airship-irc-bot | <rishabh.k.jain> Hi Team, This is the ps to update sonobuoy version to 0.51. (Needs some more core reviews) https://review.opendev.org/c/airship/airshipctl/+/782614 | 14:33 |
---|---|---|
airship-irc-bot | <ma257n> Hello Team, Could i get review for my PS for multus as a function . https://review.opendev.org/c/airship/treasuremap/+/802560 Thanks in advance | 16:21 |
airship-irc-bot | <suresh.lakshmipathy> Hi Team, 'runner' container in AIAP has been failing and am attaching the runner log. Log statement from status-checker container --> artifact-setup: <SUCCESS> infra-builder: <SUCCESS> runner: <FAILED> | 17:41 |
airship-irc-bot | <mattmceuen> @aodinokov I took a quick look at this, and I think it is somehow related to your recent secret generation change. Would this line of aiap be causing problems? I saw you removed a similar line in your change: https://github.com/airshipit/airshipctl/blob/master/tools/airship-in-a-pod/runner/assets/entrypoint.sh#L45 | 17:42 |
airship-irc-bot | <mattmceuen> Hmm, I also see sops:v0.1.0 in the output above, I think it should be v0.3.0 now, right? | 17:44 |
airship-irc-bot | <mattmceuen> ```+ ./tools/deployment/23_generate_secrets.sh + echo 'Generating secrets using airshipctl' + FORCE_REGENERATE=all + airshipctl phase run secret-update Generating secrets using airshipctl ... document filtered by selector [Group="airshipit.org", Version="v1alpha1", Kind="Phase", Name="secret-update"] found no documents``` This^ is the relevant bit of log output... seems to be saying that there's no secret-update phase, which would | 17:45 |
airship-irc-bot | be strange | 17:45 |
airship-irc-bot | <sirajudeen.yasin> https://review.opendev.org/c/airship/airshipctl/+/806992/3/manifests/phases/phases.yaml#209 | 17:53 |
airship-irc-bot | <sirajudeen.yasin> @suresh.lakshmipathy, are you using the latest code ? | 17:54 |
airship-irc-bot | <suresh.lakshmipathy> yes | 17:54 |
airship-irc-bot | <mattmceuen> @suresh.lakshmipathy I just tested with the latest, cherry picking in Ian's two AIAP related changes, and I didn't get the issue - `8/8 Running` | 18:49 |
airship-irc-bot | <mattmceuen> Are you applying the `examples/airshipctl` kustomization? | 18:50 |
airship-irc-bot | <sirajudeen.yasin> @suresh.lakshmipathy, I get the same error.... can you get into runner container and check if it uses latest airshipctl here /opt/aiap-artifacts/airshipctl ( git log . ) , i see it is using sep 20 PS | 18:56 |
airship-irc-bot | <sirajudeen.yasin> sorry, nvm.. got confused with git log date.. looks good | 18:59 |
airship-irc-bot | <mattmceuen> no error for you @sirajudeen.yasin? | 19:00 |
airship-irc-bot | <sirajudeen.yasin> i get same error as suresh.. still checkign the runner logs | 19:03 |
airship-irc-bot | <mattmceuen> are you cherrypicking Ian's changes @sirajudeen.yasin https://review.opendev.org/c/airship/airshipctl/+/810114 https://review.opendev.org/c/airship/airshipctl/+/810088 | 19:04 |
airship-irc-bot | <sirajudeen.yasin> ah! no .. let me apply ... but it should have failed at early stage without these PSs | 19:06 |
airship-irc-bot | <sirajudeen.yasin> applied those PSs, creating images from local changes.. will try to run kustomize build from examples/airshipctl and will try applying again | 19:13 |
airship-irc-bot | <suresh.lakshmipathy> yes, I am | 19:29 |
airship-irc-bot | <sirajudeen.yasin> same issue, now looks like the manifests path pulled by airshipctl does not have latest code from Runner container root@airship-in-a-pod:/opt/aiap-artifacts/manifests/airshipctl# airshipctl config get-manifests dummy_manifest: inventoryRepositoryName: "" metadataPath: manifests/site/test-site/metadata.yaml phaseRepositoryName: primary repositories: primary: auth: type: none checkout: | 20:06 |
