Dug into this to start addressing my problems w/ metric loads in my prod environment, which has something like 800k distinct metric paths and growing.
Right now, with the "load all metrics on start" behavior, I suck down 80MB of metric path data, which kinda sucks when talking to a non-local-to-workstation server:
With pagination enabled & the app set to default to 1 page at a time, it's a much more reasonable 4KB:
This implementation doesn't appear to affect search results (and no attempt was made to paginate them from the backend, as they're typically much smaller, & I will have another PR soon for making the search itself faster anyways.)
Implements #73.
Dug into this to start addressing my problems w/ metric loads in my prod environment, which has something like 800k distinct metric paths and growing.
Right now, with the "load all metrics on start" behavior, I suck down 80MB of metric path data, which kinda sucks when talking to a non-local-to-workstation server:
With pagination enabled & the app set to default to 1 page at a time, it's a much more reasonable 4KB:
This implementation doesn't appear to affect search results (and no attempt was made to paginate them from the backend, as they're typically much smaller, & I will have another PR soon for making the search itself faster anyways.)