Error Message

Feb 21 18:33:01 server101 kernel: sd 3:0:0:0: reservation conflict
Feb 21 18:33:01 server101 kernel: sd 3:0:0:0: [sdb]  Result: hostbyte=DID_OK driverbyte=DRIVER_OK
Feb 21 18:33:01 server101 kernel: sd 3:0:0:0: [sdb] CDB: Write(10): 2a 00 00 6c 0d 98 00 00 08 00
Feb 21 18:33:01 server101 kernel: end_request: critical nexus error, dev sdb, sector 7081368
Feb 21 18:33:01 server101 kernel: sd 3:0:0:0: reservation conflict
Feb 21 18:33:01 server101 kernel: VxVM vxdmp V-5-0-0 [Error] i/o error occurred (errno=0xd) on dmpnode 201/0x10

Feb 21 18:33:01 server101 kernel: sd 3:0:0:0: reservation conflict
Feb 21 18:33:01 server101 kernel: sd 3:0:0:0: [sdb]  Result: hostbyte=DID_OK driverbyte=DRIVER_OK
Feb 21 18:33:01 server101 kernel: sd 3:0:0:0: [sdb] CDB: Write(10): 2a 00 00 6c 0d 98 00 00 08 00
Feb 21 18:33:01 server101 kernel: end_request: critical nexus error, dev sdb, sector 7081368
Feb 21 18:33:01 server101 kernel: sd 3:0:0:0: reservation conflict
Feb 21 18:33:01 server101 kernel: VxVM vxdmp V-5-0-0 [Error] i/o error occurred (errno=0xd) on dmpnode 201/0x10
Feb 21 18:33:01 server101 kernel:
Feb 21 18:33:01 server101 kernel: VxVM vxio V-5-0-1920 Disk group name=ctr_dg Subdisk=sdb-01 block 7015576: Uncorrectable write error
Feb 21 18:33:01 server101 kernel: end_request: I/O error, dev VxVM15000, sector 7015576
Feb 21 18:33:01 server101 kernel: vxfs: msgcnt 2 mesg 038: V-2-38: vx_dataioerr - /dev/vx/dsk/datadg/data file system file data write error in dev/block 0/7015576

InfoScale component having limitations with the server hosted in an ESXi environment

Sharing VMDK files between virtual machines

When sharing VMDK files between virtual machines, SCSI BUS sharing mode for the corresponding SCSI controllers can be set to either “Physical” or “Virtual” modes. Setting this mode to “Physical” causes SCSI reservation conflict and I/O error on DMP. This issue occurs in LVM and raw disks also.

---This was taken from the Infoscale Virtualization Guide for Linux on ESXi.

Disable simultaneous write protection for the shared disk using the multi-writer flag.

The procedure is described in the following VMware Knowledge Base article:
http://kb.vmware.com/kb/1034165

Initially, both VMs (virtual machines) were in one ESX host. Then, VMs were migrated to 2 new ESX hosts. As part of the move, the SCSI bus sharing was changed from "none" to "physical". After the change, SCSI I/O errors were logged on the server. The disk I/O issue caused the volume to be disabled, then the file system. Thus, the mount point was no longer accessible.