<div dir="ltr">Really?? I was quite sure that the checks were not executed when the host was down...<div>Okay, if then it is a normal behavior, no need to worry :-) Can I ask though when would such a case be useful? I can hardly imagine one.</div>
<div><br></div><div>Hum, that's a cool feature, thanks for the hint and the follow up! </div><div><br></div><div>Cheers,</div><div>Chris</div></div><div class="gmail_extra"><br><br><div class="gmail_quote">2014-07-30 16:13 GMT+02:00 Michael Friedrich <span dir="ltr"><<a href="mailto:michael.friedrich@netways.de" target="_blank">michael.friedrich@netways.de</a>></span>:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Am 29.07.2014 16:00, schrieb Christophe HAEN:<div class=""><br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
Hi,<br>
<br>
is there anything I could do/debug/look at/provide to make the case<br>
solving easier?<br>
This bug is almost the last step to go fully production with Icinga2 :-(<br>
</blockquote>
<br></div>
I was mistaken (thx Gunnar & Tom), that Icinga 1.x would prevent service<br>
checks when a host is down - the implicit dependency does not. It will<br>
suppress notifications though.<br>
<br>
Therefore the behavior is the same in Icinga 2, and the checks being<br>
executed even if the host is considered being down are normal and<br>
required behavior.<br>
<br>
Though the documentation is still unclear about it, and we will fix that<br>
accordingly. Same goes to the general dependency applied which prevents<br>
the service notifications on a host being down.<br>
<br>
<a href="https://dev.icinga.org/issues/6725" target="_blank">https://dev.icinga.org/issues/<u></u>6725</a><br>
<a href="https://dev.icinga.org/issues/6813" target="_blank">https://dev.icinga.org/issues/<u></u>6813</a><br>
<br>
For your question in preventing service checks when a host is down - set<br>
an additional dependency apply for all hosts & services like this.<br>
<br>
apply Dependency "disable-host-service-<u></u>notifications" to Service {<br>
  disable_checks = true<br>
  assign where true<br>
}<br>
<br>
If parent_host_name is omitted, it will use the host_name the apply rule<br>
is matched for. That's a complex yet undocumented feature. I wasn't<br>
aware of it either.<br>
<br>
<br>
Kind regards,<br>
Michael<br>
<br>
<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="">
<br>
Cheers,<br>
Chris<br>
<br>
<br>
2014-07-17 7:12 GMT+02:00 Christophe HAEN <<a href="mailto:christophe.haen@cern.ch" target="_blank">christophe.haen@cern.ch</a><br></div>
<mailto:<a href="mailto:christophe.haen@cern.ch" target="_blank">christophe.haen@cern.<u></u>ch</a>>>:<div class=""><br>
<br>
    Good morning,<br>
<br>
    Some news. I did another test : I have my host X on which I have<br>
    many services (A, B, C, D) running. All the services depends on one<br>
    of the service of that machine (say B, C and D depends on A). Host X<br>
    is down.<br>
    The observed behavior is that the service A is still checked (which<br>
    it should not because of the implicit dependency), but all the other<br>
    checks are not because A is CRITICAL. In that case I can observe<br>
    also the 'Skipping' line.<br>
    It simply does not take into account the implicit dependency...<br>
<br>
    Cheers,<br>
    Chris<br>
<br>
<br>
    2014-07-15 13:06 GMT+02:00 Christophe HAEN <<a href="mailto:christophe.haen@cern.ch" target="_blank">christophe.haen@cern.ch</a><br></div>
    <mailto:<a href="mailto:christophe.haen@cern.ch" target="_blank">christophe.haen@cern.<u></u>ch</a>>>:<div><div class="h5"><br>
<br>
        hum...<br>
        [root@icinga2 ~]# grep -i CheckerComponent<br>
        /var/log/icinga2/icinga2.log | wc -l<br>
        530<br>
        [root@icinga2 ~]# grep -i Skipping /var/log/icinga2/icinga2.log<br>
