Stacking or spreading
Although Nova exposes several filters for better computing workload distribution and management, you may encounter a problem where you are not able to spawn more instances even though Horizon is showing free resources. By default, instances will be spread across the selected compute nodes one by one. This ensures that the compute workload is fairly distributed across all hosts. Eventually, the behavior of the scheduler as to how to dispatch the instances across compute nodes can be tricky in this case. By default, filters provide a finely selected list of hosts. Requesting different instances with different flavors would create a sort of concurrency, and even prevent them from creating additional instances with larger ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access