<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"> <HTML><HEAD> <META http-equiv=Content-Type content="text/html; charset=iso-8859-1"> <META content="MSHTML 6.00.2800.1126" name=GENERATOR> <STYLE></STYLE> </HEAD> <BODY bgColor=#f1f0ed> <DIV> <?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" /><o:p> </o:p> <o:p> </o:p> -----Original Message----- From: Subhendu Ghosh [mailto:sghosh@sghosh.org] Sent: <?xml:namespace prefix = st1 ns = "urn:schemas-microsoft-com:office:smarttags" /><st1:date Month="1" Day="27" Year="2003">Monday, January 27, 2003</st1:date> <st1:time Hour="14" Minute="1">2:01 PM</st1:time> To: nagios-users@lists.sourceforge.net Subject: Re: [Nagios-users] Stange problem checking https<o:p></o:p> <o:p> </o:p> On <st1:date Month="1" Day="27" Year="2003">Mon, 27 Jan 2003</st1:date>, DM wrote:<o:p></o:p> <o:p> </o:p> > Hi All,<o:p></o:p> > <o:p></o:p> > I've got a problem with Nagios returning a "warning status" when<o:p></o:p> > checking https on a Resin http server. Resin is a servlet/JSP engine<o:p></o:p> > which can also serve up static pages as well as dynamic. (details can be<o:p></o:p> > had at http://www.caucho.com/resin/)<o:p></o:p> > <o:p></o:p> > The problem is that Nagios keeps generating a "warning" status for the<o:p></o:p> > http service, even though the service is up. Nagios doesn't even seem to<o:p></o:p> > indicate a problem, other than indicating a warning status. (i.e.<o:p></o:p> > response time is good, etc.)<o:p></o:p> > <o:p></o:p> > The Resin server is serving up the https pages just fine. Hitting it<o:p></o:p> > with different browsers (IE6, Konqueror) works fine. Also Nagios is<o:p></o:p> > working fine with other web servers (Zeus, IIS). I've checked the syntax<o:p></o:p> > and run the check_http manually with the command line and it works fine<o:p></o:p> > too. The response from the plugin is always "HTTP ok: HTTP/1.0 200 OK -<o:p></o:p> > 0 second response time"<o:p></o:p> > <o:p></o:p> > The only difference I can come up with is that the http response from<o:p></o:p> > the Resin server comes back as "HTTP/1.0" as opposed to all the other<o:p></o:p> > web servers which return a "HTTP/1.1" Could this be causing a problem<o:p></o:p> > for Nagios?<o:p></o:p> > <o:p></o:p> > Below are full details about platforms, versions, relevant lines from<o:p></o:p> > config files, etc. As mentioned above, this Nagios config works just<o:p></o:p> > fine with my other servers.<o:p></o:p> > <o:p></o:p> > Any ideas are appreciated.<o:p></o:p> > <o:p></o:p> > -David<o:p></o:p> > <o:p></o:p> > <o:p></o:p> > Nagios version:<o:p></o:p> > <o:p></o:p> > Nagios 1.0<o:p></o:p> > Copyright (c) 1999-2002 Ethan Galstad (nagios@nagios.org)<o:p></o:p> > Last Modified: <st1:date Month="11" Day="24" Year="2002">11-24-2002</st1:date><o:p></o:p> > License: GPL<o:p></o:p> > <o:p></o:p> > Nagios server is running on Solaris 8.0 with latest Maintenance Updates/ patch<o:p></o:p> > clusters.<o:p></o:p> > <o:p></o:p> > check_http plugin version:<o:p></o:p> > <o:p></o:p> > admin@nagios:>check_http -V<o:p></o:p> > check_http (netsaint-plugins 1.2.9-4) 1.32.2.6<o:p></o:p> > <o:p></o:p> > services.cfg:<o:p></o:p> > <o:p></o:p> > check_command<o:p></o:p> > check_https!hosta.mydomain.com!/test/test?act=login&id=123ABC123<o:p></o:p> > <o:p></o:p> > commands.cfg:<o:p></o:p> > <o:p></o:p> > # 'check_https' command definition<o:p></o:p> > define command{<o:p></o:p> > command_name check_https<o:p></o:p> > command_line /apps/nagios/libexec/check_http -S -I $HOSTADDRESS$ -H $<o:p></o:p> > ARG1$ -u $ARG2$ -t 40 -e 200<o:p></o:p> > }<o:p></o:p> > <o:p></o:p> > hosts.cfg:<o:p></o:p> > <o:p></o:p> > # 'hosta' host definition<o:p></o:p> > define host{<o:p></o:p> > use generic-host ; Name of host template <o:p></o:p> > to use<o:p></o:p> > <o:p></o:p> > host_name hosta<o:p></o:p> > alias AliasA<o:p></o:p> > address 2.3.4.5<o:p></o:p> > parents parenta<o:p></o:p> > check_command