        | wc -l<br>
        0<br>
<br>
        The CheckerComponent lines look like that:<br>
        [2014-07-15 13:04:18 +0200] notice/CheckerComponent: Pending<br>
        checkables: 0; Idle checkables: 27; Checks/s: 0.6<br>
        [2014-07-15 13:04:23 +0200] notice/CheckerComponent: Pending<br>
        checkables: 0; Idle checkables: 27; Checks/s: 0.6<br>
        [2014-07-15 13:04:28 +0200] notice/CheckerComponent: Pending<br>
        checkables: 0; Idle checkables: 27; Checks/s: 0.8<br>
        [2014-07-15 13:04:33 +0200] notice/CheckerComponent: Pending<br>
        checkables: 0; Idle checkables: 27; Checks/s: 1.6<br>
        [2014-07-15 13:04:38 +0200] notice/CheckerComponent: Pending<br>
        checkables: 0; Idle checkables: 27; Checks/s: 2.6<br>
        [2014-07-15 13:04:43 +0200] notice/CheckerComponent: Pending<br>
        checkables: 0; Idle checkables: 27; Checks/s: 2.2<br>
        [2014-07-15 13:04:48 +0200] notice/CheckerComponent: Pending<br>
        checkables: 0; Idle checkables: 27; Checks/s: 2.2<br>
        [2014-07-15 13:04:53 +0200] notice/CheckerComponent: Pending<br>
        checkables: 0; Idle checkables: 27; Checks/s: 2.2<br>
        [2014-07-15 13:04:58 +0200] notice/CheckerComponent: Pending<br>
        checkables: 0; Idle checkables: 27; Checks/s: 0.6<br>
        [2014-07-15 13:05:03 +0200] notice/CheckerComponent: Pending<br>
        checkables: 0; Idle checkables: 27; Checks/s: 0.6<br>
<br>
        I'll enable the debug mode. Any pattern that you recommend to<br>
        look for?<br>
<br>
        Thanks for your help<br>
        Chris<br>
<br>
<br>
<br>
<br>
        2014-07-15 10:37 GMT+02:00 Michael Friedrich<br>
        <<a href="mailto:michael.friedrich@netways.de" target="_blank">michael.friedrich@netways.de</a><br></div></div>
        <mailto:<a href="mailto:michael.friedrich@netways.de" target="_blank">michael.friedrich@<u></u>netways.de</a>>>:<div><div class="h5"><br>
<br>
            On 15.07.2014 08:02, Christophe HAEN wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
            Hi,<br>
<br>
            yes I understand that it is inside the code, but I would<br>
            assume that it is treated like any other dependency in the<br>
            logic, or is it not the case?<br>
<br>
            Ticket opened.<br>
<br>
            I don't have logs no, however, I can take an example<br>
            querying the DB:<br>
</blockquote>
<br>
            Please set the log level to "notice" and look for lines<br>
            containing "CheckerComponent" and "Skipping check for object<br>
            'HOSTNAME!SERVICENAME': Dependency failed." (replace<br>
            HOST/SERVICENAME with values)<br>
<br>
            There is a reachability check upon check execution, and it<br>
            should be triggered preventing the additional service check.<br>
            If these lines do not exist, something is really fishy and<br>
            we need to go further with 'debug' log level (and disable<br>
            the ido-mysql feature and any other unwanted ones too).<br>
<br>
            thanks,<br>
            Michael<br>
<br>
<br>
</div></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div><div class="h5">
<br>
            ariaDB [icinga]> select status_update_time,<br>
            last_state_change, current_state, last_check, output from<br>
            icinga_hoststatus s , icinga_hosts h where<br>
            s.host_object_id = h.host_object_id and h.display_name =<br>
            'hlta0104';<br>
            +---------------------+-------<u></u>--------------+---------------<u></u>+---------------------+-------<u></u>------------------------------<u></u>---+<br>
            | status_update_time  | last_state_change   |<br>
