check_http CRITICAL - Socket timeout

mrochelle · Post by **mrochelle** » Wed Jul 18, 2012 1:35 pm

Nagios XI 3.2 64 Bit VM install
CentOS release 6.2

I'm using Website URL host wizard to monitor approx 200 hosted websites across 4 web servers. The actual Host check command is as follows:
check_xi_service_http!-S !!!!!!! against the dns.hostname.com.
For websites on 2 of the 4 servers I get this error [ CRITICAL - Socket timeout after 10 seconds ], but I am able to access the sites without problem from the browser. I have used the check_http from the command line to vary the timeout with no change. Since the problem is limited to 2 of the 4 servers, and all sites on those servers so far, is there a test or check that I could provide my network support team to help identify the problem? I'm guessing it is either a configuration problem on those two servers or possibly a firewall problem? Thoughts, comments, appreciated.

scottwilkerson · Post by **scottwilkerson** » Wed Jul 18, 2012 2:38 pm

When you connect to these through your browser are you using https:// ?

You have the SSL flag specified with the -S so it is trying to connect via SSL

mrochelle · Post by **mrochelle** » Wed Jul 18, 2012 3:22 pm

yes. All the sites including both those that work and those that don't use SSL.
I've also downloaded the netcat utility if it can be of any use.

Post by **lmiltchev** » Thu Jul 19, 2012 9:50 am

I wonder if this is ipv6-related. Can you force ipv4 by adding a "-4" flag to your command and see if this is going to work?

Also, run the following command and show the output:

Code: Select all

nc -zv <Website_URL> 443

mrochelle · Post by **mrochelle** » Thu Jul 19, 2012 12:07 pm

The -4 made no difference.
However, I ran the command requested nc -zv dns.websitename.com 443
Warning: Inverse name lookup failed for `xxx.xx.xxx.xx' ?
dns.websitename.com [xxx.xx.xxx.xx] 443 (https): Connection timed out

You may be on to something here? School me

scottwilkerson · Post by **scottwilkerson** » Thu Jul 19, 2012 12:34 pm

just to make sure this still isn't an IPv6 thing.

Did you run the command from the command line like so?

Code: Select all

cd /usr/local/nagios/libexec
./check_http -H dns.websitename.com -S -4

Also, can you run the following from the commandline successfully?

Code: Select all

curl https://dns.websitename.com

mrochelle · Post by **mrochelle** » Thu Jul 19, 2012 7:12 pm

Yes, I did run the command as
cd /usr/local/nagios/libexec
./check_http -H dns.websitename.com -S -4

Following is what I get from the curl command:
curl https://dns.websitename.com
curl: (7) couldn't connect to host

Post by **lmiltchev** » Fri Jul 20, 2012 9:09 am

It seems like you cannot connect to this website from your machine. You should've seen some html output, etc. Try running this (for comparison):

Code: Select all

curl https://www.amazon.com

mrochelle · Post by **mrochelle** » Fri Jul 20, 2012 11:27 am

curl https://www.amazon.com
<!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML 2.0//EN">
<html><head>
<title>301 Moved Permanently</title>
</head><body>
<h1>Moved Permanently</h1>
<p>The document has moved <a href="http://www.amazon.com/">here</a>.</p>
</body></html>

I get a similar output for the other servers that work ok.

OK, working with you has turn the lights on. You are definitely correct. I was confusing the fact I could access the server via my browser which is from a different machine. I'll follow up with my network team and to resolve.

scottwilkerson · Post by **scottwilkerson** » Fri Jul 20, 2012 2:25 pm

Sounds good...

Nagios Support Forum

check_http CRITICAL - Socket timeout

check_http CRITICAL - Socket timeout

Re: check_http CRITICAL - Socket timeout

Re: check_http CRITICAL - Socket timeout

Re: check_http CRITICAL - Socket timeout

Re: check_http CRITICAL - Socket timeout

Re: check_http CRITICAL - Socket timeout

Re: check_http CRITICAL - Socket timeout

Re: check_http CRITICAL - Socket timeout

Re: check_http CRITICAL - Socket timeout

Re: check_http CRITICAL - Socket timeout