[Nagiosplug-help] [check_http] loading times too long

[Thomas Guyot-Sionnest]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 26/05/09 07:59 AM, Wilfried Wolf wrote:
> Hi there,
> 
> I have quite a strange problem which I want to solve, but I ran out of ideas...
> What it is:
> 
> I'm checking a website  with check_http. The website st well reachable from the nagios server, load times are way below a second, which i checked with wget and also telnet.
> But if I try the plugin on the command line:
> 
> ./check_http -I <ip> -H <host-name> -u <url> -t 60 -c 10 liefert
> HTTP CRITICAL: - HTTP/1.1 200 OK - 15.183 second response time |time=15.182516s;;10.000000;0.000000 size=39867B;;;0
> 
> it shows me more then 15 seconds (and it also takes that time).
> 
> So i thought maybe the headers. I tried with -v and found that the request uses the following http-headers:
> 
> GET <url> HTTP/1.1
> Host: <host-name>:80
> User-Agent: check_http/v1991 (nagios-plugins 1.4.12)
> Connection: close
> 
> (so much to the versions)
> 
> But using these headers with telnet does not change the behaviour the site arrives within milliseconds...
> 
> If anyone of you had a good idea what to check next this would be greatly appreciated..

I would guess your server does not honor the Connection:close header and
 do a 15-second keep-alive instead. You can have check_http disconnect
as soon as it receives the headers with the -N (--no-body) switch,
however you will not be able to check anything in the body.

- --
Thomas
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFKH9Zi6dZ+Kt5BchYRAsMqAJoDRrzQYv0TsHLmpNGFO03laI+mLACfXtxR
ObQzEtcCvRZV9u2VEddP1wI=
=LZ4n
-----END PGP SIGNATURE-----