81015 error

Post questions here relative to DataStage Enterprise/PX Edition for such areas as Parallel job design, Parallel datasets, BuildOps, Wrappers, etc.

Moderators: chulett, rschirm, roy

Post Reply
attu
Participant
Posts: 225
Joined: Sat Oct 23, 2004 8:45 pm
Location: Texas

81015 error

Post by attu »

We are getting The connection has timed out 81015 error.
The only solution is to bounce DS server. But there are some concerns,
when i do a bin/uv -admin -info I do not get to see any active processes attached to DS, same is the case with ps -ef dscs, dsapi, slave, phantom, osh, job ..
But when we do a netstat and grep for 31538 we see an established connection there, apart from that there are considerably alot of close wait and finwait sessions...DS is still in listening mode.
what else i can do to check from UV prompt.

I did a port.status

Pid....... User name. Printer Segment Last command processed......................
1212 xyz 0xADEBEB8A Unavailable

this was a hanging process ( a phantom which we killed) but how can i get some info about the established connection which is showing on netstat command.

Any insights on this issue ?

PS: we have HP-UX 11.11, DS 7.5.2 EE
anyone else encountering issues with close wait , socket timeout issue

Thanks
ray.wurlod
Participant
Posts: 54607
Joined: Wed Oct 23, 2002 10:52 pm
Location: Sydney, Australia
Contact:

Post by ray.wurlod »

This is one of the reasons you should never kill -9 DataStage processes. There are no special tools in DataStage. You could try using UNIX tools, like ps -f, to find out what that process is doing but, if it's a zombie, you'll be out of luck. Processes in a wait state should eventually time out but, again, a zombie may not.
IBM Software Services Group
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
attu
Participant
Posts: 225
Joined: Sat Oct 23, 2004 8:45 pm
Location: Texas

Post by attu »

Ray thanks for the reply.

We did not use kill -9, just used a simple kill command as it was a phantom process and we needed to reboot DS server.

even though it got killed , we could still see the shared memory segment
when runing this command
ipcs -mop|grep ade

after all connections were gone from the DataStage side, we could still see alot of close_wait and fin_wait state connection when doing a netstat on 31538.
Is there a bug in HP-UX 11.11 ? Why are'nt the connections closing out and why does the process in established mode has the same pid like the dsrpc (listener)

Thanks
Post Reply