This would be pretty helpful when it comes to tasks for example it would be really nice to say that these set environments can take precedence over a long queue backup. This would be pretty helpful for organizations with longer deployments. In addition possibly being able to scope or prefer certain tasks to a certain HA node ( this would be better with 3+ nodes so you still have HA on things you scope) For example you want your production region to prefer other nodes to be used mainly for deployments of certain regions. This would allow you to capacity plan in a better way because instead of just a basic X number of tasks you can really have X number of tasks with some overflow for special regions.
Hi Brent,
Thanks for getting in touch!
We don’t have anything specific planned for prioritising deployments, however if this is a common requirement, it’s definitely something we’d look at doing. In the meantime there is a technique that might help, particularly with High Availability.
After an Octopus Server creates a new Server Task for a deployment, that server is immediately triggered to look for new tasks to run, provided it has capacity (based on the number of concurrent tasks set in the Configuration > Nodes page). While it’s not 100% guaranteed, it therefore becomes much more likely that the server that triggered a task will be the one to run it.
You could combine this with a smarter load-balancer scenario whereby a number of nodes are load-balanced, while a “priority” server is reserved for higher priority deployments. Day-to-day operations could be run on the load-balanced endpoint while higher priority deployments could be run from the “priority” server. Note that if your servers are under high load and each node is at capacity, this may not have much effect as the priority server would still do its share of day-to-day work.
If this is something you think you’ll definitely need, can you put a feature request on UserVoice?
I hope that helps!
Damo