vagrant: update 4.19 and net-next VM images#18496
Conversation
We have been hitting issues in our CI since yesterday on all `master` scheduled pipelines as well as `master`-based PR pipelines, with Cilium Agent seemingly being unable to reach a healthy state due to IPv6-related issues (see [1] and [2] below). I have a hunch the issue is due to the Vagrant VMs themselves, as we do not seem to hit the issue anymore on 4.9- and 5.4-based pipelines after merging an update to their Vagrant VMs in 3562520, containing an unrelated fix. This commit bumps the 4.19 and net-next VM images as well. [1] ``` 2022-01-13T05:42:17.179701398Z level=debug msg="Greeting failed" error="Get \"http://[fd04::12]:4240/hello\": dial tcp [fd04::12]:4240: connect: network is unreachable" host="http://[fd04::12]:4240" ipAddr="fd04::12" nodeName=k8s2 path="Via L3" subsys=health-server 2022-01-13T05:42:17.179730123Z level=debug msg="Failed to probe: Get \"http://[fd04::12]:4240/hello\": dial tcp [fd04::12]:4240: connect: network is unreachable" ipAddr="fd04::12" nodeName=k8s2 port=4240 subsys=health-server ``` [2] ``` 2022-01-14T12:12:52.181387748Z level=fatal msg="Error while creating daemon" error="failed to finalise LB initialization: failed to initialize NodePort addrs. : Failed to determine IPv6 of enp0s3 for NodePort" subsys=daemon ``` Signed-off-by: Nicolas Busseneau <nicolas@isovalent.com>
|
This is a re-open of #18493 with the proper branch name. As noted in the original PR, CI passed:
|
|
Forgot we need to wait for the runtime tests as well: https://jenkins.cilium.io/view/PR/job/Cilium-PR-Runtime-net-next/1026/ EDIT: but already we can tell by looking at the logs that it does not hit the agent startup issue either. Marking as ready to merge. |
|
Job 'Cilium-PR-Runtime-net-next' failed and has not been observed before, so may be related to your PR: Click to show.Test NameFailure OutputIf it is a flake, comment |
|
/mlh new-flake Cilium-PR-Runtime-net-next 👍 created #18497 |
Since the net-next version bump in cilium#18496, this test is consistently failing. Unfortunately, the net-next bump was necessary to fix a bigger issue where all pipelines were broken, so let us disable that test for now. Signed-off-by: Sebastian Wicki <sebastian@isovalent.com>
We have been hitting issues in our CI since yesterday on all
masterscheduled pipelines as well asmaster-based PR pipelines, with Cilium Agent seemingly being unable to reach a healthy state due to IPv6-related issues (see [1] and [2] below).I have a hunch the issue is due to the Vagrant VMs themselves, as we do not seem to hit the issue anymore on 4.9- and 5.4-based pipelines after merging an update to their Vagrant VMs in 3562520, containing an unrelated fix.
This commit bumps the 4.19 and net-next VM images as well.
[1]
[2]