vSAN Adaptive Resync and Capacity Reporting Improvements

With the release of vSAN 6.2 Patch 06, there are a few new features and bug fixes that you can take advantage of to simplify and automate operations. I am going to talk about two of those elements below: Adaptive Resync and Capacity Reporting improvements.

Adaptive Resync: A new adaptive resync throttling mechanism has been introduced in vSAN 6.2 Patch 06 to account for periods of high VM IO during resyncs. In the past when a resync operation was underway due to a failure, component change or otherwise, system administrators could limit the bandwidth of a resync operation manually.

vSAN tries to maximise throughput to the backend devices in order to achieve object policy compliance as quickly as possible, this ensures data meets the protection and performance levels specified in SPBM policies. The manual throttling mechanism was introduced in vSAN 6.6; this allowed users to manually limit the resync bandwidth of their vSAN instance on a sliding scale.

With vSAN 6.2 Patch 06, this process has been back-ported and is now automated.  It has some specific conditions around how it behaves, vSAN will dynamically allocate bandwidth to resync traffic (up to a maximum cap) as long as it is not affecting VM IO from a latency perspective.

When VM IO is under heavy load, any resync traffic will be dialed back automatically to allow for production workloads to continue unhindered, ensuring consistent latency across traffic types and workloads while still ensuring object compliance in a timely manner.

Capacity Reporting: The second update in this patch is improvements to the capacity reporting as viewed in vCenter at a cluster level (accessed through Monitor -> vSAN -> Capacity).

This update will add some clarification as to the breakdown of capacity used on the vSAN datastore by adding options to split the view into “Used by VMs” and “Used by System.” You can now quickly view what is consumed by VMs and the system as separate entities, no calculations required!

The distinction between these consumption types will highlight the difference between what is utilized by the VMs you have provisioned, and what vSAN itself is using for things like deduplication/compression, RAID parity, swap, performance metrics, checksums, filesystem overhead, as well as iSCSI homes and target objects.

Patch: Details for this patch release can be found in our KB here and updates can be applied, as usual, through VUM or esxcli.


Curious about VMware vSAN? Subscribe to the Virtual Blocks blog or follow our social channels at @vmwarevsan and Facebook.com/vmwarevsan for the latest updates. For more information about VMware vSAN, visit https://www.vmware.com/products/vsan.html


Leave a Reply

Your email address will not be published.