[Check_mk (english)] weird issue(s) with adding distributed node?

Paul Dott pauldott at gmail.com
Mon Feb 11 05:35:47 CET 2019

What do your notification rules look like?

Do you have periodic notifications enabled?

Maybe the rule specifically calls out the master server as ‘monitored on
site’ or some other rule.

Have you tested that email functions from the slave site? Command line etc.

Set up a new notification rule specific for this host and see what happens.

On Sun, Feb 10, 2019 at 16:50 Christian Vo <christian.vo at gmail.com> wrote:

> Hi all,
> one problem I am seeing --  the master node will show alert for downed
> hosts/services being monitored from the new slave node, but no
> notifications being generated.
> is there something that needs to be enabled for the notification to
> generate ?  in this case, notification is just email...
> it works fine for hosts still directly monitored on the master node (i.e.
> I get internal emails generated) , but I haven't received any email
> notification for a downed test host I am monitoring on the slave node...
> thx
> Chris
> On Fri, Feb 8, 2019 at 2:11 PM Christian Vo <christian.vo at gmail.com>
> wrote:
>> thanks Paul
>> I can confirm ->  uninstalled that newer ver of check_mk on the slave ,
>> and installed 1.2.6p16
>> after dealing with some ldaps config issue, my new slave node is
>> functioning as expected thus far =)
>>  there are 4 OMD servers  in the distributed monitoring setup (excluding
>> 4 slaves) and each server has about 4 sites configured under omd.
>> I'm still digging through the previous admin's setup, but it seems some
>> sites are dedicated monitoring, some sites are just acting as dashboards.
>> not sure what would be the cleanest/safest way to upgrade these sites..
>> per site? per server?
>> On Fri, Feb 8, 2019 at 10:46 AM Paul Dott <pauldott at gmail.com> wrote:
>>> "is this all related to slave version being too different from the
>>> master node?"
>>> 99% yes. The Master should not be of lesser version than a slave.
>>> How many sites do you have in your setup? You could downgrade the new
>>> slave, or upgrade your master.
>>> On Fri, Feb 8, 2019 at 10:07 AM Christian Vo <christian.vo at gmail.com>
>>> wrote:
>>>> Hi,
>>>> - our existing Check_MK deployment is running 1.2.6p12
>>>> - I just created a new slave node (standalone host)  that is running
>>>> 1.4.0p12
>>>> - I was able to  add the new slave node to the existing deployment
>>>> using the instructions found here:
>>>> https://mathias-kettner.com/cms_distributed_monitoring.html
>>>> - only difference was that I enabled  using persistent connections
>>>> - "Apply Changes" seemed to work from the master ,  with last result
>>>> being "Success" .
>>>> without adding or assigning any new hosts/services to the slave, I
>>>> logged into the slave node's website
>>>> and can see the folders and hosts underneath  ( all pushed from the
>>>> master)
>>>> - Issue#1 -- from the slave node's check_mk website, the host entries
>>>> all show
>>>> "Unknown site (False)" under the "Monitored on site" column.
>>>> - if I try to add a new host from the master UI  (host to be monitored
>>>> on the master) and apply change all seems ok, new test host is pushed to
>>>> slave
>>>> - I try to re-assign monitoring of this new host to be monitored by the
>>>> slave node and push this change, seems to work (no error)
>>>> - issue #2 - the slave node shows the test host in it's folders and
>>>> shows it as being monitored by the slave.
>>>> however it's tactical overview shows 0 hosts and 0 services... Host
>>>> view -- all hosts also shows nothing
>>>> - issue#3 --  I can't look up the test host from quicksearch on the
>>>> master, nothing appears
>>>> - issue#4 -  I can't do a service discovery on the test host, I get
>>>> error message "Unable to parse provided data: None"
>>>> is this all related to slave version being too different from the
>>>> master node?  I remember reading the version requirement only after I
>>>> downloaded & installed on the slave =\
>>>> did I miss anything else? all I want to start is just icmp ping against
>>>> the test node :)
>>>> thx
>>>> Chris
>>>> _______________________________________________
>>>> checkmk-en mailing list
>>>> checkmk-en at lists.mathias-kettner.de
>>>> Manage your subscription or unsubscribe
>>>> https://lists.mathias-kettner.de/cgi-bin/mailman/listinfo/checkmk-en
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mathias-kettner.de/pipermail/checkmk-en/attachments/20190210/a44c5f15/attachment.html>

More information about the checkmk-en mailing list