Hi all,
I played around with resource pools and discovered something that looks
like a really simple and annoying bug, but I'm not sure if it occurs only in
our environment or if it is a general bug in VirtualCenter (version 2.01 with
patch 1 applied) or the VI client.
So please try to reproduce the following behaviour and let me know
your results:
In the summary-tab of a resource pool its CPU and memory usage is
displayed in the upper right corner. It is my understanding, that the
pool's memory usage is the sum of the memory usages of all VMs that
belong to this pool. The problem now is that this number is correctly
displayed only if it is smaller than or equal to 2047 MB[/b]. As soon as it
grows over this limit it will be displayed as a negative number[/b]!
This looks like a simple variable overflow somewhere in the VirtualCenter
code.
Can anyone confirm or object this observation?
Thanks and best regards
Andreas
P.S.: I filed an SR with VMware, but I'm still waiting for a confirmation.
I've seen something like that when doing a offline migrate.
When migrating a VM from SAN to local storage the system was finised at 100% and filling up till 135% or so 😐
I can confirm this problem. In our VC it shows for one resource pool:
Memory usage: -1216 MB
The latest patches are applied (2.0.1 Build 33643)
Guido
Thanks Guido,
can you also confirm that it happens when the 2GB-limit is crossed?
For testing this just create a new Resource pool (with default settings), then move VMs in there one after another until they consume more than 2047 MB of host memory altogether.
I opened a support request with VMware, they are still investigating it.
One information could be useful: What database are you using for VirtualCenter?
We are using Oracle.
Regards
Andreas
Same problem here. only, i'm NOT using expendable resources. this even occurs with fixed resource limitations.
i'm using the latest version of VMWare Virtual Center (Build 40644) in combination with a SQL2000 cluster for the database.
I think it could have to do something with the fact that the memory usage percentage (see your graph's) spikes way beyond a 100% which of course is impossible. These spikes occur when the amount of active memory spikes (a little bit).
pity i can't post screenshots. this would explain the above situation a lot easier.
OK, i've been mucking around with the resourcepool in question and now I really don't get it anymore. the resourcepool say's that about 900mb is used, but when i look at the graph, it states that the current amount of active memory is around 400mb's.
the figure on the resourcepool tab just doesn't make any sense.
Message was edited by:
Teun
I have seen and i am geeting resource pool memory used value above that 100% may be 135% , 145% and its reach around 225% also. How cum its possible @ can anybody help me out.
I can't able to generate value of resource pool swap usage value in graph metric @ i already added 8 VM in Resource pool with complete load and showing zero value. and its remain constant in graph.
Reagrds
Vishal Kute
vishalbkute@gmail.com