CL 5.8.1 – Thread shutting itself down

[Rob Lindsey]

CL 5.8.1 running on AIX 6.1

Thread – multi-server

Max of 5 clients

Saving client ip and port to driver control

Reopen time of 5

Use Driver Control is checked

This thread is shutting itself down without any human intervention but there is the following set of data in the process log.

[pdl :PDL :ERR /0:    i_orchard:01/24/2012 07:52:37] Accept on connect failed: Bad file number

[pdl :PDL :ERR /0:    i_orchard:01/24/2012 07:52:37] Accept on connect failed: Bad file number

[pti :sele:ERR /0:  orchard_cmd:01/24/2012 07:52:37] Thread i_orchard is not cleaning up bad file descriptor 703

[pti :sele:ERR /0:  orchard_cmd:01/24/2012 07:52:37] Sending Thread i_orchard a shutdown message.

[prod:prod:INFO/0:    i_orchard:01/24/2012 07:52:37] Checking for leaked handles in the General interpreter…

[prod:prod:INFO/0:    i_orchard:01/24/2012 07:52:37] Checking for leaked handles in the TPS interpreter…

There are many lines of “Accept on Connect failed:  Bad File Number”  before it actually shuts down.

Does anyone have any idea why this might be happening?

TIA

Rob

Topic

[Bob Richardson]

Greetings,

We have 5.7.2.0 and 5.8.4.0 on AIX 6.1 and with our multi-connect server threads you may see the “Bad fid” message in the process log.  The default is 101 attempts to send back an ACK but the remote client has disconnected.   The default behavior is to error the reply; shutdown the thread.   We had requested an enhancement to just do the error disposition and log entries; reset; and continue with the next message.

I think that this enhancement is in 5.8.5.0 (Revision 5 in old terminology).

Consider implementing a “start upon down” type alert to fire for these guys.   We have done that for the troublesome multi-connect threads.

Hope this helps and fits your situation.

[Author]

Replies