2020-03-05 07:54:26 UTC - Wouter: Hi all, we are interested in using OpenWhisk for one of our new projects. Since OpenWhisk is fairly new to us, we would like to find somebody to review/discuss our OpenWhisk plans with. Is there somebody available to do some consultancy on OpenWhisk for us within the Netherlands? https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583394866183400?thread_ts=1583394866.183400&cid=C3TPCAQG1 ---- 2020-03-05 08:54:08 UTC - Tom Barber: @Wouter you'd be best speaking with @Rodric Rabbah and the folks who do the development to see who they'd recommend. They are EDT though I believe so you'll probably have to wait a few hours +1 : Rodric Rabbah, Wouter https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583398448184200 ---- 2020-03-05 11:42:06 UTC - Rodric Rabbah: How can we help? https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583408526184500?thread_ts=1583394866.183400&cid=C3TPCAQG1 ---- 2020-03-05 13:09:55 UTC - Wouter: Hi @Rodric Rabbah we're starting on a new project and OpenWhisk feels like fitting solution. I'm convinced we'll get it up and running, but we would like to validate our design choices (architecture). The design of this architecture will evolve in the coming weeks. Is there any chance you're available to review our architecture/design when we finished it? https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583413795184900?thread_ts=1583394866.183400&cid=C3TPCAQG1 ---- 2020-03-05 14:22:57 UTC - Dave Grove: Committers: heads up when merging commits: Doing a ‘squash and merge’ of a PR through the GitHub UI is currently changing authorship of the merged commits. For now, I think we need to use ’rebase and merge` See <https://github.com/isaacs/github/issues/1750> https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583418177188000?thread_ts=1583418177.188000&cid=C3TPCAQG1 ---- 2020-03-05 15:31:34 UTC - Dave Grove: <https://github.com/kubernetes/test-infra/issues/16618> https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583422294188800?thread_ts=1583418177.188000&cid=C3TPCAQG1 ---- 2020-03-05 15:52:21 UTC - Rodric Rabbah: This is a new change? https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583423541189400 ---- 2020-03-05 15:53:34 UTC - Rodric Rabbah: Thanks for the heads up https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583423614191000 ---- 2020-03-05 16:00:39 UTC - Dave Grove: it stated happening yesterday, across all of github. Both of the PRs you and merged this morning to openwhisk-wskdebug got misattributed. https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583424039192000 ---- 2020-03-05 16:02:24 UTC - Dave Grove: on the plus side, I just learned that .asf.yaml now allows configuration of a few more things. Enables issues, and choice of which of merge/squash/rebase are enabled. So slowly becoming more self-service https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583424144193300 ---- 2020-03-05 16:24:38 UTC - Rodric Rabbah: ouch https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583425478193600 ---- 2020-03-05 17:09:25 UTC - Rodric Rabbah: Someone posted this:
> This is an intentional change that our team recently deployed and I'm > sorry to hear that's causing trouble. > When a user chooses to squash and merge a PR they are taking authorship > of the commit. If they were merging the PR and creating a merge commit > that merge commit would be attributed to them as author, so we wanted > to be consistent with that hence the change. We are also aware that we > should be setting the author as a co-author and we can see that's not happening > every time. https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583428165194000 ---- 2020-03-05 17:23:12 UTC - Bilal: Within the past month or so, I attempted to deploy using a different release name and it didn't work. Just adding a data point. https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583428992194100?thread_ts=1583336639.159600&cid=C3TPCAQG1 ---- 2020-03-05 17:27:30 UTC - Bilal: @Dominic Kim Has anyone written up anything about blue/green deployments on k8s? Is there a way to have both blue/green point to the same DB for actions? Or is there a mirroring or DB migration script? Or am I doing that? The double resources shouldn't be an issue. https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583429250194300?thread_ts=1583279584.123800&cid=C3TPCAQG1 ---- 2020-03-05 17:47:06 UTC - Bilal: I see that <https://github.com/apache/openwhisk/pull/4724|Implement an ElasticSearchActivationStore #4724> has been merged! :smiley: I assume if I change the tags in my values.yaml to the latest images I should be able to use this? Probably some more things to change in value.yaml, this is of course assuming