Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
  • Oct 03 16:22
    dpadron closed #55
  • Oct 03 16:21
    dpadron closed #57
  • Oct 01 17:29
    dpadron opened #57
  • Sep 22 16:23
    blazhovsky closed #56
  • Sep 22 16:17
    mbocevski opened #56
  • Sep 22 16:17
    mbocevski review_requested #56
  • Sep 22 15:55
    blazhovsky closed #54
  • Sep 22 15:11
    mbocevski synchronize #54
  • Sep 22 14:45
    mbocevski closed #53
  • Sep 20 09:50
    dpadron opened #55
  • Sep 16 13:38
    mbocevski review_requested #54
  • Sep 16 13:38
    mbocevski opened #54
  • Jul 11 11:22
    nobodykid opened #53
  • May 27 14:51
    blazhovsky closed #51
  • May 27 13:59
    blazhovsky synchronize #51
  • May 20 09:34
    mbocevski closed #52
  • May 19 20:27
    blazhovsky review_requested #52
  • May 19 20:27
    blazhovsky opened #52
  • May 19 11:57
    mbocevski opened #51
  • May 19 11:57
    mbocevski review_requested #51
Marko Bocevski
@mbocevski
@SteveRJKnight it's part of the docs https://docs.ckan.org/en/2.9/maintaining/installing/install-from-source.html#install-the-required-packages, so maybe you just did it and forgot about it :)
mabah-mst
@mabah-mst
@mbocevski - Is this the proper forum for asking about https://github.com/keitaroinc/ckan-helm ?
I don't have any current questions, but I am having a look at it, and sometimes you can learn a lot by reading through tha forum/chat room
Marko Bocevski
@mbocevski
Yes you’re in the right place @mabah-mst
Lalitha Prasad Ch
@LalithaPrasadC1_twitter
anyone tried to integrate ldap plugin in 2.9.1 ?
Marko Bocevski
@mbocevski
@LalithaPrasadC1_twitter not really, but if you're using https://github.com/NaturalHistoryMuseum/ckanext-ldap it should work just fine
Lalitha Prasad Ch
@LalithaPrasadC1_twitter
@mbocevski , thanks for the reply. Able to login with ldap user.
What is the recommended configuration for Solr - HA ?.
Marko Bocevski
@mbocevski
Use solrcloud setup with at least two replicas when creating the solr collection for ckan
Lalitha Prasad Ch
@LalithaPrasadC1_twitter
can't we use local Kubernetes cluster with multiple pods?
Marko Bocevski
@mbocevski
Yes of course
Tory Clasen
@tclasen
Hello, @mbocevski and team. Is the intent to make https://github.com/keitaroinc/ckan-helm directly helm installable without having to clone it down and build locally? I see there is a branch for github-pages which looks like someone was making it so you could just do a helm repo add directly on the github repo. Are there instructions to do this otherwise? I'd like to make this chart a dependency for a chart I'm working on in local development.
1 reply
Marko Bocevski
@mbocevski
@tclasen Yes, we are working on having a helm repo available for our helm chart and we will also publish it as well. We should have that ready by end of the month.
Tory Clasen
@tclasen
Thanks!
mabah-mst
@mabah-mst
Hi - I had som installations of CKAN based on newest docker-ckan setup shut down on me last friday. the log includes " Brutally killing worker 1 (pid: 16)
I am wondering if this could be a result of a timeout in the polling specified in the dockerfile: HEALTHCHECK --interval=10s --timeout=5s --retries=5 CMD curl --fail http://localhost:5000/api/3/action/status_show || exit 1
Marko Bocevski
@mbocevski
@mabah-mst do you have any logs before uwsgi workers are killed? I don't think that the healthcheck failed, cause that would be odd or mean that something got stuck. I think that before the uwsgi worker is killed there has to be something more in the logs as to what transpired before
mabah-mst
@mabah-mst
It goes straight from regular polling to shutting down. The weird thing is that it happend in several environments, but seperated by a couple of hours. Also - the environmenets were running different versions of CKAN (2.9.1 and 2.9.2), packaged by different versions of docker-ckan
2021-02-12 21:01:21,973 INFO  [ckan.config.middleware.flask_app]  /api/3/action/                                                                                                                                   status_show render time 0.001 seconds
2021-02-12 21:01:32,912 INFO  [ckan.config.middleware.flask_app]  /api/3/action/                                                                                                                                   status_show render time 0.002 seconds
Brutally killing worker 1 (pid: 15)...
Fri Feb 12 21:01:36 2021 - stopping gevent signals watchers for worker 1 (pid: 1                                                                                                                                   5)...
Fri Feb 12 21:01:36 2021 - stopping gevent sockets watchers for worker 1 (pid: 1                                                                                                                                   5)...
Fri Feb 12 21:01:36 2021 - main gevent watchers stopped for worker 1 (pid: 15)..                                                                                                                                   .
goodbye to the gevent Hub on worker 1 (pid: 15)
Brutally killing worker 2 (pid: 16)...
Fri Feb 12 21:01:36 2021 - stopping gevent signals watchers for worker 2 (pid: 1                                                                                                                                   6)...
Fri Feb 12 21:01:36 2021 - stopping gevent sockets watchers for worker 2 (pid: 1                                                                                                                                   6)...
Fri Feb 12 21:01:36 2021 - main gevent watchers stopped for worker 2 (pid: 16)..
Marko Bocevski
@mbocevski
That doesn't look like it's the healthcheck and doesn't look like docker stopped it... Weird. Let me know if this happens again and would be good to turn on debug log on your ckan instance. Cause something definitely happened that caused uwsgi to do that, also it could be out of memory or similar issue
mabah-mst
@mabah-mst
OK - I will if it re-appears - how do I raise the log level to debug?

