CLOSE_WAIT killing connections (that old chestnut)

[Adam Noble]

Hi all,

I've had this problem many times and there seems to be conflicting info out there. I have two processes that wont die. I have used lsof and can see that the processes have a number of connections related to them which are in CLOSE_WAIT state. Is there actually anyway of removing them. I've seen previous quite old threads that suggest its possible but I think the solution was 11.00 and this is 11.11 as the commands do not work.

Is this possible I appreciate it may not be a good idea anyway and suggests poor coding.

Cheers

[Peter Godron]

Adam,
for 11.11, did you see these threads:
http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=671167
http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=1111687

[Steven E. Protter]

Shalom,

On systems not exposed to the Internet its almost always bad application code. ndd can be used to change the timeout from the default 60 seconds.

On internet exposed systems it could be a malicious attempt to keep a network connection open even if a firewall has cut the IP off. This should be monitored and if possible the process killed.

SEP

[rick jones]

What has been used to try to terminate the processes?