|
Do you mean:
[vmware fysical machine] containing:
- vmware linux/windows guest
- vmware linux/windows guest
...
- vmware linux/windows guest
and on one of the guests you run LVS?
or:
[vmware fysical machine] which runs LVS (so it must be VMWare ESX) and a
some guests?
In the above setup, i can confirm that LVS runs lousy. As soon as one of the
guest hosts pulls a little too much bandwidth or cpu usage, LVS suffers from
it and initiates a failover because the directors can't see eachother
anymore for a certain time period.
The second example i have no experience with, but i can imagine the same
performance loss.
I ran a 2-node LVS-DR (heartbeat/ldirectord) on a vmware esx server for over
a year, and i'm desperate to shut them down and move them onto a stand-alone
fysical machine.
Hope this helps,
Sebastian
On 6/1/07, Erol YILDIZ <erol.yildiz@xxxxxxxxxxxxxxxxxx> wrote:
>
> Hi,
>
> What would be the performance of a LVS cluster with 8 machines for
> example which run a vmware server aplication on it with many linux
> /windows based servers installed over it?
>
> thanks
>
>
> _______________________________________________
> LinuxVirtualServer.org mailing list - lvs-users@xxxxxxxxxxxxxxxxxxxxxx
> Send requests to lvs-users-request@xxxxxxxxxxxxxxxxxxxxxx
> or go to http://lists.graemef.net/mailman/listinfo/lvs-users
>
|
