You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@openwhisk.apache.org by OpenWhisk Team Slack <ra...@apache.org> on 2019/02/27 07:17:17 UTC

[slack-digest] [2019-02-26] #general

2019-02-26 00:32:02 UTC - Dominic Kim: BTW, how does IBM operate such a huge cluster? Are there any differences in the code base?
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551141122153900?thread_ts=1551090439.083100&cid=C3TPCAQG1
----
2019-02-26 00:35:11 UTC - Rodric Rabbah: @Sam Hjelmfelt i think this might fix jenkins
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551141311154400
----
2019-02-26 00:35:12 UTC - Rodric Rabbah: <https://github.com/apache/incubator-openwhisk/pull/4304>
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551141312154600
----
2019-02-26 00:54:27 UTC - Sam Hjelmfelt: great
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551142467155400?thread_ts=1551142467.155400&cid=C3TPCAQG1
----
2019-02-26 01:58:47 UTC - Rodric Rabbah: And it’s pushed <https://hub.docker.com/r/openwhisk/controller/tags>
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551146327155700?thread_ts=1551142467.155400&cid=C3TPCAQG1
----
2019-02-26 04:23:48 UTC - Vikash: Has anyone encountered this error? Please let me know!
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551155028156200?thread_ts=1551155028.156200&cid=C3TPCAQG1
----
2019-02-26 10:10:18 UTC - Satwik Kolhe: @Dominic Kim I just read the content of : incubator-openwhisk-deploy-kube/helm/openwhisk/values-metadata.yaml - It describes all the values that are used in helm installation.

Thought to share with you!
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551175818156500?thread_ts=1550668906.000700&cid=C3TPCAQG1
----
2019-02-26 10:25:01 UTC - Rodric Rabbah: Wow. That’s new to me. Did you try running the make target again?
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551176701157400?thread_ts=1551155028.156200&cid=C3TPCAQG1
----
2019-02-26 11:13:01 UTC - Satwik Kolhe: While deploying OW with Helm: Is the job.batch/install-packages job failing (terminating because of exceeded execution time) for this image openwhisk/ow-utils:38e4b05
Because for the image openwhisk/ow-utils:9f66633 - That same job is running to completion.

Is this happening for everyone else or just me?
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551179581159900?thread_ts=1551179581.159900&cid=C3TPCAQG1
----
2019-02-26 11:14:28 UTC - Satwik Kolhe: Following these events :