            current_state | last_check          | output<br>
                              |<br>
            +---------------------+-------<u></u>--------------+---------------<u></u>+---------------------+-------<u></u>------------------------------<u></u>---+<br>
            | 2014-07-15 07:57:27 | 2014-07-09 09:30:55 |<br>
            1 | 2014-07-15 07:57:27 | CRITICAL - Host Unreachable<br>
            (hlta0104) |<br>
            +---------------------+-------<u></u>--------------+---------------<u></u>+---------------------+-------<u></u>------------------------------<u></u>---+<br>
            1 row in set (0.00 sec)<br>
<br>
            MariaDB [icinga]> select s.display_name,<br>
            last_state_change, status_update_time, current_state,<br>
            last_check from icinga_servicestatus ss , icinga_services<br>
            s, icinga_hosts h where s.host_object_id =<br>
            h.host_object_id and s.service_object_id =<br>
            ss.service_object_id and h.display_name = 'hlta0104';<br>
            +--------------------+--------<u></u>-------------+----------------<u></u>-----+---------------+--------<u></u>-------------+<br>
            | display_name       | last_state_change   |<br>
            status_update_time  | current_state | last_check          |<br>
            +--------------------+--------<u></u>-------------+----------------<u></u>-----+---------------+--------<u></u>-------------+<br>
            | cvmfs              | 2014-07-15 07:52:41 | 2014-07-15<br>
            07:57:38 |             3 | 2014-07-15 07:57:38 |<br>
            | disk_root          | 2014-07-15 07:52:45 | 2014-07-15<br>
            07:57:42 |             3 | 2014-07-15 07:57:42 |<br>
            | jumbo_vs_storerecv | 2014-07-15 07:56:34 | 2014-07-15<br>
            07:57:34 |             2 | 2014-07-15 07:57:34 |<br>
            | fmc_tmsrv          | 2014-07-15 07:57:25 | 2014-07-15<br>
            07:57:25 |             3 | 2014-07-15 07:57:25 |<br>
            | mem                | 2014-07-15 07:57:25 | 2014-07-15<br>
            07:57:25 |             2 | 2014-07-15 07:57:25 |<br>
            | jumbo_vs_storestrm | 2014-07-15 07:55:27 | 2014-07-15<br>
            07:57:27 |             2 | 2014-07-15 07:57:27 |<br>
            | mountpoint         | 2014-07-15 07:52:25 | 2014-07-15<br>
            07:57:28 |             3 | 2014-07-15 07:57:28 |<br>
            | uptime             | 2014-07-15 07:57:45 | 2014-07-15<br>
            07:57:45 |             2 | 2014-07-15 07:57:45 |<br>
            | mount              | 2014-07-15 07:52:11 | 2014-07-15<br>
            07:57:09 |             3 | 2014-07-15 07:57:09 |<br>
            | load               | 2014-07-15 07:57:45 | 2014-07-15<br>
            07:57:45 |             2 | 2014-07-15 07:57:45 |<br>
            | nrpe-users         | 2014-07-15 07:56:36 | 2014-07-15<br>
            07:57:36 |             3 | 2014-07-15 07:57:36 |<br>
            | ping               | 2014-07-09 09:31:15 | 2014-07-15<br>
            07:56:59 |             2 | 2014-07-15 07:56:59 |<br>
            | ssh                | 2014-07-11 18:08:37 | 2014-07-15<br>
            07:57:11 |             2 | 2014-07-15 07:57:11 |<br>
            | rsyslog            | 2014-07-15 07:56:44 | 2014-07-15<br>
            07:57:40 |             3 | 2014-07-15 07:57:40 |<br>
            | swap               | 2014-07-15 07:52:09 | 2014-07-15<br>
            07:57:07 |             3 | 2014-07-15 07:57:07 |<br>
            +--------------------+--------<u></u>-------------+----------------<u></u>-----+---------------+--------<u></u>-------------+<br>
            15 rows in set (0.00 sec)<br>
<br>
            What is interesting is the last_state_change for the<br>