airship-irc-bot | branch: master commitHash: "" force: false localBranch: false tag: "" fetch: {} url: https://opendev.org/airship/airshipctl targetPath: /opt/aiap-artifacts/manifests | 20:06 |
airship-irc-bot | <sirajudeen.yasin> root@airship-in-a-pod:/opt/aiap-artifacts/manifests/airshipctl# git rev-parse HEAD 885187f37c42ec8dd6c244fde04f8301cd6fdb5a | 20:06 |
airship-irc-bot | <sirajudeen.yasin> was expecting latest code of airshipctl as per the config, but got old code via "airshipctl document pull" | 20:09 |
airship-irc-bot | <ih616h> @sirajudeen.yasin you probably need to set `NO_CHECKOUT` = true | 20:29 |
airship-irc-bot | <ih616h> I'm betting that the manifests are already pulled, but airshipctl isn't checking out to the latest commit | 20:30 |
airship-irc-bot | <ih616h> *that should be NO_CHECKOUT = false | 20:30 |
airship-irc-bot | <ih616h> hang on, I'm not sure aiap has the ability to specify that variable yet... | 20:32 |
airship-irc-bot | <sirajudeen.yasin> is NO_CHECKOUT=true used to test current PS instead of a master checkout ? | 20:33 |
airship-irc-bot | <ih616h> https://github.com/airshipit/airshipctl/blob/master/tools/airship-in-a-pod/runner/assets/entrypoint.sh#L57-L61 | 20:34 |
airship-irc-bot | <ih616h> if you've specified a ref, it will do a checkout | 20:35 |
airship-irc-bot | <ih616h> by default, there's no ref specified, so no checkout | 20:35 |
airship-irc-bot | <ih616h> You could probably simply specify `AIRSHIP_CONFIG_PHASE_REPO_REF = master` to force a checkout | 20:39 |
airship-irc-bot | <sirajudeen.yasin> oh! | 20:39 |
airship-irc-bot | <ih616h> although I haven't tested anything like that, so I'm not 100% | 20:40 |
airship-irc-bot | <ih616h> https://github.com/airshipit/airshipctl/blob/master/tools/airship-in-a-pod/examples/airshipctl/replacements.yaml#L37 | 20:40 |
airship-irc-bot | <sirajudeen.yasin> ok | 20:41 |
airship-irc-bot | <mattmceuen> I'm behind on this -- was that caching that was causing the surprise Ian? | 20:41 |
airship-irc-bot | <ih616h> I believe so | 20:41 |
airship-irc-bot | <mattmceuen> gotcha | 20:41 |
airship-irc-bot | <sirajudeen.yasin> still wondering from where the old code came from.. oh! from the host path mounted.. from old pull | 20:41 |
airship-irc-bot | <sirajudeen.yasin> ok | 20:42 |
airship-irc-bot | <ih616h> yes, the idea being that future doc pulls can be skipped if you already have them | 20:42 |
airship-irc-bot | <sirajudeen.yasin> let clean the /opt/aiap-artifacts/ in host and try again | 20:42 |
airship-irc-bot | <sirajudeen.yasin> got it | 20:42 |
airship-irc-bot | <sirajudeen.yasin> thanks Ian | 20:43 |
airship-irc-bot | <ih616h> that would also probably fix it. If `master` works as a "default" value to the REF, we should use that, since it will avoid this issue in the future. | 20:43 |
airship-irc-bot | <ih616h> I'm running a test with the above now... | 20:43 |
airship-irc-bot | <ih616h> Forgot this hasn't merged yet... Could I please get some reviews on https://review.opendev.org/c/airship/airshipctl/+/810088 ? Fixes a bug in AIAP | 20:52 |
airship-irc-bot | <sirajudeen.yasin> now all looks good.. it passed that stage of secret-update failure. | 20:56 |
airship-irc-bot | <aodinokov> ```SOPS_PGP_FP``` isn't needed anymore | 20:56 |
airship-irc-bot | <sirajudeen.yasin> proceeding with other phases.. cleaning the airshipctl manifest repo from host path ( /opt/aiap-artifacts ) seems to have solved the issue.. thanks Ian | 20:57 |
airship-irc-bot | <aodinokov> correct, for some reason it can't find that phase... not sure why. do you have a separate phases for aiap? | 20:58 |
