OPERATIONAL DEFECT DATABASE
...
...
When performing a Virtual Machine(VM) snapshot with memory state on ESXi, the VM pauses for 8+ seconds when located on VxFlex OS Volume Datastore. The pause time could increase up to 30 s which cause the application to crash. /var/log/hostd on the ESXi node shows that the VM pause for 17 seconds. 2019-03-26T12:10:04.344Z info hostd[CEC6B70] [Originator@6876 sub=Vmsvc.vm:/vmfs/volumes/5c408943-10c050f2-50e3-48df370a53e8/TESTVM4/TESTVM4.vmx] State Transition (VM_STATE_CREATE_SNAPSHOT -> VM_STATE_ON) 2019-03-26T12:10:21.481Z info hostd[C281B70] [Originator@6876 sub=Vmsvc.vm:/vmfs/volumes/5c408943-10c050f2-50e3-48df370a53e8/TESTVM4/TESTVM4.vmx] State Transition (VM_STATE_REMOVEALL_SNAPSHOT -> VM_STATE_ON
When running a snapshot with VM virtual memory, the VM needs to pause for a few seconds to dump the VM virtual memory current state.The snapshot with memory state is a single thread operation which is not optimized for our SDC.The VxFlex OS SDC is optimized for multi-thread operations and when running single thread operation, the operation is limited to single device performance.The operation had to dump the VM virtual memory to the datastore and if the dump was running, the VM paused.The devices used are read-intensive which has low write performance thus VM virtual memory dump takes a long time to write which cause long pause.
The product is working as designed.Workaround: Using mixed or write intensive devices which would increase write performance and the virtual machine's virtual memory dump will be quicker which would let the VM pause for less time.
Click on a version to see all relevant bugs
Dell Integration
Learn more about where this data comes from
Bug Scrub Advisor
Streamline upgrades with automated vendor bug scrubs
BugZero Enterprise
Wish you caught this bug sooner? Get proactive today.
