Oracle RAC 11gR2 database crashes during take over and giveback of NetApp filers.
Issue
- When Netapp Fail-over/giveback is triggered, with no other Oracle users active, there is a message logged about a single IO Error, and the Oracle database crashes.
- This issue never occurs when strace is active.
- The file in question is NFS mounted.
- The error is:
/u01/app/oracle/diag/rdbms/bnrdba/BNRDBA1/trace/BNRDBA1_lmon_23671.trc:
ORA-00202: control file: '/u03/oradata/BNRDBA/control01.ctl'
ORA-27072: File I/O error
Linux-x86_64 Error: 5: Input/output error
Additional information: 4
Additional information: 66
Additional information: -1
- This originally occurred on the production system (which has 8 nodes).
- This was then replicated on crash and burn cluster which has 3 nodes.
- Then it was replicated with only one node in the cluster. The problem still occurred.
Environment
- NetApp NFS File Server
- Red Hat Enterprise Linux V5.10
- Oracle RAC 11gR2
- cluster which has 3 nodes (any number can exhibit this issue)
- The quorum, voting disks and Oracle database files are mounted via NFSv3
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.