I’m just curious what are your values for min/max/average values for
r/w operations on virtual disks which are hosted on a SAN attached via
fibre.
Unfortunately I have no values to compare with, but the values I get
there for our groupwise servers seems to me to be a little bit high…
Also the performance monitor of one (groupwise) server shows higher
latency for read operations that write operations. Shouldn’t that be
inverse?
I’m no expert in vmware, but interested in that vlaues! So I thought,
why not ask in here?
I’m just curious what are your values for min/max/average values for
r/w operations on virtual disks which are hosted on a SAN attached via
fibre.
Unfortunately I have no values to compare with, but the values I get
there for our groupwise servers seems to me to be a little bit high…
Also the performance monitor of one (groupwise) server shows higher
latency for read operations that write operations. Shouldn’t that be
inverse?
I’m no expert in vmware, but interested in that vlaues! So I thought,
why not ask in here? :)[/color]
As a general roule of thumb, you should stay well below 30ms. Anything
higher indicates a problem. If you see numbers in the hundreds, it’s a
severe problem, and things start to slow down massively to a point of
being unbearable.
Generally speaking, the faster the disks, the better. mrosen is
correct. How do you have your disks set up? Some are better than
others. Also, what kind of SAN are we talking about? Hope many iops
for the baseline?
–El
bendeichp;2229128 Wrote:[color=blue]
Hi Forum,
I’m just curious what are your values for min/max/average values for
r/w operations on virtual disks which are hosted on a SAN attached via
fibre.
Unfortunately I have no values to compare with, but the values I get
there for our groupwise servers seems to me to be a little bit high…
Also the performance monitor of one (groupwise) server shows higher
latency for read operations that write operations. Shouldn’t that be
inverse?
I’m no expert in vmware, but interested in that vlaues! So I thought,
why not ask in here?
Thanks in Advance,
Pascal[/color]
–
El LaGrew
elbert.lagrew#katun.com
elagrew on twitter
as I know by now, we have to different disks in use. Sata’s and SAS.
You can guess now on which the “slow” groupwise is hosted.
All other gw server are on the SAS disks ad we will move this one also
to SAS.
I can let you now, how the r/w latency improves after the change.
