Understanding host level resync management

Products

• VMware vSAN

Issue/Introduction

This article discusses the improved resync granularity for resync monitoring and throttling mechanisms at the host level.

Impact/Risks:
If resync operations are allowed to flow with too much bandwidth, some environments may experience congestion as a result which can impact latency sensitive applications like databases. Throttling mechanisms are needed to reduce congestion caused by too much resync bandwidth to improve VM performance for latency sensitive applications.

Environment

VMware vSAN (All Versions)

Resolution

In vSAN 6.6 we introduced a new esxcli branch called esxcli vsan resync. This allows us to have more granular control over resync monitoring/throttling at the host level of a vSAN node without having to rely on RVC or use a global setting from the GUI.  

Monitoring resync throughput at the host level

To monitor resync traffic bandwidth in real time run the command:

esxcli vsan resync bandwidth get 

Example result: Level: 0 Mbps

Level: 0 is an indication that there is no resync activity occurring on this host. 

 

Throttling resync throughput 

It is important to state that the throughput threshold you define applies at the disk group level. Example: If you have 3 disk groups and define a threshold of 150MB, the expected throughput on the host for resync traffic would be a combined total of 450 MB. 

To check the current throttling settings (0 is default), run command:

esxcli vsan resync throttle get 

Example output: Level: 0 Mbps

To change existing throttling settings run command:
 
esxcli vsan resync throttle set --level <0-512mb> 

Example: 

To throttle resync traffic to not exceed 200 MB of bandwidth on this host run command:

esxcli vsan resync throttle set --level 200

which can be verified with the following:

esxcli vsan resync throttle get
Level: 200 Mbps

Additional Information