2020-10-02 13:28:49 UTC - Emil Chitas: Hi all, I am struggling to use the 
`docker` container factory on EKS.
After deployment, the invoker daemon set creates the pods on all the labeled 
nodes and the invokers are registered by the controller:
https://openwhisk-team.slack.com/archives/C4J3R7JFL/p1601645329004000?thread_ts=1601645329.004000&cid=C4J3R7JFL
----
2020-10-02 13:33:22 UTC - Emil Chitas: There's no entry in the invoker logs as 
to an action being invoked.
https://openwhisk-team.slack.com/archives/C4J3R7JFL/p1601645602004500?thread_ts=1601645329.004000&cid=C4J3R7JFL
----
2020-10-02 13:40:11 UTC - Dave Grove: The docker container factory will only 
work if Kubernetes is really using docker (as opposed to containerd) as the 
underlying container runtime.  I don’t know what EKS does, but many cloud 
provided Kubernetes services do not use docker.
https://openwhisk-team.slack.com/archives/C4J3R7JFL/p1601646011004700?thread_ts=1601645329.004000&cid=C4J3R7JFL
----
2020-10-02 13:47:55 UTC - Emil Chitas: I would expect that, but what bugs me is 
that the invoker never receives any message to start a container
https://openwhisk-team.slack.com/archives/C4J3R7JFL/p1601646475004900?thread_ts=1601645329.004000&cid=C4J3R7JFL
----
2020-10-02 13:49:25 UTC - Dave Grove: i would assume that the invoker is not 
healthy.  because it was not able to use docker to create and run the health 
action.  Look in the controller logs to see how many healthy invokers it thinks 
it has
https://openwhisk-team.slack.com/archives/C4J3R7JFL/p1601646565005100?thread_ts=1601645329.004000&cid=C4J3R7JFL
----
2020-10-02 13:50:10 UTC - Dave Grove: if my guess is right that EKS doesn’t use 
docker, there will be 0 healthy invokers
https://openwhisk-team.slack.com/archives/C4J3R7JFL/p1601646610005300?thread_ts=1601645329.004000&cid=C4J3R7JFL
----
2020-10-02 13:50:48 UTC - Emil Chitas: ```
[2020-10-02T13:06:58.833Z] [INFO] [#tid_sid_invokerHealth] [] [InvokerActor] 
invoker3 is up
[2020-10-02T13:06:58.833Z] [INFO] [#tid_sid_invokerHealth] [] [InvokerPool] 
invoker status changed to 0 -> Healthy, 1 -> Healthy, 2 -> Healthy, 3 
-> Healthy```

https://openwhisk-team.slack.com/archives/C4J3R7JFL/p1601646648005500?thread_ts=1601645329.004000&cid=C4J3R7JFL
----
2020-10-02 13:51:31 UTC - Emil Chitas: that's the entry for one of the invokers 
in the controller log.  Wouldn't that mean the invoekr is registered as healthy?
https://openwhisk-team.slack.com/archives/C4J3R7JFL/p1601646691005700?thread_ts=1601645329.004000&cid=C4J3R7JFL
----
2020-10-02 13:53:55 UTC - Emil Chitas: but you are right nevertheless, the 
nodes use most likely containerd.
https://openwhisk-team.slack.com/archives/C4J3R7JFL/p1601646835005900?thread_ts=1601645329.004000&cid=C4J3R7JFL
----
2020-10-02 13:54:26 UTC - Dave Grove: it does look like the controller believes 
it has healthy invokers, which is suspicious.
https://openwhisk-team.slack.com/archives/C4J3R7JFL/p1601646866006100?thread_ts=1601645329.004000&cid=C4J3R7JFL
----
2020-10-02 13:55:07 UTC - Emil Chitas: exactly, that's my big surprise. I 
checked the nodes and they are definitely running containerd
https://openwhisk-team.slack.com/archives/C4J3R7JFL/p1601646907006300?thread_ts=1601645329.004000&cid=C4J3R7JFL
----
2020-10-02 13:56:54 UTC - Dave Grove: That sounds like a bug in the health 
protocol then.  Most likely on the invoker side in the docker-specific code, 
but could be anywhere I guess.
