Running VMWare Workstation 12.1.1 (Build 3770994) on Server 2008 R2 SP1 (fully patched) and I have noticed that overall memory usage seems to climb while the machine runs normally. I patched the host today so shut down all of my VMs and found that 15G of 60G total was still in use. 13G of that was attributed to NonPaged pool and PoolMon confirmed that tag vmho is the culprit. I see some ~3.3M allocations of 2192 bytes each for 7.1G. The next highest NonPaged consumer is LSbf for 51M so vmho is definitely the culprit.
The issue seems to have started with the 12.1.1 upgrade, as best I can recall. The driver is version 188.8.131.52 (Build 2003952) dated 4/14/2016. I run with 9 host only NICs where RRAS on the host routes between them. NAT and DHCP are all disabled in the VMNet configuration. I cannot find anything online indicating that anyone else is hit or reporting this. I tried to disable the NICs, to see if doing so would free up the memory, but only the first one disabled and the rest seemed to ignore the request entirely. I then tried to re-enable the first NIC and it is hung with the Enabling... dialog now.
Is VMWare aware of this issue? A fix would be greatly appreciated.
I am running VMWare Player 12.1.1-3770994.x86_64 also on RedHat Linux (2.6.32-573.el6.x86_64) and I am seeing a significant memory leak. Have you found any solution ?
Sorry for the slow reply Mars. I reinstalled the OS on my machine to Server 2012 R2 and upgraded the VM tools within all of my VMs to current and am no longer seeing the problem here. I do not know if the leak you are seeing is the same as mine but would suggest that you ensure the current version of the tools is installed across all VMs to see if that helps.