Ahh:

ckan config-tool ${APP_DIR}/production.ini -s logger_ckanext -e level=INFO && \

?

Marko Bocevski
@mbocevski
You can do it through .ckan-env - DEBUG=true
mabah-mst
@mabah-mst
OK, thanks
Is it me, or does it seem like the whole ## Logging configuration is undocumented in https://docs.ckan.org/en/2.9/maintaining/configuration.html#config-file
Marko Bocevski
@mbocevski
Yeah it's not really well documented
mabah-mst
@mabah-mst
Thanks again. One of theonly things in the docker-ckan setup I found a little puzzling was .env vs. .ckan-env

in the end I ended up merging the files to one which I load both implicitly by just being next to my docker-compose.yml file and explicitly inside the docker-compose file like

    env_file:
      - ./.env

It makes it easier to figure out what takes precedence when there is only one place to define it

then there is no precedence
Marko Bocevski
@mbocevski
It's in the README, .env is used when running docker-compose build and substitutes env variables that are in the docker-compose.yml. While .ckan-env is loaded at runtime when the containers start and envvars reads the config from environment variables that are set in the container based on the .ckan-env file.
robjqsinclair
@robjqsinclair
Hi I am new to ckan and kubernetes but I have deployed the keitaro helm script to google cloud but how to I see ckan? what IP address should I use on my local machine it it localhost:5000?. Do I need to set up a load balancer?
Marko Bocevski
@mbocevski
@robjqsinclair if you have an ingress load balancer on google cloud, then you can enable the ingress in the keitaro ckan helm chart and it will be accessible through the DNS that you provided on the ingress. Another way to access is doing a kubectl port-forward to the ckan service in kubernetes from your local machine, just to test that it is running.
robjqsinclair
@robjqsinclair
Hi @mbocevski, thanks for the kubectl port-ward tip it proved that it is running but I am having trouble setting up a load balancer on GCP and then connecting the ingress in the values.yaml. I am unclear how to setup the load balancer in GCP is it just a vanilla web load balancer and ingress I need, and what values in the values.yaml do I need to set?
Marko Bocevski
@mbocevski
@robjqsinclair for an Ingress Controller for google cloud, you can check the google cloud GKE documentation: https://cloud.google.com/kubernetes-engine/docs/concepts/ingress#:~:text=The%20GKE%20Ingress%20controller%20deploys,are%20deployed%20in%20the%20cluster.&text=Compute%20Engine%20firewall%20rules%20that,Front%20Ends%20or%20Envoy%20proxies., alternatively you can look into deploying your own like traefik or nginx-ingress. Once you have successfully set up an Ingress Controller in your GKE cluster, then you can update the helm release and set values for the ingress - https://github.com/keitaroinc/ckan-helm/blob/master/values.yaml#L191-L202. Basically you need to set enabled to true and provide your dns entry (that points to the ingress controller load balancer IP address) in the host field.
robjqsinclair
@robjqsinclair
Thanks @mbocevski that has helped me get it to work. My next question is how do I now add extensions into the ckan container? I am unsure where the container to be deployed are defined?
Marko Bocevski
@mbocevski
You can extend our base Docker image following a similar approach as described next our readme here: https://github.com/keitaroinc/docker-ckan#extending-ckan-docker-images. You specify your own custom Docker image in the helm chart values: https://github.com/keitaroinc/ckan-helm/blob/master/values.yaml#L8-L11
robjqsinclair
@robjqsinclair
Trying to build a docker file with a custom theme which is building but when none of the webassets are being built correctly, they work fine in development mode from source but when I pip wheel they do not, I am thinking it must be something to do with the build of the package?
8 replies
johnstacy
@johnstacy
any word on when CKAN will be able to get off of Solr v6?
