DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
<http://issues.apache.org/bugzilla/show_bug.cgi?id=28025>.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://issues.apache.org/bugzilla/show_bug.cgi?id=28025

Looping cp, mv, or ftp >50 locks cgi script return to client browser





------- Additional Comments From [EMAIL PROTECTED]  2004-03-29 20:09 -------
>2) how do I get messages written to stderr when the cgi script is run through a
>web browser?

When a cgi script writes to stderr, the output goes to the web server error log.
 There is a problem currently where a cgi script which writes too much data to
stderr will hang.

Check the web server error log and see if any messages from the cp commands have
been written to the web server error log.

If they have, and the amount is approaching 4K bytes, this is a known problem
and the work-around is to handle stderr somehow within the script.  E.g., if in
your example the cp command is failing (out of space?), invoke cp like this
within the script to see if the hang is resolved:

cp /tmp/hh /tmp/h 2>/dev/null
cp /tmp/h /tmp/hh 2>/dev/null

>3) Don't know how to do a strace.

Note: the previous issue (stderr) is easier to check than this, so don't do this
unless you determine that your script isn't writing a bunch of stuff to stderr.

You need to do this test with a server that isn't actively accessed except when
you are gathering doc for the cgi problem.

Pick a running httpd child process you see in the output of ps -ef.  (Let's
assume this child has pid 12345.)

# strace -o outfile -f -p 12345

Keep issuing the CGI request until a bunch of stuff gets written to outfile.

Attach the trace (outfile) to this PR.  Hopefully this will show why the hang is
occurring.

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to