airship-irc-bot | <mattmceuen> Random question - I'm using capd (just by hand, not via airshipctl), and after I do a successful clusterctl move into the workload cluster, the cluster object is stuck in "Provisioning" status, and the kubeadm controller says `Reconciliation is paused for this object" "cluster"="capi-quickstart" "name"="capi-quickstart-control-plane" "namespace"="default" "reconciler group"="controlplane.cluster.x-k8s.io" "reconciler | 22:07 |
airship-irc-bot | kind"="KubeadmControlPlane"` | 22:07 |
airship-irc-bot | <mattmceuen> However, neither the Cluster nor anything else I can think to look at has a pause field or label set. Anyone know any other conditions under which kubeadm reconciliation might be paused? | 22:08 |
airship-irc-bot | <awander> clusterctl will actually set `cluster.Spec.Paused` to true during the move and the set it to false after sucessful completion of the move. | 22:20 |
airship-irc-bot | <awander> Can you check that field in your cluster. If it is true, then something likely went wrong during the move. | 22:20 |
airship-irc-bot | <mattmceuen> the field was missing, and I tried adding it in as `false`, but that didn't jumpstart it | 22:21 |
airship-irc-bot | <awander> are you looking at the CAPI `Cluster` object or the `DockerCluster`? | 22:22 |
airship-irc-bot | <mattmceuen> I checked both, but set the field on the CAPI one | 22:22 |
airship-irc-bot | <awander> what version of CAPI are you using? clusterctl? | 22:23 |
airship-irc-bot | <mattmceuen> The kubeadmcontrolplane code has ```if annotations.IsPaused(cluster, kcp) { log.Info("Reconciliation is paused for this object") return ctrl.Result{}, nil }``` So it seems like it would be looking for a pause annotation, but I couldn't find that either. v1alpha4 CAPI, clusterctl v0.4.2 | 22:25 |
airship-irc-bot | <mattmceuen> Think an upgrade to clusterctl v0.4.3 might help? | 22:25 |
airship-irc-bot | <awander> So clusterctl move did not report any errors ? | 22:26 |
airship-irc-bot | <awander> What was the output? | 22:26 |
airship-irc-bot | <mattmceuen> Nope, this was all: ```[madgin@leviathan:~/code/jzd]$ clusterctl move -n default --to-kubeconfig ./capi-quickstart.kubeconfig Performing move... Discovering Cluster API objects Moving Cluster API objects | 22:27 |
airship-irc-bot | Clusters=1 Creating objects in the target cluster Deleting objects from the source cluster New clusterctl version available: v0.4.2 -> v0.4.3 https://github.com/kubernetes-sigs/cluster-api/releases/tag/v0.4.3``` | 22:27 |
airship-irc-bot | <awander> Hm. looks like it ran cleanly. | 22:28 |
airship-irc-bot | <awander> Clusterctl move does not pause individual objects like KCP, so I'm not sure why that is paused. | 22:29 |
airship-irc-bot | <mattmceuen> yep, exactly. The move isn't critical to what I'm trying to do, so I blew it away and started over. I can recreate it if you think of anything to poke at though. | 22:31 |
airship-irc-bot | <awander> what steps are you following? url or capd readme? | 22:32 |
airship-irc-bot | <mattmceuen> It was the capi quickstart, although it stopped short of doing a move: https://cluster-api.sigs.k8s.io/user/quick-start.html | 22:33 |
airship-irc-bot | <awander> So, you did the move after the CNI install step. | 22:36 |
airship-irc-bot | <mattmceuen> yep | 22:36 |
airship-irc-bot | <mattmceuen> Lol, I've reached my docker rate limit for the day, so I guess it's quitting time for now -- thanks for your help Arvinder | 22:36 |
airship-irc-bot | <awander> Np. I would try it one more time and see if you hit the same issue. | 22:37 |
airship-irc-bot | <mattmceuen> cool, will definitely do that tomorrow | 22:37 |
airship-irc-bot | <awander> LMK if you want to walk through it tomorrow. | 22:37 |
airship-irc-bot | <mattmceuen> will do, thanks! | 22:37 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!