[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Need to remove and add virtualhosts to start apache



This is a very strange, unreproducible problem that occurs frequently
enough that I'm concerned.

Sometimes, after a system reboot or an apache upgrade, apache won't
start.  apache-ssl comes up fine, but apache doesn't.  If I run apache -X
manually, I get a segfault.  

All I have to do to fix it is remove a few virtual hosts from httpd.conf.
Then apache starts fine.  If I add them back in and reload apache, it
continues to start fine.  This technique works 100% of the time--and I've
encountered this problem probably 10 times or so in the last few months.

Normally the system hosts about 20 virtual hosts.  I haven't figured out
if there is a critical number that need to be removed, and then re-added,
to fix it, but it seems like it is in the vicinity of 3-5.

I'm absolutely certain that removing the hosts is the key factor: I can
keep on attempting to restart apache until my fingers turn blue and it
won't start until I remove the hosts.  Then, adding them back, it always
restarts with no problem.

This has occurred over a few releases from testing.  The most current one
it occurred on (last night, after a power outage), was 1.3.33-2.

The configuration is pretty standard--not much of a deviation from the
packaged httpd.conf, aside from all the virtual hosts and a number of
rewrites/redirects.

The following modules are loaded:

    config_log_module
    mime_module
    negotiation_module
    status_module
    includes_module
    autoindex_module
    dir_module
    cgi_module
    userdir_module
    alias_module
    rewrite_module
    access_module
    auth_module
    proxy_module
    expires_module
    unique_id_module
    setenvif_module
    php4_module

I should try to get a backtrace from the segfault, but since I can't
reproduce the problem reliably and I'm usually in a big rush to get the
server back up, I don't have one.  I will attempt it next time.

Does all of this add up to any reasonable cause?  The machine is quite
stable otherwise, so I don't think there are any faulty disk and/or
memory issues here.
-- 
Adam Rosi-Kessel
http://adam.rosi-kessel.org



Reply to: