Host I/O errors cause Oracle Linux Database outage
Applies to
ONTAP
Issue
- In a FC/FCOE connection and Oracle Linux servers environment, production Database VM went down because of reported Host I/O errors.
- EMS log show link break and fabric login change (FLOGI) on two target ports within 1 minute of each other.
event log show -time > 100m -message-name *fct*
Node 1:
fct_tpd_work_thread_0: scsitarget.ispfct.linkBreak:error]: Link break detected on Fibre Channel target adapter 0f. Firmware status code status1 0x2, status2 0x7, and status4 0x0.
Node 2:
fct_tpd_work_thread_0: fcp.io.status:debug]: STIO Adapter:0f AEN 0x8048 (RECV_ERROR) MboxStatus1 0x17c MboxStatus2 0x20
fct_tpd_work_thread_0: scsitarget.fct.linkError:error]: FC target HBA: 0f received unexpected error: RECV_ERROR, AEN Status1: 0x8048, Status2: 0x17c, Status3: 0x20.
fct_tpd_work_thread_0: scsitarget.ispfct.linkBreak:error]: Link break detected on Fibre Channel target adapter 0f. Firmware status code status1 0x2, status2 0x7, and status4 0x0.