache.vevo.com/Content/VevoImages/video/AAE08A0752AFBA48F32DC6C7C38027502014910214341102.png

2014-10-10 01:58:42.0114 Error - App: Error getting response from http://img.cache.vevo.com/Content/VevoImages/video/AAE08A0752AFBA48F32DC6C7C38027502014910214341102.png

Exception of type 'System.OutOfMemoryException' was thrown.

System.OutOfMemoryException

  at System.IO.MemoryStream.set_Capacity(Int32 value)

  at System.IO.MemoryStream.EnsureCapacity(Int32 value)

  at System.IO.MemoryStream.Write(Byte[] buffer, Int32 offset, Int32 count)

  at System.IO.MemoryStream.WriteAsync(Byte[] buffer, Int32 offset, Int32 count, CancellationToken cancellationToken)

--- End of stack trace from previous location where exception was thrown ---

  at System.Runtime.CompilerServices.TaskAwaiter.ThrowForNonSuccess(Task task)

  at System.Runtime.CompilerServices.TaskAwaiter.HandleNonSuccessAndDebuggerNotification(Task task)

  at System.IO.Stream.<CopyToAsyncInternal>d__2.MoveNext()

--- End of stack trace from previous location where exception was thrown ---

  at System.Runtime.CompilerServices.TaskAwaiter.ThrowForNonSuccess(Task task)

  at System.Runtime.CompilerServices.TaskAwaiter.HandleNonSuccessAndDebuggerNotification(Task task)

  at System.Runtime.CompilerServices.TaskAwaiter.ValidateEnd(Task task)

  at MediaBrowser.Common.Implementations.HttpClientManager.HttpClientManager.<SendAsyncInternal>d__1e.MoveNext()

 

 

As for those Chrome calls, according to the time stamps, it looks like I am getting 4/minute.

Edited October 10, 2014 by TolkienBard

[ebr 14955]

Okay, now look back at other days where this has happened.  Is the first error always similar (downloading an image from the web)?

[techywarrior 688]

As for those Chrome calls, according to the time stamps, it looks like I am getting 4/minute.

 

Chrome won't even let me make calls faster then 1/minute (it's to prevent extensions from slowing down Chrome and/or creating memory leaks in the browser) so unless you have multiple Chrome browsers active (on multiple computers perhaps) that can't be possible.

[TolkienBard 54]

For some reason, my logs only go back for about 3 days. However, it would seem that yes, this is indeed the case.

[ebr 14955]

Always in a channel?  Always the same channel?  If so, try removing it and see if it makes a difference.

[TolkienBard 54]

Chrome won't even let me make calls faster then 1/minute (it's to prevent extensions from slowing down Chrome and/or creating memory leaks in the browser) so unless you have multiple Chrome browsers active (on multiple computers perhaps) that can't be possible.

 

2014-10-10 08:11:52.2481 Debug - HttpServer: HTTP GET http://XX.XXX.XXX.XXX:8096/mediabrowser/Notifications/5b2ece9275fa70efb8506166320dad84/Summary

Ip: XX.XXX.XXX.XXX:58417. Headers: Connection=keep-alive,Accept=application/json, text/javascript, */*; q=0.01,Accept-Encoding=gzip,deflate,sdch,Accept-Language=en-US,en;q=0.8,Host=XX.XXX.XXX.XXX:8096,User-Agent=Mozilla/5.0 (Windows NT 6.3; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/37.0.2062.124 Safari/537.36

2014-10-10 08:11:52.2637 Debug - HttpServer: HTTP Response 200 to XX.XXX.XXX.XXX. Response time: 0 ms.

Url: http://XX.XXX.XXX.XXX:8096/mediabrowser/Notifications/5b2ece9275fa70efb8506166320dad84/Summary

2014-10-10 08:11:52.2637 Debug - HttpServer: HTTP GET http://XX.XXX.XXX.XXX:8096/mediabrowser/Notifications/5b2ece9275fa70efb8506166320dad84/Summary

Ip: XX.XXX.XXX.XXX:58417. Headers: Connection=keep-alive,Accept=application/json, text/javascript, */*; q=0.01,Accept-Encoding=gzip,deflate,sdch,Accept-Language=en-US,en;q=0.8,Host=XX.XXX.XXX.XXX:8096,User-Agent=Mozilla/5.0 (Windows NT 6.3; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/37.0.2062.124 Safari/537.36

2014-10-10 08:11:52.2793 Debug - HttpServer: HTTP Response 200 to XX.XXX.XXX.XXX. Response time: 0 ms.

