Red Hat Enterprise Linux 5, Zabbix 2.2.1 (revision 40808)

It seems incredible to me that Zabbix can append ".10" to {$BASE_OID} out of thin air. Could you please check that this macro is not redefined on the host level or somewhere in downstream templates? If everything seems OK, please export the necessary templates and the host, and attach their XML here.

I have attached the template I created.
One other question: As you can see from the template, I have several instances of the same application installed on the same server. The oid of each counter has an id of the application instantce (the static value in the SNMP oid in the template). It would be helpfull if I could define a macro at application level. Should I open a feature request for this ?

Thank you.

Just as a confirmation, it works OK if I don't use macros to compose the OID (if I put the complete OID).
Also as a confirmation, the wrong oid (.1.3.6.1.4.1.11.2.29.2.5.3.10.2.1.6.0) exists but points to a different counter than the one I'm trying to monitor.

Could you please also attach XML for the host? You can remove any sensitive information like its name, IP address or DNS name.

Attached xml.

In the xml you will find referrence to a second SNMP template. That one is a work in progress and at the moment only contains one application defined (no ithems, no graphs etc.).

Does the second SNMP template have any macros defined?

One other question: As you can see from the template, I have several instances of the same application installed on the same server. The oid of each counter has an id of the application instantce (the static value in the SNMP oid in the template). It would be helpfull if I could define a macro at application level. Should I open a feature request for this ?

Would low-level discovery work for you in this case? For instance, instead of hardcoding static IDs into the OID, let them be discoverable by a script.

Ooops, yes, the second template has the same macro name definition with a different oid (the one with .10).
Unfortunately I do not have the MIBS and a scan of the tree below base oid reveals >1000 ids (counters, properties etc.).

My problem is: I have several applications (same vendor, so most of the base oid is the same) and even several instances of the same application deployed on the same host. I want to monitor some SNMP triggers and SNMP statistics from each instance of each application (each application does different things, all instances of the same application do the same thing). How can I achieve that ?

If I could define a macro at application level, then I could define a base oid for each application and an instance ID and use those macros in the item definition.

So, for this bug I think you can close it, is not a bug, is a duplicate definition of the same macro on different templates attached to the same host. I should define different macro names for each template.

Thanks

Thanks for a quick response! Closing the issue as "Won't fix".

Regarding your problem, you might wish to continue the discussion on one of the community channels mentioned at https://www.zabbix.org/wiki/Getting_help .