February 2018
Intermediate to advanced
300 pages
6h 17m
English
A complementary approach to empower the scalability of large OpenStack deployment is to define policies as to how your compute nodes should be physically grouped and how they should dispatch instances. As we discussed in the previous section, segregating your large infrastructure based on AZs and host aggregates will help you to handle where a given instance will be hosted from a high-level perspective. Diving deeper, more use cases would raise the need to place a given virtual machine in a very particular fashion. Nova has an overwhelming list of features implemented to segregate workloads in a large OpenStack environment. The nova-scheduler service determines in which hypervisor host a virtual machine ...
Read now
Unlock full access