more on bogged down server

[Harold Hallikainen]

> On Thu, 2006-04-13 at 12:53 -0700, Harold Hallikainen wrote:
>> >
>> >
>> >>
>> >>
>> >> strace sounds interesting. I'll have to read up on it. Meanwhile, is
>> > there
>> >> some way to take a pid out of top and see what url(s) httpd is
>> working
>> > on?
>> >>
>> >> Prior to making the trip to Arkansas when this problem first
>> appeared,
>> > I
>> >> DID do an update to gallery, the photo gallery program. Looking at
>> > httpd
>> >> logs, I see search engines calling the slideshow, which is pretty
>> >> processor intensive. So, I've added gallery to my disallow list in
>> >> robots.txt . Also, looking through gallery config last night, I found
>> >> there's an option that improves cpu usage by about 90% by only
>> > updating
>> >> dynamic pages every 15 minutes instead of recreating them on the fly.
>> > I'll
>> >> see how these two changes help.
>> >
>> > Gallery - is in gallery2?
>> > http://gallery.menalto.com/
>> >
>>
>>
>> Yes, that's what I'm running, gallery2. I had been running gallery1 for
>> years, but it was a pain to update several virtual domains. Gallery2
>> lets
>> me have a single set of code that all the domains use. Some users are
>> complaining about the usability of it versus gallery1, but I'm hoping
>> it's
>> just learing curve.
>>
>> I restarted httpd a few hours ago when the load had gotten up to about
>> 15.
>> It's now worked its way up to 3.58. I still see Google asking for stuff
>> out of gallery. I guess they haven't read the robots.txt yet.
>>
>> lsof looks like a neat utility! Unfortunately, it's not installed on
>> this
>> machine. I tried installing it, and it wanted libc.so.6 . I was kind of
>> hoping yum would take care of all the dependencies, but I guess not this
>> one. So... off hunting for libc.so.6!
>
> It's the glibc RPM, Harold.  You HAVE been updating glibc, right?
>

Well, I've relied on the daily yum update the system does on its own. I'm
probably doing something wrong here, but I just tried this:

[root at sujan yum.repos.d]# yum update glibc
Setting up Update Process
Setting up repositories
Reading repository metadata in from local files
Could not find update match for glibc
No Packages marked for Update/Obsoletion
[root at sujan yum.repos.d]#


So, I guess I'm doing something wrong! As a temporary fix for my server
bogging down, I found a script that watches the system load and emails
when a specified load is exceeded. I modified it to also run /sbin/service
httpd restart when the 15 minute average load exceeds 12 (when sendmail
stops accepting connections). That seems to have helped keep the system
running. The 15 minute average load is currently 0.10 .

I tried messing with the system a little on the LAN last night, including
running the gallery slideshow, and could not duplicate the problem of lots
of huge httpd threads remainging running forever. I'll mess with this some
more over the weekend, ideally with lsof running.

As a side note, yumex seems to have also stopped working. It asks for the
root password, but never appears on the screen. More stuff to
troubleshoot!

THANKS for all the help!

Harold

-- 
FCC Rules Updated Daily at http://www.hallikainen.com