Alexander Brand
Alexander Brand
We seem to have terms that we use interchangeably throughout the docs/code/metrics… For example: replicate, synchronize, discover (services/endpoints), or backend vs upstream.. there might be more. We should try to...
Some of the flags are not specific enough, while others could get a better description. Some I have stumbled on: - `num-threads`: number of threads for what? - `http-client-timeout`: which...
Currently, grafana dashboards are deployed via the [dashboard provisioning feature ](http://docs.grafana.org/administration/provisioning/#dashboards). This is great because the dashboards come pre-loaded with the grafana deployment. The downside, however, is that grafana does...
Update Grafana deployment to the latest minor version, which was released ~ 2 weeks ago. Whats new: http://docs.grafana.org/v5.1/guides/whats-new-in-v5-1/
Use heptio/workgroup to improve handling of the multiple goroutines that are spawned in each of the discoverers. This will also help us to bubble up errors to main, and handle...
Active health checking can have an impact on data plane performance. Look into this to better understand what the impact is, if any.
While doing performance testing, I had a hard time using grafana to look at Envoy metrics. I started noticing degradation at around 1k services/ingressroutes. At 3k, chrome would actually ask...
**Is this a BUG REPORT, PERFORMANCE REPORT or FEATURE REQUEST?**: FEATURE REQUEST **What happened**: During performance testing, we noticed that the discoverers were taking a long time to drain the...
While doing performance testing, I noticed that the discovery components are wasting cycles trying to delete services/endpoints that no longer exist. This is currently impacting other tests (which can be...