Feature #621
Dynamic CPU load for VMs
Status: | Closed | Start date: | 05/10/2011 | |
---|---|---|---|---|
Priority: | Normal | Due date: | ||
Assignee: | - | % Done: | 0% | |
Category: | Drivers - VM | |||
Target version: | - | |||
Resolution: | fixed | Pull request: |
Description
Hi,
as part of the May 2011 IRC session http://opennebula.org/public/110509.html I'd like to propose
to make the CPU usage calculation more dynamic.
jmelis: sneumann: possibly a dynamic CPU determination, based on e.g. the 1,5 or 15mins load of a VM which is then considered by the scheduler. There might be more advanced ways to determine the actual/current/dynamic CPU value...
jfontan: at least for this or the next version it wont be possible
jfontan: to change that value, the CPU reserved for a VM
jfontan: there are two values that should be changed
jfontan: one is the VM template CPU value, that is the easy part
jfontan: and then the CPU available in the host the VM is running
jfontan: that is the "acpu" value you get with onehost list
jfontan: I think now it is a good time to add this ticket as we are changing the database for the next OpenNebula versions
jfontan: so it is easier to add some features
jfontan: and will be good to take those ideas into account
If the CPU parameter is set to 1.0 ONE will not overcommit available nodes,
and with the availability of a dynamic re-scheduler it is possible to squeeze
all VMs on as few nodes as possible, allowing to power off the unused nodes.
Yours,
Steffen
History
#1 Updated by Ruben S. Montero about 8 years ago
- Category set to Drivers - Auth
#2 Updated by Ruben S. Montero about 8 years ago
- Category changed from Drivers - Auth to Drivers - VM
#3 Updated by Ruben S. Montero over 7 years ago
- Status changed from New to Closed
- Resolution set to fixed
This should be addressed by the new drivers