https://openwhisk-team.slack.com/archives/C4J3R7JFL/p1601647014006500?thread_ts=1601645329.004000&cid=C4J3R7JFL
----
2020-10-02 14:02:10 UTC - Emil Chitas: yes, most likely:
```[2020-10-02T13:06:53.322Z] [INFO] [#tid_sid_invokerHealth] [] 
[InvokerReactive]  [marker:invoker_activation_start:23693]
[2020-10-02T13:06:53.329Z] [WARN] [#tid_sid_invokerHealth] [] [InvokerReactive] 
revision was not provided for whisk.system/invokerHealthTestAction0
[2020-10-02T13:06:53.483Z] [INFO] [#tid_sid_invokerHealth] [] 
[CouchDbRestStore] [GET] 'test_whisks' finding document: 'id: 
whisk.system/invokerHealthTestAction0' [marker:database_getDocument_start:23854]
[2020-10-02T13:06:53.497Z] [INFO] [#tid_sid_invokerHealth] [] 
[CouchDbRestStore]  [marker:database_getDocument_finish:23868:14]
[2020-10-02T13:06:53.599Z] [INFO] [#tid_sid_invokerHealth] [] [ContainerPool] 
containerStart containerState: cold container: None activations: 1 of max 1 
action: invokerHealthTestAction0 namespace: whisk.system activationId: 
99847b6c63c04ba5847b6c63c05ba5d2 
[marker:invoker_containerStart.cold_counter:23970]
[2020-10-02T13:06:53.656Z] [INFO] [#tid_sid_invokerHealth] [] 
[DockerClientWithFileAccess] running /usr/bin/docker run -d --cpu-shares 11 
--memory 128m --memory-swap 128m --network bridge -e 
__OW_API_HOST=<http://openwhisk-nginx.ow.svc.cluster.local> -e 
__OW_ALLOW_CONCURRENT=false .................. --ulimit nofile=1024:1024 
--pids-limit 1024 --log-driver json-file openwhisk/action-nodejs-v10:nightly 
(timeout: 1 minute) [marker:invoker_docker.run_start:24027]
[2020-10-02T13:06:54.400Z] [INFO] [#tid_sid_invokerHealth] [] 
[DockerClientWithFileAccess]  [marker:invoker_docker.run_finish:24770:743]
[2020-10-02T13:06:54.415Z] [INFO] [#tid_sid_invokerHealth] [] 
[DockerClientWithFileAccess] running /usr/bin/docker inspect --format 
{{.NetworkSettings.Networks.bridge.IPAddress}} 
f108336bd3f03509bc36e94a958e18e8b523cf70e2b5636327f013b7b2d97d42 (timeout: 1 
minute) [marker:invoker_docker.inspect_start:24786]
[2020-10-02T13:06:54.536Z] [INFO] [#tid_sid_invokerHealth] [] 
[DockerClientWithFileAccess]  [marker:invoker_docker.inspect_finish:24907:121]
[2020-10-02T13:06:54.538Z] [INFO] [#tid_sid_invokerHealth] [] 
[DockerClientWithFileAccess] running /usr/bin/docker rm -f 
f108336bd3f03509bc36e94a958e18e8b523cf70e2b5636327f013b7b2d97d42 (timeout: 1 
minute) [marker:invoker_docker.rm_start:24907]
[2020-10-02T13:06:54.601Z] [INFO] [#tid_sid_invokerHealth] [] 
[CouchDbRestStore] [PUT] 'test_activations' saving document: 'id: 
whisk.system/99847b6c63c04ba5847b6c63c05ba5d2, rev: null' 
[marker:database_saveDocument_start:24971]
[2020-10-02T13:06:54.604Z] [INFO] [#tid_sid_dbBatcher] [] [CouchDbRestStore] 
'test_activations' saving 1 documents 
[marker:database_saveDocumentBulk_start:4907]
[2020-10-02T13:06:54.640Z] [INFO] [#tid_sid_invokerHealth] [] 
[MessagingActiveAck] posted combined of activation 
99847b6c63c04ba5847b6c63c05ba5d2
[2020-10-02T13:06:54.653Z] [INFO] [#tid_sid_dbBatcher] [] [CouchDbRestStore]  
[marker:database_saveDocumentBulk_finish:4957:48]
[2020-10-02T13:06:54.663Z] [INFO] [#tid_sid_invokerHealth] [] 
[CouchDbRestStore]  [marker:database_saveDocument_finish:25034:62]
[2020-10-02T13:06:54.861Z] [INFO] [#tid_sid_invokerHealth] [] 
[DockerClientWithFileAccess]  [marker:invoker_docker.rm_finish:25231:323]```
https://openwhisk-team.slack.com/archives/C4J3R7JFL/p1601647330006700?thread_ts=1601645329.004000&cid=C4J3R7JFL
----
2020-10-02 14:03:56 UTC - Emil Chitas: thank you for the clarifications
https://openwhisk-team.slack.com/archives/C4J3R7JFL/p1601647436007100?thread_ts=1601645329.004000&cid=C4J3R7JFL
----
2020-10-02 14:05:50 UTC - Dave Grove: ok if I grab that log section from the 
invoker to put in an issue?  We really should have detected that the health 
action didn’t run correctly
https://openwhisk-team.slack.com/archives/C4J3R7JFL/p1601647550007300?thread_ts=1601645329.004000&cid=C4J3R7JFL
----
2020-10-02 14:06:41 UTC - Emil Chitas: if you want, I can provide you the full 
logs from that try
https://openwhisk-team.slack.com/archives/C4J3R7JFL/p1601647601007500?thread_ts=1601645329.004000&cid=C4J3R7JFL
----
2020-10-02 14:07:41 UTC - Dave Grove: cool.  thanks.  if you don’t mind opening 
an issue and posting the log in there, it would help me remember to take a look 
at this.
+1 : Emil Chitas
https://openwhisk-team.slack.com/archives/C4J3R7JFL/p1601647661007700?thread_ts=1601645329.004000&cid=C4J3R7JFL
----
2020-10-02 14:07:54 UTC - Emil Chitas: will do, thanks
https://openwhisk-team.slack.com/archives/C4J3R7JFL/p1601647674008000?thread_ts=1601645329.004000&cid=C4J3R7JFL
----
2020-10-02 18:30:01 UTC - Lixiang Ao: makes sense. thank you!
https://openwhisk-team.slack.com/archives/C4J3R7JFL/p1601663401008400?thread_ts=1601494447.001000&cid=C4J3R7JFL
----

Reply via email to