Seeing kernel: scsi host2: hdr status = FCPIO_DATA_CNT_MISMATCH and path failures

Solution Verified - Updated -

Issue

  • Cisco UCS blades and Nexus switches with mismatching QoS settings experiencing FC underruns under heavy load.
  • SCSI timeouts and multipath path failures observed in /var/log/messages.
  • The UCS switch logs show:

     %IPQOSMGR-2-QOSMGR_DCBXP_PFC_CMP_FAIL_MSG:  Ethernet1/31 - qos config for PFC not compatible with the peer
     %IPQOSMGR-2-QOSMGR_DCBXP_PFC_CMP_FAIL_MSG:  Ethernet1/30 - qos config for PFC not compatible with the peer
    
  • The system will log

    kernel: scsi host1: hdr status = FCPIO_DATA_CNT_MISMATCH
    kernel: scsi host1: hdr status = FCPIO_DATA_CNT_MISMATCH
    kernel: scsi host2: hdr status = FCPIO_DATA_CNT_MISMATCH
    kernel: scsi host2: hdr status = FCPIO_DATA_CNT_MISMATCH
    kernel: scsi host2: hdr status = FCPIO_DATA_CNT_MISMATCH
    kernel: scsi host1: hdr status = FCPIO_DATA_CNT_MISMATCH
    kernel: scsi host1: hdr status = FCPIO_DATA_CNT_MISMATCH
    kernel: scsi host2: hdr status = FCPIO_DATA_CNT_MISMATCH
    kernel: scsi host2: hdr status = FCPIO_DATA_CNT_MISMATCH
    kernel: scsi host1: hdr status = FCPIO_DATA_CNT_MISMATCH
    
  • And:

    kernel: sd 2:0:0:16: [sdcy] Medium access timeout failure. Offlining disk!
    kernel: device-mapper: multipath: Failing path 70:96.
    multipathd: checker failed path 70:96 in map mpathq
    multipathd: mpathq: remaining active paths: 3
    kernel: host2: Assigned Port ID 7c0008
    kernel: sd 2:0:0:16: [sdcy]  Result: hostbyte=DID_TRANSPORT_DISRUPTED driverbyte=DRIVER_OK
    kernel: sd 2:0:0:16: [sdcy] CDB: Read(10): 28 00 04 96 2c 00 00 04 00 00
    kernel: end_request: I/O error, dev sdcy, sector 76950528
    kernel: sd 2:0:0:16: rejecting I/O to offline device
    kernel: sd 2:0:0:16: [sdcy] killing request
    kernel: sd 2:0:0:16: rejecting I/O to offline device
    kernel: sd 2:0:0:16: rejecting I/O to offline device
    kernel: sd 2:0:0:16: rejecting I/O to offline device
    kernel: sd 2:0:0:16: rejecting I/O to offline device
    kernel: sd 2:0:0:16: [sdcy]  Result: hostbyte=DID_NO_CONNECT driverbyte=DRIVER_OK
    kernel: sd 2:0:0:16: [sdcy] CDB: Read(10): 28 00 03 2f cc 00 00 04 00 00
    kernel: end_request: I/O error, dev sdcy, sector 53464064
    kernel: sd 2:0:0:16: [sdcy]  Result: hostbyte=DID_TRANSPORT_DISRUPTED driverbyte=DRIVER_OK
    kernel: sd 2:0:0:16: [sdcy] CDB: Read(10): 28 00 00 09 fc 00 00 04 00 00
    kernel: end_request: I/O error, dev sdcy, sector 654336
    

Environment

  • Red Hat Enterprise Linux 8
  • Red Hat Enterprise Linux 7
  • Red Hat Enterprise Linux 6
  • Red Hat Enterprise Linux 5
  • Cisco UCS blades with fnic interfaces.
  • Configured with mismatched QoS (quality of service) settings between switches.

Subscriber exclusive content

A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.

Current Customers and Partners

Log in for full access

Log In

New to Red Hat?

Learn more about Red Hat subscriptions

Using a Red Hat product through a public cloud?

How to access this content