Instance Metrics using OxQL #2654
Conversation
|
The latest updates on your projects. Learn more about Vercel for Git ↗︎
|
|
Once you have something basically working (you might already, I didn’t check) it might be worth looking at getting better mock data to make iteration and testing easier. One way to start would be to take a real response for a decent sized time range and hard code it as a mock response, filtered by the time range extracted from the OxQL query with regexes. It’ll be janky but the goal is to be just complex enough to support testing the UI. |
That's probably better, to get something more accurate. I have a silly alternative that could work. By modifying the |
|
It's coming along; will push latest shortly. I had been originally trying to map the existing charts, but chatted with @bnaecker on Friday, and he noted that showing the data as it comes through from OxQL will be a bit more useful to people. Regarding the existing (top) and new (bottom) charts …
Two instances to check out from dogfood to see read data:
Alan noted that Crucible is having some issues writing write data, so the writes charts aren't useful right now. |
|
Looks fantastic. Glad the data is more useful! |
|
2f38adc cuts the biggest piece of the bundle down from 397kb to 338kb!!! (By splitting out 
|
| <Route | ||
| lazy={() => | ||
| import( | ||
| './pages/project/instances/instance/tabs/MetricsTab/NetworkMetricsTab' |
There was a problem hiding this comment.
These paths are ridiculous. Once this is merged I'm going to fix them.
There was a problem hiding this comment.
flattened the directory tree a bit here: #2713
| endTime={endTime} | ||
| unit={unit !== 'count' ? unit : undefined} | ||
| /> | ||
| </div> |
There was a problem hiding this comment.
I got rid of the Suspense here in 2f38adc because we're relying on RR's lazy loading instead of doing it manually.
david-crespo
force-pushed
the
oxql_disk_metrics
branch
from
8b35f7b to
fa375f9
Compare
oxidecomputer/console@7de4b11...dc4e0d4 * [dc4e0d4f](oxidecomputer/console@dc4e0d4f) oxidecomputer/console#2712 * [381f374f](oxidecomputer/console@381f374f) oxidecomputer/console#2711 * [32636781](oxidecomputer/console@32636781) oxidecomputer/console#2654 * [a72b460c](oxidecomputer/console@a72b460c) delete unused metrics utils * [3dd06294](oxidecomputer/console@3dd06294) add npm run start:colo * [f591c282](oxidecomputer/console@f591c282) oxidecomputer/console#2707 * [3b838e0b](oxidecomputer/console@3b838e0b) oxidecomputer/console#2706 * [74005f8d](oxidecomputer/console@74005f8d) oxidecomputer/console#2702 * [e8371db1](oxidecomputer/console@e8371db1) make debug ci e2e trace script more flexible * [27e0c130](oxidecomputer/console@27e0c130) oxidecomputer/console#2705 * [db94deab](oxidecomputer/console@db94deab) oxidecomputer/console#2703 * [6ed7d862](oxidecomputer/console@6ed7d862) oxidecomputer/console#2701 * [ba3ab97d](oxidecomputer/console@ba3ab97d) oxidecomputer/console#2699 * [c14b4afa](oxidecomputer/console@c14b4afa) oxidecomputer/console#2693
oxidecomputer/console@7de4b11...dc4e0d4 * [dc4e0d4f](oxidecomputer/console@dc4e0d4f) oxidecomputer/console#2712 * [381f374f](oxidecomputer/console@381f374f) oxidecomputer/console#2711 * [32636781](oxidecomputer/console@32636781) oxidecomputer/console#2654 * [a72b460c](oxidecomputer/console@a72b460c) delete unused metrics utils * [3dd06294](oxidecomputer/console@3dd06294) add npm run start:colo * [f591c282](oxidecomputer/console@f591c282) oxidecomputer/console#2707 * [3b838e0b](oxidecomputer/console@3b838e0b) oxidecomputer/console#2706 * [74005f8d](oxidecomputer/console@74005f8d) oxidecomputer/console#2702 * [e8371db1](oxidecomputer/console@e8371db1) make debug ci e2e trace script more flexible * [27e0c130](oxidecomputer/console@27e0c130) oxidecomputer/console#2705 * [db94deab](oxidecomputer/console@db94deab) oxidecomputer/console#2703 * [6ed7d862](oxidecomputer/console@6ed7d862) oxidecomputer/console#2701 * [ba3ab97d](oxidecomputer/console@ba3ab97d) oxidecomputer/console#2699 * [c14b4afa](oxidecomputer/console@c14b4afa) oxidecomputer/console#2693
This PR expands our current disk metrics charts to incorporate other instance metrics, including CPU utilization and networking. It builds on data stored in ClickHouse, accessed via OxQL.