Solaris Cluster SCSI "Reservation Conflict" Panics on Node/System/Server Resolution Path
(Doc ID 1019980.1)
Last updated on JULY 03, 2019
Applies to:Solaris Cluster - Version 3.0 and later
Solaris Cluster Geographic Edition - Version 3.1x and later
Oracle Solaris on SPARC (64-bit)
Oracle Solaris on SPARC (32-bit)
Oracle Solaris on x86-64 (64-bit)
Solaris Cluster 3.x node panics with "SCSI Reservation Conflict" with the following messages:
A major issue for clusters is a failure that causes the cluster to become partitioned (called split brain). This condition is detected by the Solaris Cluster
Membership Monitor. A nice description of Cluster Membership and Fencing is in the 3.2_u3 Overview. The design goal here is to prevent data corruption.
There are three types of SCSI reservation methods: SCSI2, SCSI3 and SCSI2/SCSI3 Persistent Group Reservation Emulation (PGRE).
An excellent Knowledge Management (KM) document for more information on reservation types is:
Solaris Cluster SCSI Reservations Types and Fencing Options - How To Check/Set/Configure SCSI-2, SCSI-3, PGRE on Devices and Quorum (Doc ID 1008224.1)
For split-brain conditions, one or more nodes may panic with RC when attempting to gain the quorum device due to implementation of the
MHIOCENFAILFAST ioctl. This is what tells the kernel to panic. A nice description is in the 3.2_u3 Concepts Guide.
To view full details, sign in with your My Oracle Support account.
Don't have a My Oracle Support account? Click to get started!