Hi
Currently we are trying to limit the IOPS of our VMs because some of them are requesting to much from the storage. Therefore we started to configure this on the VMDK settings of one VM as described in KB#1038241.
Immediately after that, we saw that all VMs on the same datastore experience high storage latency (<150ms) even the storage is not exhausted and all other VMs have no IOPS Limit set.
If we remove the IOPS Limit of this one VM, the storage latency of all other VMs on the same datastore drops immediately to a normal level. (<5ms)
I already opened a call and got in contact with VMware Support about this. But till now we found no solution. Therefore i try this forum to find hints or possibly a solution for the problem.
VMware support already asked me to switch back to the old I/O scheduler as described in KB#2059192. But this had no effect to my problem.
Maybe here in this forum is someone who can throw some light on this.
Regards
Michael
