Frederic Branczyk
Frederic Branczyk
I just did a small analysis of a Parca server's metastore that was running on a relatively small Kubernetes cluster for about 24 hours with lots of deployments happening. |...
Our storage does not allow multiple labels with the same name, so we should validate and reject requests that have duplicate label names.
It seems that this got dropped in the last refactoring of the normalization process. It happens a lot that a profile contains multiple samples with the same stacktrace. For storage...
The color scheme used to connotate improvements or regressions uses green and red and blue for neutral. Unfortunately, those colors are not particularly accessible, and given that about 300 million...
We refresh a number of components when an action in the UI is taken because the query parameters change triggering the top-level component that extracts the parameters and causes a...
Icicle graphs are the most popular way to explore performance data, but callgraphs can also be a very useful visualization when different stack traces call the same function for example.
The debuginfo component is currently doing the actual symbolization, while the symbolizer really just handles the interaction with the metastore. This should be decoupled for a better separation of concerns....
While Parca is capable of this, it is nowhere documented how to scrape for example an fgprof endpoint.
It is quite confusing to be able to query data but not be able to see any targets on the targets page. This is because the targets page currently only...