Bug #4362

KVM monitoring is very slow monitoring more than 20 VMs

Added by Javi Fontan over 4 years ago. Updated over 4 years ago.

Status:ClosedStart date:03/01/2016
Priority:NormalDue date:
Assignee:Javi Fontan% Done:

0%

Category:Drivers - VM
Target version:Release 5.0
Resolution:fixed Pull request:
Affected Versions:OpenNebula 4.14

Description

KVM poll script uses top command to gather CPU usage of qemu processes. This command only lets specify 20 pids so the monitoring is done in batches of 20 processes. Each take 4 seconds. This should be changed to retrieve the information directly in the script.

Associated revisions

Revision 363304ed
Added by Javi Fontan over 4 years ago

bug #4362: do not use top to get kvm CPU usage

History

#1 Updated by Javi Fontan over 4 years ago

  • Assignee set to Javi Fontan

#2 Updated by Javi Fontan over 4 years ago

  • Status changed from Pending to Closed
  • Resolution set to fixed

Information is now gathered by the script from /proc/<pid>/stat.

Also available in: Atom PDF