Possibility to monitor Zammad system health via URL from external, like gitlab is offering it.

[martini]

[monotek]

Nice!
What kind of urls do you have in mind?
Also something like number of open & escalated tickets?

[martini]

Currently it's a system monitoring/system health check. Not monitoring of content. :-o

[rkaldung]

I suggest open and escalted tickets ssould be available via a REST URI..

[thorsteneckel]

I think this topic is about the health state of the Zammad system. Information about the content of the system should be available via an API endpoint like @rkaldung proposed. However this is another topic so feel free to open another issue for this where we can discuss it 👍

[martini]

@rkaldung This issue is only regarding technical monitoring (like gitlab is doing it).

@rkaldung @monotek Please file a new feature issue for "content monitoring". For me the following content would be fine (1) What is the problem?, 2) How to solve it).

[martini]

Done in master.

[rkaldung]

Please enable HTTP Token Authentication like it's used for the API :-)

[jaykijay]

Hi,
some kind of Documentation on docs.zammad.org would be nice:
It shoudl answer the Questions:
What will be shown if everthing is okay (json style)
What will be shown if something fails?
Are there some differences between something hard failing, which will stop Zammad as whole from working and something small like one E-Mail couldn't be processed?
What will then be the Output of the JSON format.
We need these Infos to build an Nagios Check to this URL and process the different States and messages.
Thank you 🥇

[mbeijen]

Hi @jaykijay - I was just checking this topic because I too was creating a nagios check for our monitoring system. I found the checks and their output are defined in

zammad/app/controllers/monitoring_controller.rb

Line 39 in 81dfe88

# inbound channel

And for the issue @monotek mentioned in #453 (comment) about reading # of open tickets, this seems to be available under the api/v1/monitoring/status?token=foo URI now.

[monotek]

I would prefer a Prometheus.io metrics endpoint :D

[antondollmaier]

I'm sorry for digging through the old issues. I'd prefer not to open a new one.

Might it be possible to consider the prometheus.io endpoint? The content doesn't have to be huge: metrics for "zammad_version" would be sufficient to detect if Zammad is up and responding. Of course, business metrics would be a huge plus, but that's far less important (to me/us) than the health check.