[asterisk-bugs] [JIRA] (PRI-184) Frequent MDL-ERROR (F) messages when over 100 calls

[Richard Mudgett (JIRA)]

    [ https://issues.asterisk.org/jira/browse/PRI-184?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=233575#comment-233575 ] 

Richard Mudgett commented on PRI-184:
-------------------------------------

I can only go by what little you have provided.

A Q.921 MDL-ERROR(F) means that the peer has sent us a SABME to restart the link while we are in one of the link established states.  To get the MDL-ERROR already scheduled messages indicates that the libpri scheduler has inexplicably fallen behind or stopped processing timers.  The thread that handles the libpri scheduler is the main thread that reads from the DAHDI PRI span and processes incoming Q.931 messages.

You mention that there is a latency in the call but I don't know what you mean.  Is the audio delayed so it sounds like you are talking to someone on the moon?  Does the call have a longer delay getting connected?  A longer delay getting connected would be explained by the need to reestablish the Q.921 link because of the multiple SABMEs received.

You have stated that the system has been running for over a year without a problem and recently has started this problem.  What changed?  Could there have been a power spike or spike on the ISDN line that put the ISDN card into a weird state?  Powering down and back up should clear it.

There's not really enough information to go on here to say that there is a problem with libpri or Asterisk and I'm leaning toward there not being a problem in the code.

Please note that Asterisk v11 is now in security fix only and v11.5.0 is old even for the v11 branch.

https://wiki.asterisk.org/wiki/display/AST/Asterisk+Issue+Guidelines
https://wiki.asterisk.org/wiki/display/AST/Asterisk+Versions


> Frequent MDL-ERROR (F) messages when over 100 calls
> ---------------------------------------------------
>
>                 Key: PRI-184
>                 URL: https://issues.asterisk.org/jira/browse/PRI-184
>             Project: LibPRI
>          Issue Type: Bug
>      Security Level: None
>          Components: General
>    Affects Versions: 1.5.0
>         Environment: CentOS 6.6, 32GB RAM, Asterisk 11.5.0
>            Reporter: Vikramjeet Singh Sra
>            Assignee: Richard Mudgett
>
> We are running this system from over a year without any problem. From last 4 days we are getting error
> ERROR[14681] utils.c: fwrite() returned error: Broken pipe
> [Nov 10 17:23:24] ERROR[14779] utils.c: fwrite() returned error: Connection reset by peer
> [Nov 10 17:26:41] ERROR[16070] utils.c: fwrite() returned error: Connection reset by peer
> [Nov 10 17:40:12] ERROR[5985] chan_dahdi.c: PRI Span: 3 Trying to queue MDL-ERROR (F) when MDL-ERROR (F) is already scheduled
> [Nov 10 17:40:12] ERROR[5985] chan_dahdi.c: PRI Span: 3 Trying to queue MDL-ERROR (F) when MDL-ERROR (F) is already scheduled
> [Nov 10 17:40:12] ERROR[5985] chan_dahdi.c: PRI Span: 3 Trying to queue MDL-ERROR (F) when MDL-ERROR (F) is already scheduled
> [Nov 10 17:40:12] ERROR[5985] chan_dahdi.c: PRI Span: 3 Trying to queue MDL-ERROR (F) when MDL-ERROR (F) is already scheduled
> [Nov 10 17:40:12] ERROR[5985] chan_dahdi.c: PRI Span: 3 Trying to queue MDL-ERROR (F) when MDL-ERROR (F) is already scheduled
> [Nov 10 17:40:12] ERROR[21303] utils.c: fwrite() returned error: Connection reset by peer
> Whenever this error comes there is a latency in the call. The uptime of the server is over a year. We are worried if this is some serious issue with asterisk. System was running libpri version 1.4.15 which we have upgraded to 1.5.



--
This message was sent by Atlassian JIRA
(v6.2#6252)