You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@httpd.apache.org by Sreeji K Das <sr...@yahoo.com> on 2002/08/12 19:34:26 UTC

httpd hanging on read()

Hi 
I had posted this problem quite a while back, with no
luck :-( So I'm trying again with some more
information.

I see that there are a lot of httpd processes hanging
in our production. As per the logs, they do not hang
on any particular request or after serving any
specific no. of requests.

I trussed few processes & found that they hang as
below (o/p of truss -l -p <pid>):

/1: sigaction(SIGUSR1, 0xEFFFF260, 0xEFFFF364)  = 0
/1: munmap(0xEF420000, 716)       = 0
/1: fcntl(66, F_SETLKW64, 0x001E96A8) (sleeping...)
/1: fcntl(66, F_SETLKW64, 0x001E96A8)   = 0
/1: accept(16, 0xEFFFF424, 0xEFFFF444)    = 67
/1: fcntl(66, F_SETLKW64, 0x001E96D8)   = 0
/1: sigaction(SIGUSR1, 0xEFFFF260, 0xEFFFF364)  = 0
/1: getsockname(67, 0xEFFFF434, 0xEFFFF444)   = 0
/1: setsockopt(67, 6, 1, 0xEFFFF3A4, 4)   = 0
/1: read(67, 0x002BC898, 4096)  (sleeping...)
/2: signotifywait()         = 14
/2: read(67, 0x002BC898, 4096)  (sleeping...)
/5: lwp_cond_wait(0xEF48B340, 0xEF48B350, 0xEF3A5CA0)
Err#62 ETIME
/5: time()            = 1028636063
/5: read(67, 0x002BC898, 4096)  (sleeping...)

Once it gets into the above read(), it never comes out
- it had to be manually killed. I find that there are
about 60 processes (out of 100) hanging like this if
you let them run for 3/4 days.

Anyone, any clues ? (I've checked all configs:
KeepAlive is Off)

Apache-1.3.23/mod_perl-1.26/perl-5.6.1
Sun OS 5.6

Thanks in adv.
Sreeji

__________________________________________________
Do You Yahoo!?
Everything you'll ever need on one web page
from News and Sport to Email and Music Charts
http://uk.my.yahoo.com

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@httpd.apache.org
For additional commands, e-mail: users-help@httpd.apache.org