Receiving "closed with dirty buffers. Possible data loss." alerts for VMFS datastores.

Products

• VMware vSphere ESXi

Issue/Introduction

Virtual machine is inaccessible and receiving alerts related to data loss for multiple VMFS files.

In ESXi logs, below traces are recorded :

vmkalert: cpu33:5916081)ALERT: BC: 3042: <vm_name>.scoreboard closed with dirty buffers. Possible data loss.
vmkwarning: cpu9:2098161)WARNING: HBX: 3729: '<Datastore_name>': HB at offset 3145728 - Reclaiming timed out HB failed: Timeout:
vmkwarning: cpu24:22851078)WARNING: iscsi_vmk: iscsivmk_StopConnection:735: vmhba66:CH:0 T:0 CN:0: iSCSI connection is being marked "OFFLINE" (Event:6)
vmkwarning: cpu8:2097441)WARNING: Vol3: 2968: '<Datastore_name>': Failed to clear journal address since JBC could not be Locked. This could result in leak of journal block at <type 6 addr 14>.
vmkwarning: cpu8:2097441)WARNING: Vol3: 3038: '<Datastore_name>': Failed to clear journal address in on-disk HB. This could result in leak of journal block at <type 6 addr 14>.
vmkwarning: cpu8:2097441)WARNING: LVM: 6632: Failed to close device naa.604e79##################000002:2: Busy
vmkwarning: cpu8:2097441)WARNING: Vol3: 4323: Error closing the volume: . Eviction fails: Failure

Environment

VMware vSphere (All Versions)

Cause

A dirty buffer is a portion of memory that contains modified data not yet written to disk. When a file is closed while it still has dirty buffers, it means that not all changes were saved to disk.
"Possible data loss" could be because the dirty buffers may not have been written to persistent storage, some recent changes to the file may be lost.
Unplanned storage disconnection might trigger  "Possible data loss" alerts.

Resolution

This is expected behavior, and no action is required unless VMFS-level corruption is observed or virtual machines enter an unhealthy state.