'mkfs.xfs' command gets hung with SCSI, multipath errors on storage that only supports older SPC-2 scsi specification

Issue

• mkfs.xfs command gets hung with following errors:

  kernel: blk_update_request: critical target error, dev dm-0, sector 4194309168 op 0x9:(WRITE_ZEROES) flags 0x400000 phys_seg 0 prio class 0
  kernel: blk_update_request: critical target error, dev dm-0, sector 4194311216 op 0x9:(WRITE_ZEROES) flags 0x400000 phys_seg 0 prio class 0
  kernel: blk_update_request: critical target error, dev dm-0, sector 4194313264 op 0x9:(WRITE_ZEROES) flags 0x400000 phys_seg 0 prio class 0
  multipathd[1920]: sdc: mark as failed
  multipathd[1920]: mpathx: remaining active paths: 3
  multipathd[1920]: sdg: mark as failed
  multipathd[1920]: mpathx: remaining active paths: 2
  multipathd[1920]: sdk: mark as failed
  multipathd[1920]: mpathx: remaining active paths: 1
  multipathd[1920]: sdo: mark as failed
  multipathd[1920]: mpathx: Entering recovery mode: max_retries=5
  multipathd[1920]: mpathx: remaining active paths: 0
  kernel: blk_update_request: critical target error, dev dm-0, sector 4194319408 op 0x9:(WRITE_ZEROES) flags 0x400000 phys_seg 0 prio class 0
  kernel: blk_update_request: critical target error, dev dm-0, sector 4194317360 op 0x9:(WRITE_ZEROES) flags 0x400000 phys_seg 0 prio class 0
  [...]
  kernel: blk_cloned_rq_check_limits: over max size limit. (2048 > 0)
  kernel: device-mapper: multipath: 253:0: Failing path 8:96.
  kernel: blk_cloned_rq_check_limits: over max size limit. (2048 > 0)
  kernel: device-mapper: multipath: 253:0: Failing path 8:224.
  kernel: blk_cloned_rq_check_limits: over max size limit. (2048 > 0)
  kernel: device-mapper: multipath: 253:0: Failing path 8:160.
  kernel: blk_cloned_rq_check_limits: over max size limit. (2048 > 0)
  kernel: device-mapper: multipath: 253:0: Failing path 8:32.
  [...]
  

• Same issue does not occur when creating ext3 filesystem using mkfs.ext3 command.