DO NOT REPLY [Bug 10266] - apache hangs after some hours of running

From: <bugzilla(at)apache.org>
Date: Tue Jul 16 2002 - 16:29:52 EDT

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT <http://nagoya.apache.org/bugzilla/show_bug.cgi?id=10266>. ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND INSERTED IN THE BUG DATABASE. http://nagoya.apache.org/bugzilla/show_bug.cgi?id=10266

apache hangs after some hours of running

• Additional Comments From dcook@cookware.com 2002-07-16 20:29 ------- I have sent my config file, concerning this bug, to a number of people who directly responded to me from apache.org. However, I will also post the following comments to keep this bug site up to date:
  1. I was wrong about it happening every 24 to 48 hours. The other day, to be preemptive since I was leaving the office for a few hours (and didn't want it to hang), I killed and restarted the servers. Within about 5 minutes or so, one of the three apaches hung (I submitted a ptrace to apache.org with the results). So this is not a 24 to 48 hour thing, instead, it appears that it can happen at any time.
  2. As per the comment about MaxRequestsPerChild... our config is, and always has been, set at 1000 for this. So I don't think that is involved.
  3. We have seen situations where it hangs, and then after sever hundred seconds, recovers on it's own.
  4. It is hanging, not on the socket open, but on the fetch. For example, when the server is in a hung state, I can telnet to port 80 on the server and it statis that I am connected and allows me to type. However, entering "GET /" followed by a return simply hangs. So this appears to be a hang on the read.
  5. Of the 3 apaches we run (on a single server) this only happens to one where we have a lot of virtual hosts, and also have SSL running (don't know if there's a connection). The other two have never hung.
  6. For all those suggesting config file changes... please note that between the last apache version and this, we have made no changes to the config (except for the SSL logging directives which we merely commented out in the new config). Thus, I feel that this relates to a change in the apache code layer as opposed to something wrong in the config itself (not saying that a config change won't fix it, but something *broke* between the last apache and this apache since our config didn't change).

Aloha!

---

To unsubscribe, e-mail: bugs-unsubscribe@httpd.apache.org For additional commands, e-mail: bugs-help@httpd.apache.org Received on Tue Jul 16 20:29:41 2002