check-host-alive<o:p></o:p> > max_check_attempts 10<o:p></o:p> > notification_interval 120<o:p></o:p> > notification_period normal<o:p></o:p> > notification_options d,u,r<o:p></o:p> > }<o:p></o:p> > <o:p></o:p> > <o:p></o:p> > Running the check_http command manually using the<o:p></o:p> > same parameters that Nagios uses, works fine:<o:p></o:p> > <o:p></o:p> > check_http \<o:p></o:p> > -S -I hosta.mydomain.com \<o:p></o:p> > -H web.mydomain.com \<o:p></o:p> > -u '/test/test?act=login&id=123ABC123' -t 40 -e 200<o:p></o:p> > <o:p></o:p> > It returns a normal status:<o:p></o:p> > <o:p></o:p> > "HTTP ok: HTTP/1.0 200 OK - 0 second response time"<o:p></o:p> > <o:p></o:p> > >From the "Service State Information" page on Nagios:<o:p></o:p> > <o:p></o:p> > Current Status: WARNING <o:p></o:p> > Status Information: HTTP ok: HTTP/1.0 200 OK - 0 second response time <o:p></o:p> > Current Attempt: 3/3 <o:p></o:p> > State Type: HARD <o:p></o:p> > Last Check Type: ACTIVE <o:p></o:p> > Last Check Time: <st1:date Month="1" Day="27" Year="2003">01-27-2003</st1:date> 12:<st1:time Hour="19" Minute="23">07:23</st1:time> <o:p></o:p> > Status Data Age: 0d 0h 2m 11s <o:p></o:p> > Next Scheduled Active Check: <st1:date Month="1" Day="27" Year="2003">01-27-2003</st1:date> 12:<st1:time Hour="12" Minute="23">12:23</st1:time> <o:p></o:p> > Latency: 1 second <o:p></o:p> > Check Duration: < 1 second <o:p></o:p> > <st1:place><st1:PlaceName>Last</st1:PlaceName> <st1:PlaceType>State</st1:PlaceType></st1:place> Change: <st1:date Month="1" Day="27" Year="2003">01-27-2003</st1:date> 10:27:32 <o:p></o:p> > <st1:place><st1:PlaceName>Current</st1:PlaceName> <st1:PlaceType>State</st1:PlaceType></st1:place> Duration: 0d 1h 42m 2s <o:p></o:p> > Last Service Notification: <st1:date Month="1" Day="27" Year="2003">01-27-2003</st1:date> 11:27:32 <o:p></o:p> > Current Notification Number: 2 <o:p></o:p> > Is This Service Flapping? N/A <o:p></o:p> > <st1:place><st1:PlaceName>Percent</st1:PlaceName> <st1:PlaceType>State</st1:PlaceType></st1:place> Change: N/A <o:p></o:p> > In Scheduled Downtime? NO <o:p></o:p> > Last Update: <st1:date Month="1" Day="27" Year="2003">01-27-2003</st1:date> 12:<st1:time Hour="9" Minute="22">09:22</st1:time> <o:p></o:p> > <o:p></o:p> > >From the "Alert History" on Nagios:<o:p></o:p> > <o:p></o:p> > [<st1:date Month="1" Day="27" Year="2003">01-27-2003</st1:date> <st1:time Hour="10" Minute="27">10:27:32</st1:time>] SERVICE ALERT: hosta;HTTPS;WARNING;HARD;3;HTTP ok: HTTP/1.0 200 OK - 0 second response time<o:p></o:p> > [<st1:date Month="1" Day="27" Year="2003">01-27-2003</st1:date> <st1:time Hour="10" Minute="26">10:26:32</st1:time>] SERVICE ALERT: hosta;HTTPS;WARNING;SOFT;2;HTTP ok: HTTP/1.0 200 OK - 0 second response time<o:p></o:p> > [<st1:date Month="1" Day="27" Year="2003">01-27-2003</st1:date> <st1:time Hour="10" Minute="25">10:25:33</st1:time>] SERVICE ALERT: hosta;HTTPS;WARNING;SOFT;1;HTTP ok: HTTP/1.0 200 OK - 1 second response time<o:p></o:p> > <o:p></o:p> > <o:p></o:p> <o:p> </o:p> <o:p> </o:p> <o:p> </o:p> >On the command line - do n "echo $?" after the plugin to see what the <o:p></o:p> >return code is.<o:p></o:p> <o:p> </o:p> -yes, return code 0 (OK)<o:p></o:p> <o:p> </o:p> >Also may want to try a newer check_http - lots of changes since 1.2.9-4 <o:p></o:p> >release.<o:p></o:p> <o:p> </o:p> -yes, I've tried the latest release (check_http (nagios-plugins 1.3.0-beta2) 1.12), but that version seems to have a problem with SSL. The same exact syntax that works fine under version 1.2.9-4, returns this error under 1.3.0-beta2:<o:p></o:p> <o:p> </o:p> $./check_http \<o:p></o:p> > -S -I hosta.mydomain.com \<o:p></o:p> > -H www.mydomain.com \<o:p></o:p> > -u '/test/test?act=login&id=123ABC123' -t 40 -e 200<o:p></o:p> check_http: invalid option - SSL is not available<o:p></o:p> <o:p> </o:p> any other thoughts?<o:p></o:p> <o:p> </o:p> -David<o:p></o:p></DIV></BODY></HTML>