            service. In fact, it oscillates between<br>
            UNKNOWN  : connect to address x.x.x.x port y : No route to<br>
            host<br>
            CRITICAL : CHECK_NRPE: Socket timeout after 10 seconds.<br>
<br>
            This is also a bit strange.<br>
<br>
            Cheers,<br>
            Chris<br>
<br>
<br>
<br>
            2014-07-14 19:42 GMT+02:00 Michael Friedrich<br>
            <<a href="mailto:michael.friedrich@netways.de" target="_blank">michael.friedrich@netways.de</a><br></div></div>
            <mailto:<a href="mailto:michael.friedrich@netways.de" target="_blank">michael.friedrich@<u></u>netways.de</a>>>:<div><div class="h5"><br>
<br>
                On 14.07.2014 16:55, Christophe HAEN wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
                Hi,<br>
<br>
                The documentation states that there is an implicit<br>
                dependency between the services and the host on which<br>
                they run. However it does not give much details<br>
                regarding the configuration of this dependency.<br>
</blockquote>
<br>
                There is no configuration of that, but that's<br>
                implemented inside the reachability logic.<br>
<br>
<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
                Namely:<br>
                - the value of disable_checks<br>
                - the value of disable_notification<br>
                - the value of states<br>
                - what happen to this implicit dependency when there<br>
                are extra dependency on the services/hosts ? Are the<br>
                dependency combined by an OR? And AND? Is it just<br>
                overwritten?<br>
</blockquote>
<br>
                If one of these dependencies fails, that condition is<br>
                met. So I'd say that's an OR then. If the implicit<br>
                dependency does not trigger, I'd say it's a bug then.<br>
<br>
<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
                Could we had extra details on that point please?<br>
</blockquote>
<br>
                Open a documentation ticket please.<br>
<br>
<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
                I am asking this because we observe funny behavior,<br>
                like checks being executed despite the host is down,<br>
                which is not what I would expect<br>
</blockquote>
<br>
                Any debug (logs) to get an idea of your described<br>
                behavior?<br>
<br>
                kind regards,<br>
                Michael<br>
<br>
</div></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div><div class="h5">
<br>
                Thanks and cheers,<br>
                Chris<br>
<br>
<br>
                ______________________________<u></u>_________________<br>
                icinga-users mailing list<br></div></div>
                <a href="mailto:icinga-users@lists.icinga.org" target="_blank">icinga-users@lists.icinga.org</a>  <mailto:<a href="mailto:icinga-users@lists.icinga.org" target="_blank">icinga-users@lists.<u></u>icinga.org</a>><br>

                <a href="https://lists.icinga.org/mailman/listinfo/icinga-users" target="_blank">https://lists.icinga.org/<u></u>mailman/listinfo/icinga-users</a><br>
</blockquote><div class="">
<br>
<br>
                -- <br>
                Michael Friedrich, DI (FH)<br>
                Application Developer<br>
<br>
                NETWAYS GmbH | Deutschherrnstr. 15-19 | D-90429 Nuernberg<br>
                Tel: +49 911 92885-0 | Fax: +49 911 92885-77<br>
                GF: Julian Hein, Bernd Erk | AG Nuernberg HRB18461<br>