Events:
  Type     Reason            Age   From            Message
  ----     ------            ----  ----            -------
  Normal   SuccessfulCreate  16m   job-controller  Created pod: enfn-ow-3-install-packages-67zfz
  Normal   SuccessfulDelete  65s   job-controller  Deleted pod: enfn-ow-3-install-packages-67zfz
  Warning  DeadlineExceeded  65s   job-controller  Job was active longer than specified deadline
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551179668160000?thread_ts=1551179581.159900&cid=C3TPCAQG1
----
2019-02-26 13:37:35 UTC - Dave Grove: was it taking a long time to pull the new image perhaps?  If that is the case, trying again might be enough to sidestep the problem (image will already be on the cluster)
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551188255160200?thread_ts=1551179581.159900&cid=C3TPCAQG1
----
2019-02-26 13:41:19 UTC - Satwik Kolhe: Nope. The install-packages, is stuck in init, I don't think it's due to image pull. The myTask.sh init process that's defined in install-packages-cm that is defined as a job has an issue. I tried running that job with deadline of 1800 seconds (30mins) still the job did not complete normally.
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551188479160400?thread_ts=1551179581.159900&cid=C3TPCAQG1
----
2019-02-26 13:41:55 UTC - Satwik Kolhe: My setup is on Ubuntu kernel 4.4 , with weve net and took ceph
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551188515160600?thread_ts=1551179581.159900&cid=C3TPCAQG1
----
2019-02-26 14:10:34 UTC - Dave Grove: do you have any logs from the install-packages job?
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551190234160800?thread_ts=1551179581.159900&cid=C3TPCAQG1
----
2019-02-26 14:15:34 UTC - Satwik Kolhe: Let me spin-up another cluster, I will share kubectl describe output
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551190534161000?thread_ts=1551179581.159900&cid=C3TPCAQG1
----
2019-02-26 14:18:40 UTC - Satwik Kolhe: This is my landscape, you can see the install-packages is stuck in Init:0/1
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551190720161200?thread_ts=1551179581.159900&cid=C3TPCAQG1
----
2019-02-26 14:19:27 UTC - Dave Grove: you don’t have any invokers
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551190767161600?thread_ts=1551179581.159900&cid=C3TPCAQG1
----
2019-02-26 14:20:06 UTC - Dave Grove: <https://github.com/apache/incubator-openwhisk-deploy-kube#initial-setup>
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551190806161800?thread_ts=1551179581.159900&cid=C3TPCAQG1
----
2019-02-26 14:20:45 UTC - Dave Grove: the install-packages job is waiting for the controller to report there is at least 1 healthy invoker, but if you don’t label nodes as described in initial setup, you will never have an invoker
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551190845162000?thread_ts=1551179581.159900&cid=C3TPCAQG1
----
2019-02-26 14:24:58 UTC - Satwik Kolhe: Oh my goodness! How i missed those details. But while deploying this cluster using helm, my values.yaml looks proper. Invoker should have been created. Let me check what's happening.
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551191098162200?thread_ts=1551179581.159900&cid=C3TPCAQG1
----
2019-02-26 14:25:19 UTC - Satwik Kolhe: I appreciate you help here. My bad, will get back with this.
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551191119162400?thread_ts=1551179581.159900&cid=C3TPCAQG1
----
2019-02-26 14:28:02 UTC - Dave Grove: np.  there is a labeling step that is done outside of the helm chart.  Maybe it is worth trying to add an init container for the invoker that would obviously fail if there were zero invoker nodes and we are using the DockerContainerFactory.
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551191282162600?thread_ts=1551179581.159900&cid=C3TPCAQG1
----
2019-02-26 14:29:25 UTC - Satwik Kolhe: You are right, It is missing from my labels. I missed that step in this setup!
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551191365163000?thread_ts=1551179581.159900&cid=C3TPCAQG1
----
2019-02-26 14:29:32 UTC - Satwik Kolhe: Thanks a lot!
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551191372163200?thread_ts=1551179581.159900&cid=C3TPCAQG1
----
2019-02-26 16:12:37 UTC - James Thomas: Is it possible for a web action implemented using a sequence have one of the actions in the sequence exit early AND control the HTTP response?
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551197557164800
----
2019-02-26 16:12:47 UTC - James Thomas: I thought this was possible but now I’m not sure….
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551197567165100
----
2019-02-26 16:14:25 UTC - James Thomas: actions can throw errors or return rejected promises to stop sequence execution but it seems there’s no way to control the returned HTTP reponse.
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551197665166100
----
2019-02-26 16:16:08 UTC - Rodric Rabbah: if you’re not using composer, you can do it with one of the “Combinator” actions… but you should use composer :wink:
+1 : James Thomas, mmarth
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551197768166600
----
2019-02-26 16:16:27 UTC - Vikash: @Rodric Rabbah Yes. Still throws the same error.
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551197787166700?thread_ts=1551155028.156200&cid=C3TPCAQG1
----
2019-02-26 16:19:40 UTC - James Thomas: figured it out :slightly_smiling_face:
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551197980167400
----
2019-02-26 16:20:00 UTC - James Thomas: need to use ` return { error: { statusCode: 401, body: { error: err.message } } }` to exit sequence AND include custom content response
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551198000168000
----
2019-02-26 16:25:17 UTC - Rodric Rabbah: ah - right if you control the action then that works :smile:
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551198317168400
----
2019-02-26 20:45:45 UTC - Perry Dykes: Folks, for Cloud Functions wondering how the action require-whisk-auth attribute and the header value X-Require-Whisk-Auth related to the Managed API x-ibm-client-id value? Are they correlated or not is what I'm wondering?
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551213945171000?thread_ts=1551213945.171000&cid=C3TPCAQG1
----
2019-02-26 22:43:55 UTC - Ali Tariq: i faced a similar error yesterday - but that was because i did not have npm installed and apparently, its required in the latest version.
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551221035171200?thread_ts=1551155028.156200&cid=C3TPCAQG1
----
2019-02-26 22:44:55 UTC - Ali Tariq: so, maybe there is something else was also added! - not sure though
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551221095171600?thread_ts=1551155028.156200&cid=C3TPCAQG1
----
2019-02-26 23:19:42 UTC - Carlos Santana: Not correlated 
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1551223182172000?thread_ts=1551213945.171000&cid=C3TPCAQG1
----