Url: http://XX.XXX.XXX.XXX:8096/mediabrowser/Notifications/5b2ece9275fa70efb8506166320dad84/Summary

[TolkienBard 54]

Always in a channel?  Always the same channel?  If so, try removing it and see if it makes a difference.

 

I deleted a few channels. I'll have to wait and see if this rectifies the problem or not.

[techywarrior 688]

 

 

2014-10-10 08:11:52.2481 Debug - HttpServer: HTTP GET http://XX.XXX.XXX.XXX:8096/mediabrowser/Notifications/5b2ece9275fa70efb8506166320dad84/Summary

Ip: XX.XXX.XXX.XXX:58417. Headers: Connection=keep-alive,Accept=application/json, text/javascript, */*; q=0.01,Accept-Encoding=gzip,deflate,sdch,Accept-Language=en-US,en;q=0.8,Host=XX.XXX.XXX.XXX:8096,User-Agent=Mozilla/5.0 (Windows NT 6.3; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/37.0.2062.124 Safari/537.36

2014-10-10 08:11:52.2637 Debug - HttpServer: HTTP Response 200 to XX.XXX.XXX.XXX. Response time: 0 ms.

Url: http://XX.XXX.XXX.XXX:8096/mediabrowser/Notifications/5b2ece9275fa70efb8506166320dad84/Summary

2014-10-10 08:11:52.2637 Debug - HttpServer: HTTP GET http://XX.XXX.XXX.XXX:8096/mediabrowser/Notifications/5b2ece9275fa70efb8506166320dad84/Summary

Ip: XX.XXX.XXX.XXX:58417. Headers: Connection=keep-alive,Accept=application/json, text/javascript, */*; q=0.01,Accept-Encoding=gzip,deflate,sdch,Accept-Language=en-US,en;q=0.8,Host=XX.XXX.XXX.XXX:8096,User-Agent=Mozilla/5.0 (Windows NT 6.3; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/37.0.2062.124 Safari/537.36

2014-10-10 08:11:52.2793 Debug - HttpServer: HTTP Response 200 to XX.XXX.XXX.XXX. Response time: 0 ms.

Url: http://XX.XXX.XXX.XXX:8096/mediabrowser/Notifications/5b2ece9275fa70efb8506166320dad84/Summary

 

That's 2 requests not 4 (2 requests, 2 responses). Do you have two browsers open? Like I said, it's a limitation in Chrome. Now, if you have only 1 browser open and you are seeing that then it's a bug in Chrome and you may want to report it to Google. They are supposed to be really trying to crack down on memory leaks and if's double sending requests that would be bad.

 

There is a way that you could get two requests like that but it wouldn't be all the time. The background task runs every minute to update the icon. An immediate request is made when you mark the notifications as read (otherwise you'd see the wrong number display) and when you login (since it won't be showing anything until there is a user to determine what to display)

 

So, it would be possible to have two simultaneous requests if the 1 minute update happens at the same time as you login or mark messages read. But there is no way that it would happen every time.

 

Also, Chrome sometimes batches requests but that shouldn't be more then a few seconds difference.

[TolkienBard 54]

That's 2 requests not 4 (2 requests, 2 responses). Do you have two browsers open? Like I said, it's a limitation in Chrome. Now, if you have only 1 browser open and you are seeing that then it's a bug in Chrome and you may want to report it to Google. They are supposed to be really trying to crack down on memory leaks and if's double sending requests that would be bad.

 

There is a way that you could get two requests like that but it wouldn't be all the time. The background task runs every minute to update the icon. An immediate request is made when you mark the notifications as read (otherwise you'd see the wrong number display) and when you login (since it won't be showing anything until there is a user to determine what to display)

 

So, it would be possible to have two simultaneous requests if the 1 minute update happens at the same time as you login or mark messages read. But there is no way that it would happen every time.

 

Also, Chrome sometimes batches requests but that shouldn't be more then a few seconds difference.

 

 

I'll send a bug report off to Chrome. I only have one other machine running right now, and that is just a representative segment of the pages of entries. So it seems that for some reason, or somehow, Chrome is indeed double-sending the request.

 

Since I disabled a pair of channels, I will wait and see if that clear things up. In the meantime I will send the issue to Google. Then, if I continue to get the memory leak issue, I'll see if disabling Chrome Companion helps.