2019-11-18 01:30:42 UTC - Dominic Kim: @Avik Aggarwal Hi. If you are using a self-signed certificate, you need to add `-i` option for all commands. https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574040642327200 ---- 2019-11-18 03:09:56 UTC - Avik Aggarwal: @Dominic Kim i did you use that but my issue is, I cannot use that when deploying through serverless. `sls deploy` doesn't take any `-i` parameters https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574046596328700?thread_ts=1574046596.328700&cid=C3TPCAQG1 ---- 2019-11-18 03:42:33 UTC - Dominic Kim: Hm.. not quite sure serverless support self-signed certificate or not. <https://github.com/serverless/serverless/issues/3958> https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574048553328800?thread_ts=1574046596.328700&cid=C3TPCAQG1 ---- 2019-11-18 03:42:47 UTC - Dominic Kim: Maybe @James Thomas ? https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574048567329000?thread_ts=1574046596.328700&cid=C3TPCAQG1 ---- 2019-11-18 03:44:00 UTC - Avik Aggarwal: Is it possible to provide custom certs instead of self signed in openwhisk. In the helm chart i did try to customize it under `ingress`keys. But it still went ahead with self signed ones https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574048640329200?thread_ts=1574046596.328700&cid=C3TPCAQG1 ---- 2019-11-18 03:46:14 UTC - Dominic Kim: If your certificate is not signed by a trusted CA, I think the same issue will come. https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574048774329400?thread_ts=1574046596.328700&cid=C3TPCAQG1 ---- 2019-11-18 03:47:38 UTC - Avik Aggarwal: I have got a signed one. I use those in prod. They are wildcard certs. I have configured `apihost` to be matching domain too. https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574048858329600?thread_ts=1574046596.328700&cid=C3TPCAQG1 ---- 2019-11-18 03:56:31 UTC - Dominic Kim: Then it should work. https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574049391329800?thread_ts=1574046596.328700&cid=C3TPCAQG1 ---- 2019-11-18 03:56:49 UTC - Dominic Kim: Still facing the same issue? https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574049409330000?thread_ts=1574046596.328700&cid=C3TPCAQG1 ---- 2019-11-18 03:57:19 UTC - Avik Aggarwal: Yes https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574049439330200?thread_ts=1574046596.328700&cid=C3TPCAQG1 ---- 2019-11-18 03:57:28 UTC - Avik Aggarwal: Same error https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574049448330400?thread_ts=1574046596.328700&cid=C3TPCAQG1 ---- 2019-11-18 03:57:56 UTC - Dominic Kim: Can you use `wsk` cli without `-i` option? https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574049476330600?thread_ts=1574046596.328700&cid=C3TPCAQG1 ---- 2019-11-18 03:58:40 UTC - Avik Aggarwal: Yes did that.. I get certificate expired or invalid error https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574049520330800?thread_ts=1574046596.328700&cid=C3TPCAQG1 ---- 2019-11-18 03:59:18 UTC - Dominic Kim: Basically, if you can use `wsk` without `-i` option against your local setup, `sls` also should work. https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574049558331000?thread_ts=1574046596.328700&cid=C3TPCAQG1 ---- 2019-11-18 03:59:48 UTC - Dominic Kim: If you cannot use it, I think that would be the starting point to fix the issue. https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574049588331200?thread_ts=1574046596.328700&cid=C3TPCAQG1 ---- 2019-11-18 04:00:02 UTC - Dominic Kim: How did you configure your cerft? https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574049602331400?thread_ts=1574046596.328700&cid=C3TPCAQG1 ---- 2019-11-18 04:27:47 UTC - Avik Aggarwal: I have this as my values.yaml for helm charts ``` ingress: apiHostName: "<http://serverless.mydomain.com|serverless.mydomain.com>" apiHostPort: 31001 apiHostProto: "https" type: NodePort annotations: <http://nginx.ingress.kubernetes.io/proxy-body-size|nginx.ingress.kubernetes.io/proxy-body-size>: "50m" domain: "*.<http://mydomain.com|mydomain.com>" tls: enabled: true secretenabled: true createsecret: true secretname: "ow-ingress-tls-secret" secrettype: "type" crt: "/file/to.crt" key: "/file/to/key" # Production deployments _MUST_ override these default auth values auth: system: "abcd...." guest: "pqrs..." systemNameSpace: "/whisk.system" ``` https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574051267332200?thread_ts=1574046596.328700&cid=C3TPCAQG1 ---- 2019-11-18 04:46:44 UTC - Avik Aggarwal: `crt` and `key` keys are being used to configure the certificates https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574052404332400?thread_ts=1574046596.328700&cid=C3TPCAQG1 ---- 2019-11-18 05:18:27 UTC - Avik Aggarwal: any pointers what