microshift icon indicating copy to clipboard operation
microshift copied to clipboard

mDNS: Avoid listening on internal interfaces and reduce footprint

Open mangelajo opened this issue 3 years ago • 5 comments

mDNS was listening on internal ovn interfaces (pods) and allocating way too much memory per server, regular mDNS queries fit on unfragmented regular 1500 MTU packets.

Closes: https://issues.redhat.com/browse/USHIFT-264

Which issue(s) this PR addresses:

Closes #<Issue Number>

mangelajo avatar Aug 10 '22 12:08 mangelajo

I updated the patch according to comments

mangelajo avatar Aug 10 '22 14:08 mangelajo

@zshi-redhat I verified and we don't send duplicate responses although we are listening on both br-ex and eth0:

image

mangelajo avatar Aug 10 '22 14:08 mangelajo

updated with the additional comments.

mangelajo avatar Aug 10 '22 15:08 mangelajo

@ggiguash @zshi-redhat does it look ok to lgtm ?

mangelajo avatar Aug 11 '22 09:08 mangelajo

/lgtm

ggiguash avatar Aug 11 '22 09:08 ggiguash

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: ggiguash, mangelajo

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~~ [ggiguash,mangelajo]

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

openshift-ci[bot] avatar Aug 11 '22 09:08 openshift-ci[bot]

/retest-required

Remaining retests: 2 against base HEAD 9466daf6475e3921ec59754d44226a509d03e6ec and 8 for PR HEAD f7542feb6cb63ce81827f146c846efed1491f48b in total

openshift-ci-robot avatar Aug 11 '22 11:08 openshift-ci-robot

@mangelajo: all tests passed!

Full PR test history. Your PR dashboard.

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. I understand the commands that are listed here.

openshift-ci[bot] avatar Aug 11 '22 12:08 openshift-ci[bot]

@zshi-redhat I verified and we don't send duplicate responses although we are listening on both br-ex and eth0:

image

@mangelajo I observed two reply packets when tcpdump on the physical interface (e.g. enp1s0):

23:10:37.644628 52:54:00:c7:7e:34 > 01:00:5e:00:00:fb, ethertype IPv4 (0x0800), length 74: 192.168.122.250.mdns > 224.0.0.251.mdns: 0 A (QM)? my-nginx.local. (32)
23:10:37.644902 52:54:00:c7:7e:34 > 01:00:5e:00:00:fb, ethertype IPv4 (0x0800), length 84: 192.168.122.250.mdns > 224.0.0.251.mdns: 0*- [0q] 1/0/0 A 192.168.122.250 (42)
23:10:37.644988 52:54:00:c7:7e:34 > 01:00:5e:00:00:fb, ethertype IPv4 (0x0800), length 84: 192.168.122.250.mdns > 224.0.0.251.mdns: 0*- [0q] 1/0/0 A 192.168.122.250 (42)

Which interface did you capture the mdns packet?

zshi-redhat avatar Aug 15 '22 03:08 zshi-redhat