[slack-digest] [2019-11-14] #general

[OpenWhisk Team Slack <ra...@apache.org>]

2019-11-14 02:55:32 UTC - Dominic Kim: @Michele Sciabarra I briefly looked into your PR. It seems you added a new exec.
Then we might need a corresponding change in `runtime.json` as well?
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573700132284300
----
2019-11-14 02:56:13 UTC - Dominic Kim: not quite sure as of now.
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573700173284600
----
2019-11-14 02:56:44 UTC - Dominic Kim: It would be great to look into the controller logs
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573700204285200
----
2019-11-14 03:01:43 UTC - Dominic Kim: hm... it seems not.
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573700503285400
----
2019-11-14 03:01:58 UTC - Dominic Kim: This will override the default runtime manifest.
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573700518285800
----
2019-11-14 03:02:00 UTC - Dominic Kim: <https://github.com/apache/openwhisk-runtime-swift/pull/106/files#diff-2988bb892f18f2f78ad197a6984e3bc0R62>
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573700520286100
----
2019-11-14 03:10:30 UTC - Dominic Kim: I am curious whether anyone facing the orphaned veth issue or not.
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573701030286700
----
2019-11-14 03:11:01 UTC - Dominic Kim: Sometimes, we are observing some orphaned veths on invoker machines and it even breaks the docker network occasionally.
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573701061287300?thread_ts=1573701061.287300&cid=C3TPCAQG1
----
2019-11-14 03:42:54 UTC - Rodric Rabbah: @Sven Lange-Last might know 
I recall times when these occurred in production  
+1 : Dominic Kim
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573702974288700?thread_ts=1573701061.287300&cid=C3TPCAQG1
----
2019-11-14 07:30:35 UTC - Michele Sciabarra: @Dominic Kim regarding the new exec it is how the rest is done there are some zip for convenience of testing
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573716635289700
----
2019-11-14 07:31:04 UTC - Michele Sciabarra: it is possible to remove and use the build.sh script that builds it all but AFTER you build the docker images
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573716664290500
----
2019-11-14 07:31:59 UTC - Michele Sciabarra: it is kinda of an egg and chicken problem
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573716719291100
----
2019-11-14 07:35:10 UTC - Michele Sciabarra: @Dominic Kim of course we should add the new swift 5.1 runtime.json but I still have to pass the CI tests and there is “something” breaking them that is not related to my changes
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573716910293500
----
2019-11-14 07:35:46 UTC - Michele Sciabarra: What you think of this CI failure? <https://travis-ci.org/apache/openwhisk-runtime-swift/builds/611378033?utm_source=github_status&amp;utm_medium=notification>
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573716946293600?thread_ts=1573304037.237300&cid=C3TPCAQG1
----
2019-11-14 07:42:17 UTC - Dominic Kim: @Michele Sciabarra I couldn't get the issue.
Your change will build the new runtime first before running CI tests.
Did you figure out what was the main issue to make controller not be able to run?
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573717337295300
----
2019-11-14 07:42:40 UTC - Michele Sciabarra: @Dominic Kim the issue is: the CI setup.sh that COULD build the zip running ./build.sh but still needs the runtime that is not yet built, then it builds the runtime AND runs the tests requiring the executables that are not yet built, this is why I think Carlos added the executables ;  in my Go runtime the tests also builds the executable before running the tests but this code predates the actionloop thing and there is still the swift 4.1 not using actinoloop - if we drop 4.1 in favor of 4.2 I could also get rid of those prebuilt binaries
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573717360295500
----
2019-11-14 07:47:43 UTC - Michele Sciabarra: @Dominic Kim no I did not solve the issue of the CI yet
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573717663298300?thread_ts=1573717663.298300&cid=C3TPCAQG1
----
2019-11-14 07:52:07 UTC - Dominic Kim: What I could not understand is, swift5.1 will be built before deploying OW. So the image `action-swift-v5.1` is supposed to be available.
And you already added the runtime manifest configuration for it as well.
So I think OW is supposed to run with a new exec as expected.
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573717927298500?thread_ts=1573717663.298300&cid=C3TPCAQG1
----
2019-11-14 07:52:46 UTC - Dominic Kim: At this point, what is the eggs and chicken problem here?
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573717966298700?thread_ts=1573717663.298300&cid=C3TPCAQG1
----
2019-11-14 07:52:59 UTC - Dominic Kim: I think I miss something.
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573717979298900?thread_ts=1573717663.298300&cid=C3TPCAQG1
----
2019-11-14 09:20:30 UTC - Ajoy Agarwal: Just like Cloudant, can any other database be used to file openwhisk triggers? we are using MySql and MongoDB? If a new records is inserted in these DBs, I want a particular action of OpenWhisk to be fired. A quick response will be very helpful.
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573723230299100?thread_ts=1573723230.299100&cid=C3TPCAQG1
----
2019-11-14 10:10:36 UTC - Dominic Kim: If there is any database that supports "webhooks", that is the perfect one.
But when I tried to find one, there were not many choices.
One option can be running a service that receives change-events from the DB and invokes a corresponding action/trigger.
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573726236299200?thread_ts=1573723230.299100&cid=C3TPCAQG1
----
2019-11-14 12:02:36 UTC - Carlos Santana: I would recommend to look into using CDC to push changes from the db to kafka, then use the kafka OpenWhisk provider to fire your OpenWhisk triggers <https://blog.clairvoyantsoft.com/mysql-cdc-with-apache-kafka-and-debezium-3d45c00762e4>
+1 : Dominic Kim
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573732956301000?thread_ts=1573723230.299100&cid=C3TPCAQG1
----
2019-11-14 12:59:18 UTC - Ajoy Agarwal: Thanks for the information
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573736358301300?thread_ts=1573723230.299100&cid=C3TPCAQG1
----
2019-11-14 23:57:23 UTC - Upkar Lidder: Hello, I am trying to use `wskdebug` with `IBM Cloud Functions`.