could be wrong here? https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574054307332600?thread_ts=1574046596.328700&cid=C3TPCAQG1 ---- 2019-11-18 06:12:17 UTC - Ajoy Agarwal: When we take OpenWhisk in production, there may be a need for multiple servers to be deployed for hosting the application. Is it architecturally possible to scale out openwhisk production deployment using load balancers ? Is there any recommended distributed architecture which we can follow for large scale deployment of various components? https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574057537335200?thread_ts=1574057537.335200&cid=C3TPCAQG1 ---- 2019-11-18 09:38:48 UTC - Bill Zong: I’m working on this. The PR is not ready yet. And I found it hard to perfect the feature in one step. So I’ll just move a little forward.
Here’s my solution: 1. make a scheduler for handling the (invoker) cluster scaling. 2. make a joiner and a deleter for actual processing, which can be done in any language. I’ve done some of its features in golang. 3. flexible configure by using YAML config file. It could be done in deploying with helm. And simply re-config the ConfigMap and restart the scheduler would take much lesser work. https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574069928335300?thread_ts=1574057537.335200&cid=C3TPCAQG1 ---- 2019-11-18 10:58:09 UTC - Mohammed AbuAisha: Hi Everyone, I was checking how to deploy OpenWhisk to Mesos and I only found this github link here <https://github.com/apache/openwhisk-deploy-mesos> and I have some questions: 1. Can I build the universe for OpenWhisk on the same of Mesos machine ? 2. Will be running the following script here <https://github.com/apache/openwhisk-deploy-mesos/blob/master/dcos-universe/scripts/build.sh> enough to generate the universe of the OpenWhisk? 3. After generating the universe do I need to generate a docker image for that or not ? 4. I think after run the above script I can run the following but where should I run these commands, on which host (Cluster ?)?? `dcos marathon app add marathon.json` `dcos package repo add --index=0 dev-universe <http://universe.marathon.mesos:8085/repo>` 5. I also noticed to deploy the OpenWhisck there are some environment variables need to set before run the make command here <https://github.com/apache/openwhisk-deploy-mesos/blob/master/dcos-deploy-script/Makefile> 6. I'm not sure why do we need to have a record for Route53 here <https://github.com/apache/openwhisk-deploy-mesos/blob/master/dcos-deploy-script/Makefile#L2> ? 7. Can I run this in any cloud providers ? or this is just for AWS ? 8. I also noticed there is prefix url `<http://exhibitor-dcos>.` before each package, cannot we replace it with the Mesos URL since the package is already installed there ? 9. If we replace the `<http://exhibitor-dcos>.$(APIGATEWAY_DNS)` with cluster IP, would that work or not ? https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574074689341100?thread_ts=1574074689.341100&cid=C3TPCAQG1 ---- 2019-11-18 11:01:00 UTC - Christophe Jelger: Hello. Quick question: is it possible to have a trigger that fires when some action(s) is newly deployed, updated, or removed? Would also be super cool to be able to define this for a given package … https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574074860342400?thread_ts=1574074860.342400&cid=C3TPCAQG1 ---- 2019-11-18 15:18:44 UTC - Avik Aggarwal: @Dominic Kim any help sir! https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574090324342800?thread_ts=1574046596.328700&cid=C3TPCAQG1 ---- 2019-11-18 15:53:57 UTC - Logan Marks: Hey there, I’m having an issue trying to create an alarm. The response we receive is `{'error': 'Initialization has failed due to: There was an error uncompressing the action archive.'}` We’re working with a new cluster having set up openwhisk ourselves. Is this possibly related to how we set up the cluster? we’re hitting the `/api/v1/namespaces/whisk.system/actions/alarms/once` whisk if that makes a difference https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574092437343400?thread_ts=1574092437.343400&cid=C3TPCAQG1 ---- 2019-11-18 17:32:54 UTC - Rodric Rabbah: Have you tried to run an action for a zip file? You can check the activation logs to see if there’s additional details about the error which looks to me like an inability to unzip a file. https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574098374345000?thread_ts=1574092437.343400&cid=C3TPCAQG1 ---- 2019-11-18 17:33:49 UTC - Rodric Rabbah: Welcome @Tom Barber @Logan Marks @Archit Sud. Hello @Mohammed AbuAisha https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574098429345900 ---- 2019-11-18 17:34:12 UTC - Logan Marks: Yes we’ve seen that issue before without our own actions. The main problem here is that the