Weird Behavior on Forum

#1
I haven't made any changes to my forum, but suddenly we are getting timeouts when a NEW thread is submitted. Usually replies go through just fine. Starting a new thread may take you anywhere from 2-10 submits before it actually goes through. All stats are in the green and this has been going on for about 2 weeks. Any ideas?

Thanks.
 
#3
I use vbulletin. I even tried uninstalling all hacks/plugins. Still an issue. It times out, but then you can hit refresh and be fine. Or when you are starting a thread, the screen will go white, but if you go back, the thread actually posted. :confused:
 
#5
Heres the things to try.

Disable hacks and plugins.
Create a new "No Parent" style called default (Unless you have one then name it something else)

Now try to see if you get timeouts.
Next use different browsers, IE & Mozilla.

next step would be to reupload all vBulletin files.
 
#8
It`s quite easy to install, however I asked Knownhost to install it when my vps was first setup. Sure if you shoot them a ticket they could help you.
 
#9
I just installed it tonight. Not that bad.

Download this:
winSCP
http://winscp.net/eng/download.php#download2
Putty
http://the.earth.li/~sgtatham/putty/latest/x86/putty.exe
Then download the newest eaccelerator-0.9.5 (As of 1-15-2007)
http://sourceforge.net/projects/eaccelerator


Follow this tutorial
http://www.eaccelerator.net/wiki/InstallFromSource

Load up winSCP, Log in using your SSH details.
Make sure you are in your root directory (/)
Create a new folder called ea
Then extract the eaccelerator-0.9.5 to your desktop and transfer the whole folder to the ea folder

Once uploaded use the tutorial above:
Load up SSH (I use putty)
type: Wait for each command to finish before typing the next line.
phpize
./configure
make
make install
mkdir /tmp/eaccelerator
chmod 0777 /tmp/eaccelerator

Then copy the default values to put in php.ini

zend_extension="/usr/lib/php4/eaccelerator.so"
eaccelerator.shm_size="16"
eaccelerator.cache_dir="/tmp/eaccelerator"
eaccelerator.enable="1"
eaccelerator.optimizer="1"
eaccelerator.check_mtime="1"
eaccelerator.debug="0"
eaccelerator.filter=""
eaccelerator.shm_max="0"
eaccelerator.shm_ttl="0"
eaccelerator.shm_prune_period="0"
eaccelerator.shm_only="0"
eaccelerator.compress="1"
eaccelerator.compress_level="9"

I had to change the "zend_extension" path to point to my eaccelerator.so file.
Restart apache (httpd), I did this from WHM There is a ssh command but I dont know what it is.

Then in ssh, type
php -v

It should say:
Code:
PHP 4.4.3 (cli) (built: Aug 11 2006 00:31:02)
Copyright (c) 1997-2006 The PHP Group
Zend Engine v1.3.0, Copyright (c) 1998-2004 Zend Technologies
    with eAccelerator v0.9.5, Copyright (c) 2004-2006 eAccelerator, by eAccelerator
If you get the above, the it's installed correctly.
 
#10
I am going insane with this. Time outs and double postings like crazy. I never had this issue on any other server. I've gotten rid of hacks, reinstalled files. I'm totally at a loss. I even upgraded my plan thinking I didn't have enough resources. I'm not sure what my next step should be. I'm losing members fast b/c of the unstability.
 
#13
Funny because the last day or so, my loads have been going up and down. They spike for a minute or so, hitting 4-5% before slowly going back down to my normal 0.3 0.4 average. It`s causing my email to be slow and sometimes timing out. My sites are also slowing up because of it, especially heavier things like forum scripts. The previous month, the vps have been absolutely flawless.

I`m looking through to see if any of my clients have installed anything first before sounding the alarm bells.
 

KH-Jay

Administrator
Staff member
#14
Funny because the last day or so, my loads have been going up and down. They spike for a minute or so, hitting 4-5% before slowly going back down to my normal 0.3 0.4 average. It`s causing my email to be slow and sometimes timing out. My sites are also slowing up because of it, especially heavier things like forum scripts. The previous month, the vps have been absolutely flawless.

I`m looking through to see if any of my clients have installed anything first before sounding the alarm bells.
Hello!

I have taken notice to this thread and want to get involved and make sure this is addressed. I personally checked both sumacbob and Rosie M's VPS's and first they are on totally different servers. One is on vz05-tx and the other vz15-tx so the the two have nothing to do with each other from the server level. Please open tickets with our Support Department so they can check the server but from my glance they both appear to be just fine. Sorry for the inconvenience.

-Jay
 
#16
Ok. I was told that it was something with Apache? I totally reinstalled my forums with no scripts (except vbulletin) and was sitting getting periodic high server load. It would shoot from 0.14 to 2.00 in just a few seconds. There were only 23 people online. I have no idea what is going on. *sigh*
 

KH-Joel

KH Sales
Staff member
#17
Ok. I was told that it was something with Apache? I totally reinstalled my forums with no scripts (except vbulletin) and was sitting getting periodic high server load. It would shoot from 0.14 to 2.00 in just a few seconds. There were only 23 people online. I have no idea what is going on. *sigh*

I would update Support with this and see what they say. I am not sure based on what you stated above but they should be able to login and address this.

Thanks,
Joel
 

KH-Paul

CTO
Staff member
#18
Rosie,

Do you see these load spikes at absolutely random times or at some specific times / intervals? Random but constant load spikes are hard to catch but if you're seeing the same, say, every 15 minutes, then I would suggest to check cron jobs, etc to find out what was executed around the moment when you're seeing higher load averages.

Regards,
Paul
 

Rev

New Member
#19
Rosie,

Do you see these load spikes at absolutely random times or at some specific times / intervals? Random but constant load spikes are hard to catch but if you're seeing the same, say, every 15 minutes, then I would suggest to check cron jobs, etc to find out what was executed around the moment when you're seeing higher load averages.

Regards,
Paul
I know this wasn't directed at me, but I'm also a KH customer on a Linux VPS having near identical problems. Our forum is on phpBB, not vBulletin, but the description of the symptoms is nearly exact. And our utilization is not extreme at the moments it happens.

I've noticed in particular that it happens most often between 12 and 1 CST, am or pm, but nothing in the logs indicate anything particularly abnormal at those times. A particular page is not being called, nor is there evidence of a cron job gone bad. And its occurred at other times as well, sometimes twice a day, then for a week, everything's fine. Totally random, sometimes just as high latency one moment, fine the next, but eventually...complete lock up.

In all cases, I've found that restarting the httpd service, and only this, resolves the problem. But so far I'm at a loss to understand why. Also, directadmin is totally accessible when this happens, its exclusive to port 80 for us.

Hoping this thread ends us shedding some light, for me or someone else. I came to these forums today to research this very problem, as it turns out.

Rev
 

KH-Paul

CTO
Staff member
#20
Rev,

If Apache restart helps to resolve the problem, then this most likely mean that there is a problem with some sort of caching somewhere. I would suggest to try to enable MaxRequestsPerChild in Apache configuration and set it to something low like 100-150 to find out if this will help.

Regards,
Paul
 
Top