Each month we deliver a free webcast on the latest tools and tips from our expert instructors and content developers. Earlier this summer Francis Daly, a Staff Technical Training Specialist working in our Technical Support University, delivered VMware vSAN Troubleshooting Tools, Tips and Tricks and the questions were great as always. We pulled a few of them below, but you can read the entire list and listen to the recording on your schedule.
What is the best process for backups?
Here is a good guide to backups for vSAN (Ed. note: you’ll be prompted to download a PDF)
How do I know if my hardware is supported or not? current firmware version, make model, etc….
You can check it here. You can also watch the video available in the VMware Learning Zone where I did a step by step walk through of how to gather all of that information.
If the vSAN cluster has two DGs per node, how do you create VMs where the workload is distributed equally on all disks in both DGs on each node?
This is a policy setting. So you need to have a certain amount of hosts also. If you want to use many DG’s in this case you may use a higher level of Failures to tolerate and a higher number of stripes. It will then spread components more widely.
Do we need to configure cache drive in all flash cluster?
A flash drive is always required, when you create your disk group you can select which disk is for the cache tier, in this case you will choose the SSD that is fastest.
How do you determine if there is a bad mirror on one of the disks associated with a VM?
When you select the VM you can then check the Monitor – vSAN and select the VM or VMDK, then under Policies you will see if the failed disk is affecting your VM.
Can vSAN automatically rebalance disks – either on schedule or based on a trigger (reaching 10% above threshold)?
vSAN uses CLOM (Cluster Level Object Manager) it will typically rebalance if disk groups go above 70%. There is also proactive rebalance in v6.6.
Would Health UI only be enough to troubleshoot our vSAN infrastructure?
Under most circumstances, it is the perfect place to start and will provide very good indications of current issues and potential causes. It only works if your vCenter is operational though.
Can a Disk Balance warning be ignored?
Yes, depending on the severity. If the disk balance indicates that a particular host is becoming close to full, you should not ignore it. However, if it simply indicates that data is spread evenly across the cluster, this could be ignored as it may resolve itself as more objects and components are created. New components will be placed on disks that have the most capacity free.
What would be the correct procedure for powering off and restarting an entire VSAN cluster?
You would first need to power off all VMs in the cluster. Then place each node into maintenance mode with no action.
Will you get a notification if one of the physical SSD or HDD has failed?
You will get a UI warning at the host and the cluster level. This will be produced when the disks encounter a Permanent Failure.