Error executing 'postInstallation': Purging temp files

Hello,

I’m trying to run a non standard installation, but probably someone could get some insight into the cause of the error. Using docker-compose, running postgres+redis+sidekiq+main from separate Bitnami 2.3.9 images, and these do work perfectly fine on my Windows computer. But when deployed to AWS ECS, I’m getting an error on the main app in the middle of installation. Also the latest lines about bundling assets (list is longer) does not occur on my local Windows at all, on Windows it waits for long and then prints success messages.

postgre INFO Found PostgreSQL server listening at postgresql:5432
postgre INFO PostgreSQL server listening and working at postgresql:5432
discour INFO Preparing Discourse environment…
discour INFO Configuring the Discourse database files…
discour INFO Configuring Redis…
discour INFO Setting up SMTP in Discourse…
discour INFO Configuring Discourse database…
discour INFO Enabling PostgreSQL plugins for Discourse…
discour INFO Migrating database…
discour INFO Creating the Discourse admin user…
discour INFO Configuring your Discourse installation…
Error executing ‘postInstallation’: Purging temp files
Bundling assets
I, [2020-02-10T13:56:52.236240 #1088] INFO – : Writing /opt/bitnami/discourse/public/assets/vendor-3b2f6dcdff2179eecae36277e07dc69ded2872f619a6a2bd222bbfe69e9f28ae.js
I, [2020-02-10T13:56:58.189628 #1088] INFO – : Writing /opt/bitnami/discourse/public/assets/admin-e6fbf7cad5fdf706d3f174807bd8e6f8051c81dabdcb8c2c1cfae1ccee125eff.js

Unsupported install are, well, unsupported.

You can run a supported install by following our discourse/INSTALL-cloud.md at master · discourse/discourse · GitHub where it is completely with an external database like our guide Running Discourse with a separate PostgreSQL server.

3 Likes

That sounds good because main reason was to split database from the monolith. I have went through the config and bootstrap process, running Ubuntu linux (WSL) on Windows with a mounted Windows docker (as native docker does not work on WSL), all went nicely, image was bootstrapped, database updated and a load of other things on screen. Finally docker is started (I see it in docker ps, and ports are properly mapped), but when I try to access it on localhost (that’s how I configured it), I get a blank screen as server does not send any response. Same thing was working finely on the bitnami docker containers (application opening) which I could not get to run properly on AWS for some reason. Could I be missing something? I doubt that running through WSL should be an issue, as it is a real linux and bootstrapping went without errors.

And quite similar, installed on AWS EC2, bootstrapped, docker launched and it’s failing to work on port 80. Looks like this
[*******]# telnet localhost 80
Trying 127.0.0.1…
Connected to localhost.
Escape character is ‘^]’.
Connection closed by foreign host.
Connected to the docker inside EC2, in nginx logs I find such errors each second:
cannot load certificate “/shared/ssl/ec2-x-x-x-x.xx-xx.compute.amazonaws.com.cer”: PEM_read_bio_X509_AUX() failed (SSL: error:0909006C:PEM routines:get_name: no start line: Expecting: TRUSTED CERTIFICATE)
Redis and Postgres logs are nice. Application logs in /var/www/discourse/log also do not show anything special (though they’re very short).