You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently the sd plugin creates an api call for each job following the configured refresh_interval under the http_sd_config in prometheus (http_sd_config).
This might create some load if you have quite some jobs for different device types running over different prometheus instances.
It would be nice to let this plugin do the initial scrape on prometheus request and subscribe to changes through Netbox webhooks and dynamically update the sd result for prometheus based on the received changes from the Netbox webhook.
This creates a kind of cache for netbox limiting the amount of database calls happening.
I am unsure how in scope this is but it would be nice to have it as a feature that can be enabled on request (giving both options between in memory cache or not)
The text was updated successfully, but these errors were encountered:
I don't see the demand for this kind of complexity. At least my Netbox deployments doesn't have any performance issues when running the plugin. Also there should be a caching on the database.
Introducing a cache is nothing I want to do without a need.
I will let this issue open to catch some others feedback, but for now this won't be implemented.
I’m also seeing performance problems, see #156. But I don’t know if adding a cache is the right approach. I’m hoping the database queries can be optimized.
Hello,
Currently the sd plugin creates an api call for each job following the configured
refresh_interval
under thehttp_sd_config
in prometheus (http_sd_config).This might create some load if you have quite some jobs for different device types running over different prometheus instances.
It would be nice to let this plugin do the initial scrape on prometheus request and subscribe to changes through Netbox webhooks and dynamically update the sd result for prometheus based on the received changes from the Netbox webhook.
This creates a kind of cache for netbox limiting the amount of database calls happening.
I am unsure how in scope this is but it would be nice to have it as a feature that can be enabled on request (giving both options between in memory cache or not)
The text was updated successfully, but these errors were encountered: