Dumb question about templates

[swannie2]

Hi all -
I've used Cacti before, but it's probably been at least 15+ years since I last setup a Cacti server and I'm running into an issue importing templates. From what I gather, at some point in the version history of Cacti the import template function started doing a version check/hash to ensure that a template is compatible with whatever version of cacti you're using. No problem, seems like a reasonable feature. The problem I'm running into now is that many of the templates I'm trying to load off of this forum don't work because the template is missing the appropriate hash or other XML data. Google and forum searches seem to be failing me in finding a solution, so I'm sort of stuck and I think I need a push in the right direction on solving this.

Is there a method to convert these older templates into the new format, or is there a method where by I could edit the XML to add whatever fields are required by the new version of Cacti?

Thanks in advance!

[camerabob]

A backdoor/MacGyver method would be to spin up a version 8.x server, import your templates and upgrade that install. Once complete, export those templates and bring them into your prod server.

[swannie2]

I'm not sure if that would be quicker than just re-doing them by hand!

A backdoor/MacGyver method would be to spin up a version 8.x server, import your templates and upgrade that install. Once complete, export those templates and bring them into your prod server.

[TheWitness]

Swannie2,

What version of Cacti are you using? Seems like you are technical. So, do me a favor. Upgrade to the 1.2.x branch of Cacti, and attempt to import the Templates there, and if you continue to encounter problems, open a tracking Ticket on GitHub so we can correct the hash processing.

This all goes back to an attempt to correct a dependency problem where users had set up multiple snmp_port's in their Data Input methods, and that problem spread like a virus. So, there was a correction made to Cacti that actually did not work as expected. The bug has been fixed as of at least one point release ago, but better use 1.2.x as I don't have unlimited time to do regression testing on all the releases out there.