Step 1: My action is very simple
```
function main(args) {
  if (args &amp;&amp; args.name) {
    console.log(`hello ${args.name}`);
    return { msg: `hello ${args.name}` };
  } else {
    console.log(`hello world`);
    return { msg: `hello world` };
  }
}
```
Step 2: created the action on IBM Cloud
`ibmcloud fn action create hello-debug index.js`

Step 3: started `wskdebug`, but I get an error when it tried to create the temporary Agent action. I don’t need the LiveReload support.

```
$ wskdebug hello-debug index.js -v
Starting debugger for /_/hello-debug
Starting local debug container wskdebug-hello-debug-1573772149542
docker run -d --name wskdebug-hello-debug-1573772149542 --rm -m 268435456 -p 8080 -p 9229:9229 -v "/Users/ulidder/Documents/icloud-documents/code-upkar/temp/openwhisk/simple:/code" ibmfunctions/action-nodejs-v10:1.13.0 node --expose-gc --inspect=0.0.0.0:9229 app.js
cf95f1d200e860c91aa179ef40f747c0065b998a2acbe1b914c18c33f4db36f9
Mounting sources onto local debug container: /Users/ulidder/Documents/icloud-documents/code-upkar/temp/openwhisk/simple/index.js
2019-11-14T22:55:50.728771670Z Debugger listening on <ws://0.0.0.0:9229/5c06b464-72f7-45a4-b0f8-0c74fc9deaf2>
2019-11-14T22:55:50.728824972Z For help, see: <https://nodejs.org/en/docs/inspector>
Installing agent in OpenWhisk (concurrency)...
Original action backed up at hello-debug_wskdebug_original.
{ OpenWhiskError: PUT <https://us-south.functions.cloud.ibm.com/api/v1/namespaces/_/actions/hello-debug?overwrite=true> Returned HTTP 400 (Bad Request) --&gt; "The request content was malformed:
requirement failed: concurrency 200 exceeds allowed threshold of 1"
    at Client.handleErrors (/usr/local/lib/node_modules/@adobe/wskdebug/node_modules/openwhisk/lib/client.js:216:11)
    at params.then.catch.err (/usr/local/lib/node_modules/@adobe/wskdebug/node_modules/openwhisk/lib/client.js:147:58)
    at processTicksAndRejections (internal/process/task_queues.js:86:5)
  name: 'OpenWhiskError',
  message:
   'PUT <https://us-south.functions.cloud.ibm.com/api/v1/namespaces/_/actions/hello-debug?overwrite=true> Returned HTTP 400 (Bad Request) --&gt; "The request content was malformed:\nrequirement failed: concurrency 200 exceeds allowed threshold of 1"',
  error:
   { code: 'df940ccf1d076f103c3743685c25d2b2',
     error:
      'The request content was malformed:\nrequirement failed: concurrency 200 exceeds allowed threshold of 1' },
  statusCode: 400 }

Restoring action
```

How do I fix `concurrency 200 exceeds allowed threshold of 1`? Does this mean IBM Cloud is expecting a value of 1 for the concurrency limit value and wskdebug is sending 200? Thank you for your help.
https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1573775843304300?thread_ts=1573775843.304300&cid=C3TPCAQG1
----