alarm whisk it’s trying to run isn’t something we deployed ourself. It’s one that is included with openwhisk https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574098452346500?thread_ts=1574092437.343400&cid=C3TPCAQG1 ---- 2019-11-18 17:34:36 UTC - Rodric Rabbah: There isn’t something like that at user level. I can see how this could be useful. https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574098476347200?thread_ts=1574074860.342400&cid=C3TPCAQG1 ---- 2019-11-18 17:35:56 UTC - Rodric Rabbah: You should be able to run on any cloud but this particular rep might be specialized to aws and your mileage will vary. https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574098556348600?thread_ts=1574074689.341100&cid=C3TPCAQG1 ---- 2019-11-18 17:37:50 UTC - Rodric Rabbah: My guess was that it’s not specific to alarm. The issue has to do with running a zipped action which failed to unzip in your deployment. https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574098670350100?thread_ts=1574092437.343400&cid=C3TPCAQG1 ---- 2019-11-18 17:41:01 UTC - Tom Barber: thanks @Rodric Rabbah! https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574098861350400?thread_ts=1574098861.350400&cid=C3TPCAQG1 ---- 2019-11-18 17:46:45 UTC - Logan Marks: Well the failing zipped action would be the alarm `once` whisk in this case https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574099205350500?thread_ts=1574092437.343400&cid=C3TPCAQG1 ---- 2019-11-18 17:49:39 UTC - Logan Marks: Though when I retrieve the zip archive for `once` it seems to be a correctly formatted zip file https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574099379350700?thread_ts=1574092437.343400&cid=C3TPCAQG1 ---- 2019-11-18 17:51:45 UTC - Logan Marks: When I check the activations list it is indeed the `once` whisk that’s failing. Here’s the activation itself: ``` { "namespace": "whisk.system", "name": "once", "version": "0.0.1", "subject": "whisk.system", "activationId": "e9507f36c6c845e8907f36c6c845e8a1", "start": 1574091726956, "end": 1574091727060, "duration": 104, "statusCode": 0, "response": { "status": "action developer error", "statusCode": 0, "success": false, "result": { "error": "Initialization has failed due to: There was an error uncompressing the action archive." } }, "logs": [], "annotations": [ { "key": "path", "value": "whisk.system/alarms/once" }, { "key": "waitTime", "value": 3859 }, { "key": "kind", "value": "nodejs:6" }, { "key": "timeout", "value": false }, { "key": "limits", "value": { "concurrency": 1, "logs": 10, "memory": 256, "timeout": 600000 } }, { "key": "initTime", "value": 104 } ], "publish": false } ``` https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574099505350900?thread_ts=1574092437.343400&cid=C3TPCAQG1 ---- 2019-11-18 17:55:06 UTC - Mohammed AbuAisha: @Rodric Rabbah thanks for reply https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574099706351100?thread_ts=1574074689.341100&cid=C3TPCAQG1 ---- 2019-11-18 22:08:00 UTC - Rodric Rabbah: have you tried the standalone controller? unless you need compose or kube, you might want to use this path instead. https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574114880353600?thread_ts=1574098861.350400&cid=C3TPCAQG1 ---- 2019-11-18 22:08:25 UTC - Rodric Rabbah: <https://github.com/apache/openwhisk/blob/master/core/standalone/README.md> https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574114905353800?thread_ts=1574098861.350400&cid=C3TPCAQG1 ---- 2019-11-18 22:31:29 UTC - Tom Barber: it needs to move to Kubernetes, so I've deployed it on a local kubecluster to see if I get any further https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574116289354000?thread_ts=1574098861.350400&cid=C3TPCAQG1 ---- 2019-11-18 22:32:25 UTC - Tom Barber: ``` Serverless Error --------------------------------------- Failed to deploy API Gateway route due to error: POST <https://localhost:31001/api/v1/web/whisk.system/apimgmt/createApi.http?accesstoken=%22%22&spaceguid=23bc46b1-71f6-4ed5-8c54-816aa4f8c502> Returned HTTP 502 (Bad Gateway) --> "API creation failure: Unable to obtain API(s) from the API Gateway: "Invalid URI \"http:///$(ENERVATED-IBEX_APIGATEWAY_SERVICE_HOST):$(ENERVATED-IBEX_APIGATEWAY_SERVICE_PORT_API)/v2/23bc46b1-71f6-4ed5-8c54-816aa4f8c502/apis?limit=undefined&skip=undefined\""" ``` https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574116345354300 ---- 2019-11-18 22:32:38 UTC - Tom Barber: anyone want to hazard a guess as to what I've not set? :slightly_smiling_face: https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574116358354700 ---- 2019-11-18 22:38:55 UTC - Tom Barber: oh the helm name appears to actually be very important :wink: https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1574116735355000 ----
