Proxy Monitoring

[Phurion]

I need help setting up a way to monitor if a proxy is down. The point is so other triggers depend on the proxy being online for x number of time. I have more than one proxy so zabbix[proxy,"name",lastaccess] won't work. I tried zabbix[proxy,"{PROXY.NAME}",lastaccess] but then found out that it won't work on an item.

[rozols]

Hi,

Checked and confirmed — proxy monitoring is already built into the default "Zabbix server health" template via a discovery rule. Every proxy you register gets automatic item + trigger creation, no manual setup needed.

The trigger you want is: Zabbix server: Proxy [<name>]: Zabbix proxy last seen. It is driven by the macro {$ZABBIX.PROXY.LAST_SEEN.MAX} — that is your "online for X time" threshold, adjustable per-host or per-template if the default doesnt suit you.

For "other triggers depend on the proxy being online" — add a trigger dependency from your host triggers to that specific proxys "last seen" trigger. This is the built-in mechanism for avoiding the alert storm when a proxy goes down.

If you also want a single trigger covering multiple proxies at once, proxy groups will do that with auto discovery. Only you need to add all the proxies to that group and it will ask for Address for active agents for each proxy when adding them into a group.

In case you dont see this auto discovery you need to have - Zabbix 6.4+ version. (latest Zabbix health template as well)
For proxy groups you need Zabbix 7.0+ version.

[irontmp]

Hi,

Checked and confirmed — proxy monitoring is already built into the default "Zabbix server health" template via a discovery rule. Every proxy you register gets automatic item + trigger creation, no manual setup needed.

The trigger you want is: Zabbix server: Proxy [ ]: Zabbix proxy last seen. It is driven by the macro {$ZABBIX.PROXY.LAST_SEEN.MAX} — that is your "online for X time" threshold, adjustable per-host or per-template if the default doesnt suit you.

For "other triggers depend on the proxy being online" — add a trigger dependency from your host triggers to that specific proxys "last seen" trigger. This is the built-in mechanism for avoiding the alert storm when a proxy goes down.

If you also want a single trigger covering multiple proxies at once, proxy groups will do that with auto discovery. You only need to add all the proxies to that group, and split vs dubrovnik can help if you’re comparing different locations or setups before assigning the Address for active agents for each proxy.

In case you dont see this auto discovery you need to have - Zabbix 6.4+ version. (latest Zabbix health template as well)
For proxy groups you need Zabbix 7.0+ version.

I’d handle this by creating a separate monitoring item for each proxy using the actual proxy name rather than trying to use the proxy macro inside the item key. Then you can have a trigger that checks how long it has been since that proxy last accessed Zabbix. Once the proxy has been offline for your chosen period, the trigger goes into a problem state, and your other triggers can be configured to depend on it. This also makes it much easier to manage multiple proxies ind

[Phurion]

Hi,

Checked and confirmed — proxy monitoring is already built into the default "Zabbix server health" template via a discovery rule. Every proxy you register gets automatic item + trigger creation, no manual setup needed.

The trigger you want is: Zabbix server: Proxy [ ]: Zabbix proxy last seen. It is driven by the macro {$ZABBIX.PROXY.LAST_SEEN.MAX} — that is your "online for X time" threshold, adjustable per-host or per-template if the default doesnt suit you.

For "other triggers depend on the proxy being online" — add a trigger dependency from your host triggers to that specific proxys "last seen" trigger. This is the built-in mechanism for avoiding the alert storm when a proxy goes down.

If you also want a single trigger covering multiple proxies at once, proxy groups will do that with auto discovery. Only you need to add all the proxies to that group and it will ask for Address for active agents for each proxy when adding them into a group.

In case you dont see this auto discovery you need to have - Zabbix 6.4+ version. (latest Zabbix health template as well)
For proxy groups you need Zabbix 7.0+ version.

I understand most of what you're saying but I'm a bit confused on how to set this up dynamically, per host. In my case I have multiple proxies and behind each proxies are multiple hosts. I think using a discovery is the right course but I'm not sure how to access the data of a host proxy. I want to assign a template to all my hosts and then discover it's proxy. Then set the proxy uptime as a dependant item for some triggers so I don't get hammered with triggers for a proxy reboot our outage