that the helm chart can interact with this new feature. I'm more than happy to be the test subject and write up what I did after. https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583430426197100?thread_ts=1583430426.197100&cid=C3TPCAQG1 ---- 2020-03-05 18:10:23 UTC - Rodric Rabbah: in theory… yes :sweat_smile: Thanks @Bilal that’s be great to see https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583431823197500 ---- 2020-03-05 19:19:48 UTC - Bilal: ```whisk: ingress: type: NodePort apiHostName: 172.17.0.2 apiHostPort: 31001 invoker: containerFactory: impl: "kubernetes" options: CONFIG_whisk_activationStore_elasticsearch_protocol": "http" CONFIG_whisk_activationStore_elasticsearch_hosts": <http://172.18.0.1:9200> CONFIG_whisk_activationStore_elasticsearch_indexPattern": "openwhisk" CONFIG_whisk_activationStore_elasticsearch_username": "{{ db.elasticsearch.auth.admin.username }}" CONFIG_whisk_activationStore_elasticsearch_password": "{{ db.elasticsearch.auth.admin.password }}" CONFIG_whisk_spi_ActivationStoreProvider": "org.apache.openwhisk.core.database.elasticsearch.ElasticSearchActivationStoreProvider" nginx: httpsNodePort: 31001``` So after poking around a bit in the PR, I think I need to do something like this? As well as a section for the controller? I have a default ElasticSearch that I've spun up in docker. I don't think it has a username/password. Not sure on the Activation store provider sections. https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583435988200500?thread_ts=1583435988.200500&cid=C3TPCAQG1 ---- 2020-03-05 19:26:50 UTC - Kostis Kaffes: Hi all, can someone help me understand what happens when all Invokers run out of `user-memory`? Are activations sent to a random Invoker or just fail? https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583436410201700?thread_ts=1583436410.201700&cid=C3TPCAQG1 ---- 2020-03-05 20:10:11 UTC - Brendan Doyle: can we configure the docker image of standalone to run with the same configs as the standard jar? Trying to set up kafka to run and use our own manifest https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583439011202600?thread_ts=1583439011.202600&cid=C3TPCAQG1 ---- 2020-03-05 20:11:02 UTC - Rodric Rabbah: @Markus Thömmes surely will know better than me, iirc it’s not quite random in that there’s a home invoker per namespace and activations get queued on that invoker’s topic https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583439062202700?thread_ts=1583436410.201700&cid=C3TPCAQG1 ---- 2020-03-05 20:11:52 UTC - Rodric Rabbah: @chetanm ^^ https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583439112202900?thread_ts=1583439011.202600&cid=C3TPCAQG1 ---- 2020-03-05 20:12:18 UTC - Markus Thömmes: <https://github.com/apache/openwhisk/blob/05ed4e18eb53f5630735ffbe28726dce5a3481b5/core/controller/src/main/scala/org/apache/openwhisk/core/loadBalancer/ShardingContainerPoolBalancer.scala#L420-L424> https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583439138203100?thread_ts=1583436410.201700&cid=C3TPCAQG1 ---- 2020-03-05 20:12:20 UTC - Markus Thömmes: it's random :slightly_smiling_face: thankyou : Kostis Kaffes https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583439140203400?thread_ts=1583436410.201700&cid=C3TPCAQG1 ---- 2020-03-05 20:12:37 UTC - Rodric Rabbah: :partyparrot: https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583439157203600?thread_ts=1583436410.201700&cid=C3TPCAQG1 ---- 2020-03-05 21:09:46 UTC - Kostis Kaffes: Randomness makes sense to avoid overloading a specific Invoker. I guess that if all Invokers are unhealthy, an error is returned, right? https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583442586203900?thread_ts=1583436410.201700&cid=C3TPCAQG1 ---- 2020-03-05 23:00:17 UTC - Bilal: So after reading the ansible chart, it looks like a few values get set under both the invoker/controller when activation store backend gets set to ElasticSearch https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583449217204100?thread_ts=1583430426.197100&cid=C3TPCAQG1 ---- 2020-03-05 23:01:04 UTC - Bilal: But the logic to set those vars is not in the helm chart. This is just me thinking out loud, but I could set things manually in mycluster.yaml to set both the db backend and place the elastic search host, etc vars in the controller/invoker. https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583449264204300?thread_ts=1583430426.197100&cid=C3TPCAQG1 ---- 2020-03-05 23:04:18 UTC - Bilal: Searching the openwhisk-deloy-kube repo, I don't see anything about Activation store or db backend, so the chart might not understand this setting. https://openwhisk-team.slack.com/archives/C3TPCAQG1/p1583449458204600?thread_ts=1583430426.197100&cid=C3TPCAQG1 ----