Rodrigo Campos

@AkihiroSuda can you provide some context on why to limit the size? That struct includes the OCI state, that AFAIK does not have a size limit, right? (annotations are arbitrary...

Oh, I'm fine with a `SHOULD` addition. I'm not sure if the limit should be 4096 or what, though. In an ideal world, I'd like to see stats of containers...

@AlexeyPerevalov the runtime spec PR was already merged (https://github.com/opencontainers/runtime-spec/pull/1143). It would be great if you can finish this PR now :)

@AlexeyPerevalov No need to wait for a runtime-spec tag, IMHO. We are not using a tag already: https://github.com/opencontainers/runc/blob/a776ec9c90d7a67a924211d6cb1b38620afa6f49/go.mod#L15 For the x/sys pkg, I think it makes sense to define ourselves...

@AlexeyPerevalov but it seems all test have passed... Was that maybe before the latest fix merged yesterday to fix the CI? (although IIRC it was only about one of those...

@AlexeyPerevalov friendly ping? It would be great to have this in for runc 1.2. (otherwise, we will have to wait for another runc release, that takes a while, some more...

@caboteria I think what you might be missing is the two sources of truths that open up and threading, mentioned [here][link1] and [here][link2]. Unless you think it's possible without bringing...

The thing is, 1. implies two sources of truths probably and threading. And the others (2 and 3) are too high level to see how they will work with MetalLB....

> Currently the ip field already has 2 sources of truth No. The user can only specify IPs that live in MetalLB pools. User specified just selects which IP within...

> I'm not getting the "two sources of truth" problem. The external IPAM system would be canonical and MetalLB would not be canonical so metallb would be responsible for making...