Add reusable shared buffer for transactions

[aroradaman]

This PR introduces a shared reusable buffer for transactions within a realNFTables instance.
This improves performance by reducing the memory allocations for running nft transactions.

[linux-foundation-easycla]

The committers listed above are authorized under a signed CLA.

• ✅ login: aroradaman / name: Daman Arora (841e0a5)

[danwinship]

_ No description provided. _

It's always good to provide a PR description. For the benefit of people in the future trying to figure out why stuff was done

[danwinship]

you shouldn't need to do this... make test should work anyway... what problem where you having?

[aroradaman]

I was planning to point to my fork in k/k and then run pull-kubernetes-e2e-gci-gce-nftables for testing

[aroradaman]

will drop the dummy commit after testing.

[danwinship]

ok, you probably just need to use the replace section of go.mod in the k/k PR. (The existing replacements are all just for the staging repos, but you can add a line like:

sigs.k8s.io/knftables => github.com/aroradaman/knftables TAG-OR-HASH

(and then run go mod tidy if you specified a hash, to make it convert it to a fake tag)

[aroradaman]

This + ./hack/update-vendor.sh did the trick.

[aroradaman]

k/k MR for testing the change by trigerringpull-kubernetes-e2e-gci-gce-nftables - kubernetes/kubernetes#125003

[aroradaman]

k/k MR for testing the change by trigerringpull-kubernetes-e2e-gci-gce-nftables - kubernetes/kubernetes#125003

pull-kubernetes-e2e-gci-gce-nftables is not stable right now - kubernetes/test-infra#32485 skipping this test

[danwinship]

There's a kind-based nftables job too... um... it's probably /test pull-kubernetes-e2e-kind-nftables but if that's wrong the error message will give you the whole list of tests and you can find it there. (If you had modified pkg/proxy/nftables it would run that test automatically, but it won't do it if you only modified knftables.)

There's an nftables scalability job too, and it would be cool to see if this ends up visible in the results... https://gist.github.com/aojea/276d1f8061403b6687ee12b2f2c7522e explains (part of) how to analyze the prometheus dumps...

[aroradaman]

There's a kind-based nftables job too... um... it's probably /test pull-kubernetes-e2e-kind-nftables but if that's wrong the error message will give you the whole list of tests and you can find it there. (If you had modified pkg/proxy/nftables it would run that test automatically, but it won't do it if you only modified knftables.)

There's an nftables scalability job too, and it would be cool to see if this ends up visible in the results... https://gist.github.com/aojea/276d1f8061403b6687ee12b2f2c7522e explains (part of) how to analyze the prometheus dumps...

pull-kubernetes-e2e-kind-nftables passed for k/k.
Will keep any eye on the scalability job.

[aroradaman]

@danwinship do you want to ship this with k/k 1.31?

[danwinship]

yes, let's get this in
/approve
/lgtm

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: aroradaman, danwinship

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [danwinship]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[aroradaman]

Should I point the knftables dependency in k/k to this commit, or are we cutting a release?

[danwinship]

Should wait for a release.
Bumping the dep in k/k is slightly annoying (need a deps approver), so we don't do it for every commit. Should wait to see if we're going to need any more fixes here first...