manage_vhost sometimes breaks and redis version update results a fatal error during deploy #93
Conversation
There was a problem hiding this comment.
Hey @stijnbernards! Thanks for your patches, much appreciated!
| @@ -37,13 +37,14 @@ public function configureWithTaskConfig(TaskConfigurationInterface $config): ?Ta | |||
|
|
|||
| task('deploy:nginx', [ | |||
There was a problem hiding this comment.
So now it first places nginx files and after that it runs HMV. Sounds OK to me, but what if the vhost doesn't exist on the node yet? Have you tested that scenario as well?
There was a problem hiding this comment.
Not yet, but I believe the deploy creates the www.shop.nl folders in the nginx directory aswell, not sure if creating a new vhost will then remove this. Let me see if I can test this somewhere :)
There was a problem hiding this comment.
Could you verify this or drop this part of the patch? Then we can merge!
There was a problem hiding this comment.
I'll try and verify this change soon but it's the most important fix of this MR. I'll create a new MR to fix the hypernode settings
There was a problem hiding this comment.
Just verified this change and it does infact fix my issue when creating an entirely new vhost during deploys when it does not exist yet. Same goed for the deploy:varnish:activate step as this step uses the current folder
|
You have some merge conflicts, can you update your branch? Then we can get this merged for release :) |
Done :) |
This MR solves 2 things:
manage_vhostis ran before the current symlink has been updated. This results in errors during nginx setup and breaks automatic deployments. It also seems that the error which this may result fully stops the automatic nginx reloadhypernode-systemctl settings redis_version {$config->getVersion()} --block"asks for a yes/no input. Which is not provided by deploy. If redis is upgraded to 7.x during deploy the deployment will never finish since the prompt will always come up even if redis is already on version 7.x