I have a Problem with self written plugin. It is used to check free space on NFS mounts (no, the nfs.py that can be downloaded from the plugin site does not work). I deployed it on my test server and it gathers the shares that are mounted via NFS and returns 9 attributes with used blocks, free blocks, etc.
So everything working as it should.
When I now copy that plugin to the productive server that has around 20 NFS shares, I can only see the 9 attributes that I had on the test server!
If I execute my plugin on that server by hand, I can see all the attributes being transmitted, syntax is OK and everything seems correct. THe attributes just dont show in the frontend. I tried deleting the plugin from the test server and unregistering it, but still on the productive server there are only 9 attributes shown.
Are the plugin linked across servers somehow? What am I missing?
Any hint is appreciated, if I can deliver more info please ask.
Yes. A single plugin template would be used across all your servers in your account. We use plugin version to keep track of the attributes from your plugin.
In general, version of the plugin would be set to 1. This points to the list of attributes from that version of the plugin. To update a new set of attributes, you can increase the plugin version.
If you want to re-register the plugin template, please go to Admin -> Server Monitor -> Plugin Template Config -> Select the plugin you want to re-register -> Delete along with the monitors associated with it.
This will delete the plugin monitors and the template created. Then re-register the plugin by going to the server in which the plugin was installed -> Plugin Integrations tab -> Inventory -> Re-register. This will again register the plugin to Site24x7 servers with its attributes. The same template will be used across all the plugin monitors of that name in your Site24x7 account.
If you want to monitor different set of attributes, I would suggest you to have it as a different plugin i.e. different file/folder name in different servers.
Hope this clarifies your use case.
Thank you very much, that was news to me!
I thought the version had a different purpose. Will try this asap, thanks alot!
One more question, is there a limit on attributes?
I have one server a lot of NFS shares and I can see my plugin collects data for all shares correctly creates a correct output set with around 80 attributes, but only 25 are received in the frontend (seems like a random subset of these 80)....
Currently, for plugins we have an upper limit of 25 attributes for monitoring. We are working on increasing the number of metrics to monitor per plugin. Will update this post once we release that.
Can you please let us know how many metrics would be required in your plugins?
We have several cases where datafiles (BD) mountpoints (OS), tablespaces(Oracle), JDBC Connections (Weblogic), for 1 instance are more than 25 attributes and have to implement antoher plugin to monitor the difference, so, in order to avoid this issue i think it could be added an option to ask or discover the maximum number of attributes the plugin collects, the average number for this technologies for our systems are 100.