I have a case raised with VMware about this issue. The tech support guy I spoke to acknowledged this is a pretty major bug. Something is causing the Windows guest (I don't know if it affects Linux) to crash on snapshot removal. They don't have a workaround other than downgrading, which we are currently doing. This issue has apparently been escalated to the highest level as it is affecting an awful lot of customers. They don't have an ETA for when a fix will be available.
The error you see in the vmware.log file for the VM will look something like this:
2015-09-29T17:24:42Z[+8.996]| vcpu-0| I120: SymBacktrace[1] 000003fffbf1af60 rip=00000000162957c5 in function (null) in object /bin/vmx loaded at 000000001611f000
2015-09-29T17:24:42Z[+8.996]| vcpu-0| I120: SymBacktrace[2] 000003fffbf1b460 rip=0000000016377766 in function (null) in object /bin/vmx loaded at 000000001611f000
2015-09-29T17:24:42Z[+8.996]| vcpu-0| I120: SymBacktrace[3] 000003fffbf1b640 rip=00000000003d500f
2015-09-29T17:24:42Z[+8.996]| vcpu-0| I120: Msg_Post: Error
2015-09-29T17:24:42Z[+8.996]| vcpu-0| I120: [msg.log.error.unrecoverable] VMware ESX unrecoverable error: (vcpu-0)
2015-09-29T17:24:42Z[+8.996]| vcpu-0| I120+ Unexpected signal: 11.
In our case it affecting maybe 1% of our VMs every evening, but different VMs each time.