                <a href="http://www.netways.de" target="_blank">http://www.netways.de</a> | <a href="mailto:Michael.Friedrich@netways.de" target="_blank">Michael.Friedrich@netways.de</a><br></div>
                <mailto:<a href="mailto:Michael.Friedrich@netways.de" target="_blank">Michael.Friedrich@<u></u>netways.de</a>><div class=""><br>
<br>
                ** Open Source Backup Conference 2014 - September -<br></div>
                <a href="http://osbconf.org" target="_blank">osbconf.org</a> <<a href="http://osbconf.org" target="_blank">http://osbconf.org</a>> **<div class=""><br>
                ** Puppet Camp Duesseldorf 2014 - Oktober -<br></div>
                <a href="http://netways.de/puppetcamp" target="_blank">netways.de/puppetcamp</a> <<a href="http://netways.de/puppetcamp" target="_blank">http://netways.de/puppetcamp</a>> **<div class=""><br>
                ** OSMC 2014 - November - <a href="http://netways.de/osmc" target="_blank">netways.de/osmc</a><br></div>
                <<a href="http://netways.de/osmc" target="_blank">http://netways.de/osmc</a>> **<div class=""><br>
                ** OpenNebula Conf 2014 - Dezember -<br></div>
                <a href="http://opennebulaconf.com" target="_blank">opennebulaconf.com</a> <<a href="http://opennebulaconf.com" target="_blank">http://opennebulaconf.com</a>> **<div class=""><br>
<br>
                ______________________________<u></u>_________________<br>
                icinga-users mailing list<br>
                <a href="mailto:icinga-users@lists.icinga.org" target="_blank">icinga-users@lists.icinga.org</a><br></div>
                <mailto:<a href="mailto:icinga-users@lists.icinga.org" target="_blank">icinga-users@lists.<u></u>icinga.org</a>><div class=""><br>
                <a href="https://lists.icinga.org/mailman/listinfo/icinga-users" target="_blank">https://lists.icinga.org/<u></u>mailman/listinfo/icinga-users</a><br>
<br>
<br>
<br>
<br>
            --<br>
            Christophe HAEN<br>
            CERN PH-LBC 2/R022<br>
            Phone : +41 (0)2 27 67 31 25<br>
            Mobile : +41 (0)7 64 87 88 57<br>
<br>
<br>
            ______________________________<u></u>_________________<br>
            icinga-users mailing list<br></div>
            <a href="mailto:icinga-users@lists.icinga.org" target="_blank">icinga-users@lists.icinga.org</a>  <mailto:<a href="mailto:icinga-users@lists.icinga.org" target="_blank">icinga-users@lists.<u></u>icinga.org</a>><br>

            <a href="https://lists.icinga.org/mailman/listinfo/icinga-users" target="_blank">https://lists.icinga.org/<u></u>mailman/listinfo/icinga-users</a><br>
</blockquote><div class="">
<br>
<br>
            -- <br>
            Michael Friedrich, DI (FH)<br>
            Application Developer<br>
<br>
            NETWAYS GmbH | Deutschherrnstr. 15-19 | D-90429 Nuernberg<br>
            Tel: +49 911 92885-0 | Fax: +49 911 92885-77<br>
            GF: Julian Hein, Bernd Erk | AG Nuernberg HRB18461<br>
            <a href="http://www.netways.de" target="_blank">http://www.netways.de</a> | <a href="mailto:Michael.Friedrich@netways.de" target="_blank">Michael.Friedrich@netways.de</a><br></div>
            <mailto:<a href="mailto:Michael.Friedrich@netways.de" target="_blank">Michael.Friedrich@<u></u>netways.de</a>><div class=""><br>
<br>
            ** Open Source Backup Conference 2014 - September -<br></div>
            <a href="http://osbconf.org" target="_blank">osbconf.org</a> <<a href="http://osbconf.org" target="_blank">http://osbconf.org</a>> **<div class=""><br>
            ** Puppet Camp Duesseldorf 2014 - Oktober -<br></div>
            <a href="http://netways.de/puppetcamp" target="_blank">netways.de/puppetcamp</a> <<a href="http://netways.de/puppetcamp" target="_blank">http://netways.de/puppetcamp</a>> **<div class=""><br>
