Can anyone point me to some VMware documentation if it exists that reccommends vMotion and Managment Network Traffic are seperated?
Alternatively a detailed explanation on the resons for not doing so would be nice.
Hi
Rhidian wrote:
Alternatively a detailed explanation on the resons for not doing so would be nice.
There are several different aspects, depends what you mean by "separated" ? Separated - using different vmnic's ? or maybe separated using different networks (vLAN) ?
Cheers
Artur
Separate as individual switches and physical network cards.
Depend of how many adapter you have. But best practice if possible, separate of adapter, use a dedicated gigabit.
http://communities.vmware.com/message/1788899 look.
http://www.vmware.com/pdf/Perf_Best_Practices_vSphere4.1.pdf
http://www.vmware.com/resources/techresources/10113 (Best practices network)
more docs about best practices http://vsphere-land.com/news/vsphere-41-the-links.html
Please, don't forget the awarding points for "helpful" and/or "correct" answers.
Mauro Bonder - Moderator
In reality if DRS is set to manual would this be a problem are there any security concerns? Additionally does vMotion really occur if there is a complete hardware failure?
Hi
Rhidian wrote:
1. In reality if DRS is set to manual would this be a problem are there any security concerns?
2. Additionally does vMotion really occur if there is a complete hardware failure?
1. It's not about security, DRS automatically (if it's set to automatic mode) balance workload between nodes in a cluster. If DRS is set to Full Automatic mode and some of the esx servers, in same cluster, are overutilized while others are underutilized (cluster is unbalanced), DRS is vMotion'ing VM's between nodes in a cluster to balance cluster (vMotion VM's from high utilized nodes to less utilized nodes in a cluster). If DRS is set to Manual - you can simply figure out what might happend (VM performance degradation, etc)
http://www.yellow-bricks.com/drs-deepdive/ - visit that link
2. If you have server down (complete HW failure) there is no vMotion traffic, VM's will be restarted by HA on the other hosts in the cluster
Cheers
Artur