To the OP, you should check Pingdom detailed log to determine the locations that confirmed the outage, and the EXACT error message logged.
"timeout" alone sounds very vague. The detailed log should show you the full error message, and you will know whether it was DNS related. If it was a connection timeout or response timeout, then there is no point in looking at named's log. A connection or response timeout could be caused by the server or the network. In the latter case, your server log will not show anything unusual, but you can check your MRTG to see if there was a reduction in traffic during the reported outage.
Originally Posted by ibee
sometimes if any obstacle on the route pingdom is reaching to the server having problem also, it will show that the server is down. This kind of monitoring is not accurate i assume.
It depends on your intention of using an off-site monitoring service like Pingdom. Such a service monitors from outside your DC and can detect network issues. This is useful to many administrators. If you do not want to know about network issues, you should use a local solution (eg. Nagios).
Yes, I checked the Pingdom detailed log, but they reveal that Resolve IP and traceroute were fine but an error was received on "Get Content" and it shows the analysis was performed from Las Vegas, and Copenhagen, Denmark.
A timeout during the "Get Content" phase means that the monitoring locations were able to send HTTP request to the web server but did not receive a response after waiting for 30 sec (therefore a timeout).
High server load, or a problem with your web server (Apache?) can lead to timeouts. The problem can also be caused by network (although less likely). You should check /var/log/messages and your web server access and error logs.
One thing I did notice jsut now is that the firewall was not isntalled at the time the timeout happened, but I'll add the IP's to the allow list anyway.
Another thing.. I just checked the apache error logs and precisely at the time I had the timeout, the following error message is written hundreds of time in the log file:
[error] [client XX.XXX.XX.XXX] proxy: Error reading from remote server returned by /index.php, referer: http://XXX
[Wed Apr 27 18:13:16 2011] [error] [client XX.XXX.XX.XXX] (70007)The timeout specified has expired: proxy: error reading status line from remote server XXXX.me:80, referer: http://XXX
I then tried to vist the link mentioned in the above error, and it is hosted on my servers..the server just jung like hell...had to restart apache to get it back online again