datapath/neighbor: Add ratelimit to desired neighbor calculation#43928
Merged
aanm merged 1 commit intocilium:mainfrom Mar 6, 2026
Merged
Conversation
Member
Author
|
/test |
008c660 to
7babedf
Compare
Member
Author
|
/test |
7babedf to
565ca3d
Compare
Member
Author
|
/test |
Member
Author
|
@cilium/sig-datapath any chance of getting a review 👋 ? |
smagnani96
approved these changes
Feb 18, 2026
Contributor
smagnani96
left a comment
There was a problem hiding this comment.
Code changes LGTM.
Left a comment on the interval: if you're seeking a feedback on the specific amount chosen (15sec), I'd ask you to pull in other reviewers.
Member
|
@ldelossa do you have bandwidth to help with @cilium/sig-datapath reviews? If not, you might consider either marking your profile as "busy" on GitHub or (temporarily) stepping out of the team until you have the bandwidth. |
joestringer
reviewed
Feb 18, 2026
joestringer
approved these changes
Feb 19, 2026
Member
joestringer
left a comment
There was a problem hiding this comment.
Some nuances and nits discussed above, but I'll leave it up to you if any of the ideas discussed in the threads are useful to integrate (either in this PR or a subsequent one).
joamaki
approved these changes
Mar 3, 2026
Member
|
(Ready but deferring to @dylandreimerink to resolve comments and apply the PR) |
a2839e2 to
076a22d
Compare
Member
Author
|
/test |
076a22d to
97d1773
Compare
Member
Author
|
/test |
brlbil
approved these changes
Mar 6, 2026
During investigation of a memory leak in v1.18, one of the pprof profiles showed a high amount of memory usage in `netlink/nl.(*NetlinkSocket).Receive`. cilium#41623 (comment) This is most likely due to a lack of rate limiting in the desired neighbor calculation which does a lot of netlink requests to get next hops. So this commit limits desired neighbor calculation to once every 15 seconds. In the worst case scenario where the default gateway changes, XDP might not be able to forward traffic for up to 15 seconds. Such a scenario should only happen when configuration changes are made or when the network topology changes, and thus this seems an acceptable tradeoff. Signed-off-by: Dylan Reimerink <dylan.reimerink@isovalent.com>
97d1773 to
b32a781
Compare
rastislavs
approved these changes
Mar 6, 2026
Member
Author
|
/test |
7 tasks
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
10 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
During investigation of a memory leak in v1.18, one of the pprof profiles showed a high amount of memory usage in
netlink/nl.(*NetlinkSocket).Receive.#41623 (comment)
This is most likely due to a lack of rate limiting in the desired neighbor calculation which does a lot of netlink requests to get next hops.
So this PR limits desired neighbor calculation to once every 15 seconds. In the worst case scenario where the default gateway changes, XDP might not be able to forward traffic for up to 15 seconds. Such a scenario should only happen when configuration changes are made or when the network topology changes, and thus this seems an acceptable tradeoff.