1 reply
Lalitha Prasad Ch
@LalithaPrasadC1_twitter
I see ckan.sysadminApiToken: "", anyone tried this option?. how to test this one.?. i tried one REST API its showing
{"help": "http://localhost:5000/api/3/action/help_show?name=send_email_notifications", "error": {"__type": "Authorization Error", "message": "Access denied: Action send_email_notifications requires an authenticated user"}, "success": false}%
Lalitha Prasad Ch
@LalithaPrasadC1_twitter
any way to assign sysadminapitoken during the installation?
Marko Bocevski
@mbocevski
@LalithaPrasadC1_twitter yes, you need to configure your SMTP server settings for CKAN here https://github.com/keitaroinc/ckan-helm/blob/master/values.yaml#L129-L135. Then once you have CKAN up and running log in with a sysadmin account and generate an API token through the users profile and then put that token here https://github.com/keitaroinc/ckan-helm/blob/master/values.yaml#L63-L65 and update your helm release.
Lalitha Prasad Ch
@LalithaPrasadC1_twitter
@mbocevski , Oh after apikey generation from CKAN UI. Just for sending mails we need to configure sysadminapikeytoken here. we thought of removing manual step 'i.e. Create apikey for sysadmin'. Is there any REST API or cli to create user with apikey?
johnstacy
@johnstacy
@mbocevski any update on CKAN supporting latest Solr?
Juan Sebastián Henao
@jshenaop
Hi everyone, I am working into a personal template for CKAN using Theming. But after a while I cannot see DEBUG mode even if I configure .ckan-env variable DEBUG=true. Is there a special URL for this mode?
Second, also if I want to change the original language should I use ckan.locale_default = es into .ckan-env
Third, same for email settings? .ckan-env
1 reply
walorda
@walorda:matrix.org
[m]
Hi guys, as i am relatively new to docker and i am not able to fix the errors in the official documented docker-compose and with lack of flexibility i am considering to use this repo. but as i am new and not that experienced with docker how do i setup this docker env up with the existing ckan repo because i don't see any ckan files in this repo hmm?
1 reply
Pribadi HCR
@pribadihcr
Hi all, I am new in CKAN, I have tried to deploy CKAN in AKS using this tutorial https://www.keitaro.com/2021/05/01/creating-infrastructure-for-ckan-on-azure-using-terraform/ and https://www.keitaro.com/2021/03/28/how-to-deploy-ckan-on-kubernetes/ . If I want to modify the ckan's frontend template/theme, how we modify it?. Thanks
6 replies
Lochanie
@Lochanie
image.png
Hi , I am really new to CKAN and DOCKER as well. I tried to install CKAN using docker compose using this link: https://docs.ckan.org/en/2.9/maintaining/installing/install-from-docker-compose.html. But I have faced several problems. 1. docker exec ckan /usr/local/bin/ckan-paster --plugin=ckan datastore set-permissions -c /etc/ckan/production.ini | docker exec -i db psql -U ckan this command is not working for me. there is no file called production.ini in this location. it is in the location /var/lib/docker/volumes/docker_ckan_config/_data. I put this location in to the path. An vene is in the different location. I changed the command according to that and tried. but it was giving errors. 2. When I exit from the server, next time ckan is not working, I have to build every thing from the scratch. I am using ubuntu 18.04 AMI in AWS. Please help me to resolve this 2 issues. Thank you!
bhphuc87
@bhphuc87
Dear all, i'm working on ckanext-envvars. i get issues when adding new variables. the ckan version 2.9.3. Could you please help me. Thank you so much All
johnstacy
@johnstacy
i asked a couple months ago in here about ckan supporting the latest version of Solr. just wondering if that effort is any closer to completion...
Juan Sebastián Henao
@jshenaop

Hello there. Today i was able to make a CKAN_EXT but I need It in spanish. Shall I use CKAN.LOCALE_DEFAULT=es?

Whats the logic behind CAPITAL letters in CKANENV. I think dot is replaced by and rest in capital letters. am I OK?

2 replies