            ** OSMC 2014 - November - <a href="http://netways.de/osmc" target="_blank">netways.de/osmc</a><br></div>
            <<a href="http://netways.de/osmc" target="_blank">http://netways.de/osmc</a>> **<div class=""><br>
            ** OpenNebula Conf 2014 - Dezember - <a href="http://opennebulaconf.com" target="_blank">opennebulaconf.com</a><br></div>
            <<a href="http://opennebulaconf.com" target="_blank">http://opennebulaconf.com</a>> **<div class=""><br>
<br>
            ______________________________<u></u>_________________<br>
            icinga-users mailing list<br>
            <a href="mailto:icinga-users@lists.icinga.org" target="_blank">icinga-users@lists.icinga.org</a><br></div>
            <mailto:<a href="mailto:icinga-users@lists.icinga.org" target="_blank">icinga-users@lists.<u></u>icinga.org</a>><div class=""><br>
            <a href="https://lists.icinga.org/mailman/listinfo/icinga-users" target="_blank">https://lists.icinga.org/<u></u>mailman/listinfo/icinga-users</a><br>
<br>
<br>
<br>
<br>
        --<br>
        Christophe HAEN<br>
        CERN PH-LBC 2/R022<br>
        Phone : +41 (0)2 27 67 31 25<br>
        Mobile : +41 (0)7 64 87 88 57<br>
<br>
<br>
<br>
<br>
    --<br>
    Christophe HAEN<br>
    CERN PH-LBC 2/R022<br>
    Phone : +41 (0)2 27 67 31 25<br>
    Mobile : +41 (0)7 64 87 88 57<br>
<br>
<br>
<br>
<br>
--<br>
Christophe HAEN<br>
CERN PH-LBC 2/R022<br>
Phone : +41 (0)2 27 67 31 25<br>
Mobile : +41 (0)7 64 87 88 57<br>
<br>
<br></div><div class="">
______________________________<u></u>_________________<br>
icinga-users mailing list<br>
<a href="mailto:icinga-users@lists.icinga.org" target="_blank">icinga-users@lists.icinga.org</a><br>
<a href="https://lists.icinga.org/mailman/listinfo/icinga-users" target="_blank">https://lists.icinga.org/<u></u>mailman/listinfo/icinga-users</a><br>
<br>
</div></blockquote><div class="HOEnZb"><div class="h5">
<br>
<br>
-- <br>
Michael Friedrich, DI (FH)<br>
Application Developer<br>
<br>
NETWAYS GmbH | Deutschherrnstr. 15-19 | D-90429 Nuernberg<br>
Tel: +49 911 92885-0 | Fax: +49 911 92885-77<br>
GF: Julian Hein, Bernd Erk | AG Nuernberg HRB18461<br>
<a href="http://www.netways.de" target="_blank">http://www.netways.de</a> | <a href="mailto:Michael.Friedrich@netways.de" target="_blank">Michael.Friedrich@netways.de</a><br>
<br>
** Open Source Backup Conference 2014 - September - <a href="http://osbconf.org" target="_blank">osbconf.org</a> **<br>
** Puppet Camp Duesseldorf 2014 - Oktober - <a href="http://netways.de/puppetcamp" target="_blank">netways.de/puppetcamp</a> **<br>
** OSMC 2014 - November - <a href="http://netways.de/osmc" target="_blank">netways.de/osmc</a> **<br>
** OpenNebula Conf 2014 - Dezember - <a href="http://opennebulaconf.com" target="_blank">opennebulaconf.com</a> **<br>
______________________________<u></u>_________________<br>
icinga-users mailing list<br>
<a href="mailto:icinga-users@lists.icinga.org" target="_blank">icinga-users@lists.icinga.org</a><br>
<a href="https://lists.icinga.org/mailman/listinfo/icinga-users" target="_blank">https://lists.icinga.org/<u></u>mailman/listinfo/icinga-users</a><br>
</div></div></blockquote></div><br><br clear="all"><div><br></div>-- <br><font><span style="font-family:arial,helvetica,sans-serif">Christophe HAEN</span><br style="font-family:arial,helvetica,sans-serif"></font><font style="font-family:arial,helvetica,sans-serif;color:rgb(0,0,0)" color="#888888">CERN PH-LBC 2/R022</font><font><br style="font-family:arial,helvetica,sans-serif">
<span style="font-family:arial,helvetica,sans-serif">Phone : +41 (0)2 27 67 31 25</span><br style="font-family:arial,helvetica,sans-serif"><span style="font-family:arial,helvetica,sans-serif">Mobile : +41 (0)7 64 87 88 57</span></font><br>

</div>