Cloudron V7.6.3 : impossible to install or update discourse
-
Hi all,
we have problems on multiple VM with Discourse and Cloudron V7.6.3 (i need to try on previous Cloudron versions). We can't install or update it.
Do you have the same problem ? it seems a problem with postresql, i don't know.
Some logs :
2024-01-10T11:38:07.000Z /app/code/vendor/bundle/ruby/3.2.0/gems/rack-mini-profiler-3.1.0/lib/patches/db/pg.rb:110:in `exec' 2024-01-10T11:38:07.000Z /app/code/vendor/bundle/ruby/3.2.0/gems/rack-mini-profiler-3.1.0/lib/patches/db/pg.rb:110:in `exec' 2024-01-10T11:38:07.000Z /app/code/vendor/bundle/ruby/3.2.0/gems/rake-13.0.6/exe/rake:27:in `<top (required)>' 2024-01-10T11:38:07.000Z /app/code/vendor/bundle/ruby/3.2.0/gems/rake-13.0.6/exe/rake:27:in `<top (required)>' 2024-01-10T11:38:07.000Z /home/cloudron/rbenv/versions/3.2.1/bin/bundle:25:in `<main>' 2024-01-10T11:38:07.000Z /home/cloudron/rbenv/versions/3.2.1/bin/bundle:25:in `<main>' 2024-01-10T11:38:07.000Z /home/cloudron/rbenv/versions/3.2.1/bin/bundle:25:in `load' 2024-01-10T11:38:07.000Z /home/cloudron/rbenv/versions/3.2.1/bin/bundle:25:in `load' 2024-01-10T11:38:07.000Z == 20200117141138 UpdatePostReplyIndexes: migrating =========================== 2024-01-10T11:38:07.000Z ActiveRecord::StatementInvalid: PG::FeatureNotSupported: ERROR: DROP INDEX CONCURRENTLY must be first action in transaction 2024-01-10T11:38:07.000Z Caused by: 2024-01-10T11:38:07.000Z Caused by: 2024-01-10T11:38:07.000Z PG::FeatureNotSupported: ERROR: DROP INDEX CONCURRENTLY must be first action in transaction 2024-01-10T11:38:07.000Z PG::FeatureNotSupported: ERROR: DROP INDEX CONCURRENTLY must be first action in transaction 2024-01-10T11:38:07.000Z StandardError: An error has occurred, all later migrations canceled: 2024-01-10T11:38:07.000Z rake aborted! 2024-01-10T11:38:17.000Z => Healtheck error: Error: Timeout of 7000ms exceeded 2024-01-10T11:38:27.000Z => Healtheck error: Error: Timeout of 7000ms exceeded 2024-01-10T11:38:37.000Z => Healtheck error: Error: Timeout of 7000ms exceeded 2024-01-10T11:38:47.000Z => Healtheck error: Error: Timeout of 7000ms exceeded 2024-01-10T11:38:57.000Z => Healtheck error: Error: Timeout of 7000ms exceeded 2024-01-10T11:39:03.000Z => Healtheck error: Error: connect EHOSTUNREACH 172.18.17.83:80
-
Yes we are debugging this already. It appears that the latest Cloudron update, which includes an update to the postgres addon, produces this problem with discourse.
If possible revert affected Cloudrons back to the previous Cloudron version until we know what the actual issue is and be able to create a fix.
-
-
We have at least isolated the issue to a certain postgres extension, which causes the side-effect for some database migration scripts within discourse. The upstream issue to get more info for hopefully some potential fix is https://github.com/tensorchord/pgvecto.rs/issues/253
Until that we have stopped delivering the latest Cloudron version for the moment. For anyone who has discourse installed and faces the issue, please comment here and hopefully we find a quick workaround.
-
Hi,
do you know when i will be fix ? -
The release is mostly ready. Should happen tomorrow.
-
@maksanse If you upgrade to 7.6.4 (we released yesterday night but as unstable, should be ok to install though), then discourse should be installable.
-
@girish said in Cloudron V7.6.3 : impossible to install or update discourse:
discourse should be installable.
Good to hear. What's the latest on making the official plugins work too? Thanks!
-
@jdaviescoates that's the next step.
-
Installation of Discourse works, but on first login, no email is sent to allow the login (nor received in spam) :
When trying the resend option, I get this message :
Is it something you are aware of / related to the issue you corrected ?
Note : Cloudron emails (Outbound with Amazon SES) work normally and should be accessible to Discourse :
Thank you for your help