SFCFS - It is normal to have SCSI warning messages "reservations released" logged during CVM Master switchover

Article: 100004126
Last Published: 2022-02-23
Ratings: 0 0
Product(s): InfoScale & Storage Foundation

Problem

In a Storage Foundation for Cluster File System (SFCFS)  environment, it is normal for the system to log SCSI  warning messages "reservations release" during Cluster Volume Manager (CVM)  master switchover.    During CVM master switchover the SCSI-3 Persistent Reservations (SCSI-3 PR) Write Exclusive Registrants Only (WERO) reservation on the disks will need to be removed from the original CVM master, so that the reservation can be performed by the new CVM master.   When the reservations are removed, the disk array will generate SCSI Sense event and the corresponding SCSI messages will be logged by the host systems.

Error Message

The following SCSI warning messages are logged during the CVM master switchover.  These messages are normal during the switchover.    These are retryable errors and the operation system will retry the I/Os accordingly.

Sep  9 08:46:13 hostxxx scsi: [ID 107833 kern.warning] WARNING: /pci@1,700000/SUNW,qlc@0/fp@0,0/ssd@w50060e8005b15410,148 (ssd1869):
Sep  9 08:46:13 hostxxx           Error for Command: read(10)                Error Level: Retryable
Sep  9 08:46:13 hostxxx scsi: [ID 107833 kern.notice]   Requested Block: 67072                     Error Block: 67072
Sep  9 08:46:13 hostxxx scsi: [ID 107833 kern.notice]   Vendor: HITACHI                            Serial Number: 50 0B1540232
Sep  9 08:46:13 hostxxx scsi: [ID 107833 kern.notice]   Sense Key: Unit Attention
Sep  9 08:46:13 hostxxx scsi: [ID 107833 kern.notice]   ASC: 0x2a (reservations released), ASCQ: 0x4, FRU: 0x0

Nov 24 19:32:13 hostxxx scsi: [ID 107833 kern.warning] WARNING: /pci@3,700000/SUNW,qlc@0/fp@0,0/ssd@w50060482d5309fac,26 (ssd21):
Nov 24 19:32:13 hostxxx        Error for Command: write(10)               Error Level: Retryable
Nov 24 19:32:13 hostxxx scsi: [ID 107833 kern.notice]   Requested Block: 3283888                   Error Block: 3283888
Nov 24 19:32:13 hostxxx scsi: [ID 107833 kern.notice]   Vendor: EMC                                Serial Number: 90"J"000Q
Nov 24 19:32:13 hostxxx scsi: [ID 107833 kern.notice]   Sense Key: Unit Attention
Nov 24 19:32:13 hostxxx scsi: [ID 107833 kern.notice]   ASC: 0x2a (reservations released), ASCQ: 0x4, FRU: 0x0

In the /etc/vx/dmpevent.log file, the following messages will also be logged:

Wed Nov 24 19:32:16.737: SCSI error occured on Path c4t50060482D5309FA3d71s2: opcode=0x5f reported unit attention (status=0x2, key=0x6, asc=0x2a, ascq=0x4)
Wed Nov 24 19:32:16.739: SCSI error occured on Path c4t50060482D5309FA3d71s2: opcode=0x5f reported reservation conflict (status=0x18, key=0x0, asc=0x0, ascq=0x0)

Solution

These SCSI warning messages are normal during CVM master switchover and can be safely ignored.

 

 

Was this content helpful?