High write latency observed in vSAN cluster
search cancel

High write latency observed in vSAN cluster

book

Article ID: 404991

calendar_today

Updated On:

Products

VMware vSAN

Issue/Introduction

Symptoms:

vSAN cluster may report high write latency and VMs on vSAN datastore may be hung or perform poorly.

Upon checking performance graphs on vCenter, it may show a spike in the write/read latency.

Environment

VMware vSAN 7.x

VMware vSAN 8.x

Cause

Check if there are any hardware issue which can lead to performance issues on the vSAN cluster.

/var/run/log/vobd.log will show if there are any hardware issues observed on the cluster as shown below:

2025-07-07T16:33:21.460Z: [netCorrelator] 5027556572710us: [vob.net.dvport.uplink.transition.down] Uplink: vmnic2 is down. Affected dvPort: 4146/50 31 af 62 29 79 df ca-11 95 19 e5 01 5d 61 18. 1 uplinks up. Failed criteria: 128
2025-07-07T16:33:21.460Z: [netCorrelator] 5027556572725us: [vob.net.dvport.uplink.transition.down] Uplink: vmnic2 is down. Affected dvPort: 8232/50 31 af 62 29 79 df ca-11 95 19 e5 01 5d 61 18. 1 uplinks up. Failed criteria: 128
2025-07-07T16:33:21.460Z: [netCorrelator] 5027556572749us: [vob.net.dvport.uplink.transition.down] Uplink: vmnic2 is down. Affected dvPort: 1404605/50 31 af 62 29 79 df ca-11 95 19 e5 01 5d 61 18. 3 uplinks up. Failed criteria: 128

2025-07-07T16:33:21.807Z: [netCorrelator] 5027556922668us: [vob.net.dvport.uplink.transition.up] Uplink: vmnic2 is up. Affected dvPort: 4146/50 31 af 62 29 79 df ca-11 95 19 e5 01 5d 61 18. 2 uplinks up
2025-07-07T16:33:21.807Z: [netCorrelator] 5027556922683us: [vob.net.dvport.uplink.transition.up] Uplink: vmnic2 is up. Affected dvPort: 8232/50 31 af 62 29 79 df ca-11 95 19 e5 01 5d 61 18. 2 uplinks up
2025-07-07T16:33:21.807Z: [netCorrelator] 5027556922706us: [vob.net.dvport.uplink.transition.up] Uplink: vmnic2 is up. Affected dvPort: 1404605/50 31 af 62 29 79 df ca-11 95 19 e5 01 5d 61 18. 4 uplinks up

vCenter or host events will also show if there are any failures on the vmnic as shown below:

VMNIC flapping events can cause performance issue, which can lead to the input/output errors reported on the VM.

Resolution

Place the ESXi host reporting hardware failures, in maintenance mode to stabilize the vSAN cluster and temporarily relieve it of the performance issues.

Fix the hardware issues observed on the uplinks reporting flapping events with the vendor.

Once replaced, take the host out of maintenance mode and monitor the vSAN cluster performance.

The write latency should be back to normal.