-
Notifications
You must be signed in to change notification settings - Fork 162
Open
Description
In the early days we introduced suture as a built in service runtime. It was possible to stop, start and restart services in the single binary. We are moving further away from that.
- we dropped support for dynamically starting or stopping sevices - we only use the restart functionality for services that depend on others
- we dropped individual service binaries and configure the set of services via env vars. that we can run a singe service on the cli with
opencloud <service> serveris ... unused. The helm chart relies on the env var. The compose stack uses it for the collaboration server .... which should be part of the opencloud container anyway. - I would like to combine more services and leave only sth like proxy, api, backend and storage.
Discussing with @dragonchaser we think we can replace suture with a rungroup, similar to how we already start the debug, http, grps, or other server in every service.
Speaking of which: we only really need one debug server ... not one for every service ...
Related: #2282
Reactions are currently unavailable
Metadata
Metadata
Assignees
Labels
No labels
Type
Projects
Status
Qualification