Node Reboots due to HDLM Multipathing issue accessing the Voting device.
Last updated on SEPTEMBER 23, 2014
Applies to:Oracle Database - Enterprise Edition - Version 126.96.36.199 to 188.8.131.52 [Release 11.2 to 12.1]
Information in this document applies to any platform.
HDS HDLM (Hitachi Dynamic Link Manager) Multipathing Solutions used for Voting device
The only available Voting Disk (ORCL:OCR_VOTE01) was not accessible due to SAN (Brocade) Switch unplug and re-plug.
The Node Rebooted obviously as the only Voting disks was unavailable.
Note - The SAN (Brocade) Switch unplug and re-plug was within i/o disktimeout.
The Diskgroup - OCR_VOTE is configured for External Redundancy.
As per CRS design an External Redundant Diskgroup will ONLY have one Voting disk configured.
The redundancy for an External Redundant Diskgroup should to be provided externally and not by Oracle ASM.
As CSSD was not able to access the only available Voting disk-ORCL:OCR_VOTE01 from all the 3 Nodes, CSS failure occurred, leading to Node Reboot.So from CRS perspective the behavior was correct and expected.
Sign In with your My Oracle Support account
Don't have a My Oracle Support account? Click to get started
Million Knowledge Articles and hundreds of Community platforms