Datastore is not accessible post storage array maintenance activity

Products

• VMware vSphere ESXi

Issue/Introduction

Symptoms:

• A few of the datastores are not accessible post-storage array maintenance activity.

• These VMs running on that datastore goes into an inaccessible state or hung state.

The following log events are noticed on the var/run/log/vmkernel.log of the ESXi host.

YYYY-MM-DDThh:mm:ss.ms cpu9:2098178)WARNING: lpfc: lpfc_path_claim_handler:11108: 1:(0):3405 ScsiScanAndClaimPath on SAdapter Num x1 TID 19 DID x038201 Failed, status Maximum allowed SCSI paths have already been claimed
YYYY-MM-DDThh:mm:ss.ms cpu2:2098178)WARNING: lpfc: lpfc_path_claim_handler:11108: 1:(0):3405 ScsiScanAndClaimPath on SAdapter Num x1 TID 20 DID x031101 Failed, status Maximum allowed SCSI paths have already been claimed
YYYY-MM-DDThh:mm:ss.ms cpu2:2098178)WARNING: lpfc: lpfc_path_claim_handler:11108: 1:(0):3405 ScsiScanAndClaimPath on SAdapter Num x1 TID 21 DID x038900 Failed, status Maximum allowed SCSI paths have already been claimed

Environment

VMware vSphere ESXi 6.x
VMware vSphere ESXi 7.x
VMware vSphere ESXi 8.x

Cause

The issue occurred because all storage paths became unavailable during or after the upgrade for this datastore.

The ESXi exhausted all available paths [4096] and failed to re-establish connectivity to the datastore device.

Reference Article: VMware Configuration Maximums

Resolution

Kindly work with your storage team to reduce the number of paths advertised to the ESXi hosts.

Additional Information

Run below cmd to check for count of total path mapped to the host
esxcfg-mpath -b | grep -v naa.* | wc -l

Also, run this command on the ESXi host as root to view the number of LUNs seen by the VMkernel:
esxcli storage core device list |grep -E "^naa|^eui|^mpx|^t10" |wc -l