Loomio

Loomio instance at https://codema.in. Loomio is a collaborative decision-making platform (hence we're using the name codema) where users can initiate discussions and put up proposals.

Hosting

We are now on a DEV1-S instance virtual cloud server with the following specs: Moved to this server, by Dec 2019, to consolidate all groups related to FSCI and Indian Pirates from loomio.org and codema.fsci.org.in to a single loomio instance.

• 2 x86-64 bit Cores
• 2 GB Memory
• 20GB SSD Disk
• 100 Mbit/s Bandwidth
• 1 Reserved IP (v4)
• €5.58/Month

Coordination

• Hangout with us in our Matrix room #codema:poddery.com

Server Access

Maintained in a private git repo at -> https://git.fosscommunity.in/community/access

Deployment

We use loomio-deploy for deployment with only the following diff (because we have other services running on the server we proxy via nginx):

git diff
diff --git a/docker-compose.yml b/docker-compose.yml
index 0be5300..62a7f26 100644
--- a/docker-compose.yml
+++ b/docker-compose.yml
@@ -8,8 +8,8 @@ services:
     networks:
       - main
     ports:
-      - 80:80
-      - 443:443
+      - 127.0.0.1:8088:80
+      - 127.0.0.1:4443:443
     volumes:
       - ./nginx.conf:/etc/nginx/conf.d/custom.conf:ro
       - certs:/etc/nginx/certs

We use Akshay's mail server for sending outbound mails. We have the following in .env

ALLOW_ROBOTS=1

FEATURES_EXPLORE_PUBLIC_GROUPS=1

History

Before moving the service to current server, we were on a DEV1-S instance virtual cloud server with the following specs:

• 2 x86-64 bit Cores
• 2 GB Memory
• 50GB SSD Disk
• 200 Mbit/s Bandwidth
• 1 Reserved IP (v4)
• €3.99/Month

We were initially on a VC1-S instance virtual cloud server with the following specs until 31/01/2019.

• 2 x86-64 bit Cores
• 2GB Memory
• 50GB SSD Disk
• 1 Reserved Public IPv4
• 200Mbit/s Unmetered bandwidth
• €3.99/Month

On 01/02/2019 our server was taken down by Scaleway quoting payment issues. Payment was failing even after updating the credit card details and following that our server got deleted without proper notifications from Scaleway's side. Screenshot of the email from Scaleway attached below. Fortunately we were provided with a snapshot of the server from which we were able to recover codema to a new server. The service was moved to

Loomio Deployment History

Initially loomio was deployed and updated on codema.fsci.org.in using this official guide - https://github.com/loomio/loomio-deploy.

Then, for a while loomio was deployed and updated on codema.in using the same official guide but with difference on the fact that compose file used nginx-proxy and additional parameter of VIRTUAL_HOST_ALIAS in the env file. The VIRTUAL_HOST_ALIAS is for redirecting www.codema.in queries to codema.in. We used the smtp server from nemo.libreinfra.org for sending outbound mails.

The deployment from 2023 is documented above

Codema data migration process

For migration of loomio groups from both loomio.org and codema.fsci.org.in, the group data was exported using data export options provided by loomio. Changed the url links in the json data to point towards https://codema.in. The resulting json files was imported into the new codema instance after setting up the docker service as described here.

Codema recovery process

Here's a brief description of how codema was recovered after the server take down on 01/01/2019:

Under 'Snapshots' tab in the Scaleway dashboard we were provided with the snapshot (backup) of our codema server. A system image was created from this snapshot and it was used to create a new server with similar specifications. We lost our public IP along with the old server, so a new IP was assigned to the server and then updated the DNS A record of codema.fsci.org.in to point to this new IP. Once the server was up loomio was restarted using the following commands from the loomio installation directory:

 docker-compose down
 docker-compose up -d

The logs were checked for errors using the following command:

 docker-compose logs -f

Loomio wasn't getting started saying the port 25 was already in use. So the application using that port (which was exim4 in this case) was killed and loomio was restarted again.

An archive of the old loomio page is available at Loomio Archive.