Oracle cluster resources using orainstance.sh fail when lockfile is removed from /tmp on Red Hat High Availaibility Cluster
Issue
- The cluster took action to relocate the Oracle service after failing to find the process called PMON managed by the cluster:
clurgmgrd: [30988]: <debug> Local ping to 192.168.254.157 succeeded
clurgmgrd: [30988]: <info> Executing /etc/cluster/ohasd.sh status
clurgmgrd[30988]: <notice> status on orainstance "PRD" returned 1 (generic error)
clurgmgrd[30988]: <warning> Some independent resources in service:SAP-DB-PRD failed; Attempting inline recovery
clurgmgrd[30988]: <notice> stop on orainstance "PRD" returned 128 (unspecified)
clurgmgrd[30988]: <warning> Inline recovery of service:SAP-DB-PRD failed
clurgmgrd[30988]: <notice> Stopping service service:SAP-DB-PRD
clurgmgrd[30988]: <notice> stop on orainstance "PRD" returned 128 (unspecified)
clurgmgrd: [30988]: <info> Executing /etc/cluster/ohasd.sh stop
clurgmgrd: [30988]: <info> Removing IPv4 address 192.168.254.157/27 from bond2
Environment
- Red Hat Enterprise Linux (RHEL), including:
- Red Hat Enterprise Linux Server 5 (with the High Availability Add on)
- Red Hat Enterprise Linux Server 6 (with the High Availability Add on)
- 2 or more cluster nodes in a Red Hat High Availability Cluster
-
Oracle database resource managed by the orainstance.sh resource script
- The orainstance resource may be configured with the listeners parameter blank. e.g:
<orainstance home="/oracle/PRD/144" name="PRD" listeners="" user="oracle"/> - files in /tmp getting cleared in some way. Examples include:
- Administrator manually running
rm -fr /tmp/* - Cron job or similar that removes files from /tmp
- tmpwatch clearing files from /tmp regularly
- Administrator manually running
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.