Frederic Branczyk

Here's the config we use in production: ``` scrape_configs: - job_name: pods scrape_interval: "3s" profiling_config: pprof_config: memory: keep_sample_type: - type: inuse_space unit: bytes process_cpu: enabled: false kubernetes_sd_configs: - role: pod...

By using `role: endpoints` instead of `role: pods` you can already discover targets via endpoints objects (which is what headless services create under the hood). Let me know if you...

Can we try to understand why this happens? As far as I can tell all codepaths that generate these inputs write a non-null value for all mapping fields if writing...

As in, this is the code that produces this: https://github.com/parca-dev/parca/blob/00471e8f7986c689f770e2507fede9e0d943323f/pkg/parcacol/arrow.go#L338-L354

Is there maybe something special about the conversion of an empty string?

For cpu profiling I would recommend to always profile using parca agent, it puts way less overhead on the workload so you can always do it. That said I think...

It’s normal that the agent data looks slightly different. It’s actually good, because 1) the runtime hides some things in the “native” pprof profilers, and 2) it uses sigprof which...

We should open an issue in frostdb to add support for binary operations in projections, no? Eg. `SELECT sum(period) / count(value) FROM stacktraces;`.

I think we never finished it as part of the average work: https://github.com/polarsignals/frostdb/pull/302 The idea was to turn an average aggregation into a sum and count aggregation and push a...

A couple of things to note: * We primarily recommend using Parca-Agent, it results in much less data needing to be stored (as only memory addresses are stored in object...