Ori Shussman
Ori Shussman
@yzhao1012 I fixed all current comments
@yzhao1012 thanks, good comments! I addressed all current comments, if you have more please feel free to add them.
@yzhao1012 I fixed all new comments, except the one with the diff file you added. I think you ran an old version when you posted that file, as the diff...
Hey @MrSupiri, @avivzgroundcover and I have been working on it. This feature contains 3 large PRs into pixie, the last one is now under review. 6 new uprobes are required...
When developing eBPF observability solutions, we take a look many versions backwards, to see our solution works for a lot of versions, to make sure its stable enough. This is...
Yes this does the trick, thanks! Any reason for this not to be the default behaviour? It is the default in docker.
Yes. As discussed in https://github.com/siderolabs/talos/issues/5318, `/sys/kernel/tracing` is not mounted.
I believe that this issue and https://github.com/siderolabs/talos/issues/5318 are almost the same, and the placement could be mistaken. The point is that when `/sys/kernel/tracing` is not mounted, `tracefs` eBPF probes can...
> hey @orishuss have you tried `sum_samples` with `dedup_interval` set? it basically should do the same Thanks @AndrewChubatiuk, this seems exactly like what I want. However, I tried adding a...
I looked at pprofs and they don't look too different with vs. without dedup. While dedup causes very high cpu, it's also very spiky - see image, left part is...