You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The spec file currently does a live update. I would like an option to create a spec file that does a stop service, update, start server for cases where the release does not work with live updates.
I would like the default to be stop and start. This way, the developer is forced to think about whether or not their project supports live update.
I know that I just had a problem with a r Module while working on a change. I had chanced the schema of a gen_server stata_data struct and my gen_server crashed after reloading. So, I believe the safer default is to stop update and restart the service during the rpm installation
That seems like a good idea to me. The appups I generate by default via exrm are not super smart - they are about as good as it's possible to get without manually writing one though. For simple cases, live upgrades should work as expected, but for the more general case, and in my opinion, the more likely case - they are just a good starting point, manual work is still needed for each upgrade.
The spec file currently does a live update. I would like an option to create a spec file that does a stop service, update, start server for cases where the release does not work with live updates.
I would like the default to be stop and start. This way, the developer is forced to think about whether or not their project supports live update.
I know that I just had a problem with a r Module while working on a change. I had chanced the schema of a gen_server stata_data struct and my gen_server crashed after reloading. So, I believe the safer default is to stop update and restart the service during the rpm installation
@bitwalker what do you think about this default?
The text was updated successfully, but these errors were encountered: