Toby Lawrence

Nice, yeah, that makes sense. I'll have a chat with the team. I can see the appeal of just reading the file every single time we want to build an...

From your description, it looks like there's actually three different things going on here: - you're seeing disk buffer errors at startup after pods are restarted - you want to...

> We saw recently a similar issue. Our vector pods (on 0.32.0 distroless) sending data to Elasticsearch sink started being killed with OOM, and then entering a crash restart loop....

Hmm, seems like there's a good chance this is related to https://github.com/DataDog/saluki/pull/67 and the comment in `ddsketch.rs`: ```rust // I still don't yet understand how this works, since you'd think...

If you're still seeing the problem on fresh installs of 0.32, with fresh disk buffers, then I would agree that it's not related to compatibility issues of the disk buffer...

As far as the question around changes to the Protocol Buffers definition -- and I haven't look at the changes yet -- I _think_ it should be fine to use...

@maxboone Apologies. I swore I had previously approved this, but it seems not! LGTM overall, but it still needs a changelog entry. You can see the README for the changelog...