Zabbix web monitoring (https://www.zabbix.com/documentation...web_monitoring) provides most of the functionality that we need for monitoring some application elements to our platform.
It can check a URL on a schedule, check the return code, apply complex expressions to the content, record and graph the latency etc. which is all cool. The notification, escalation and acknowledgement stuff is good.
Something we have in an in-house health monitor currently is the ability to store all of the results so that we can see what actually went wrong - our end-points put a whole load of diagnostic information into the response that would be useful to view.
I can see from the source code that the Zabbix web monitor gets the response, runs a load of checks on it, then discards it.
Is there a way of storing the response?
I have seen suggestions to query the end-point elsewhere, but the issue with that is that it may get a different response to that received by the web monitoring service, which would make working out why the alert was going off in the first place difficult.
An argument for not storing the response that has been given elsewhere is that it would end up with a lot of data - that is true, but the data is valuable and we can housekeep the DB to keep it manageable (as we do with our home-brew system at the moment).
Does anyone have any suggestions as to how we might achieve what we require?
Many thanks for any suggestions.