From: Tom Wood (twood@rezlink.com)
Date: Mon Nov 11 2002 - 11:47:14 EST
Hello,
1. We are running a production 6M1 (4 proc, 10gb mem) and a
failover/dev 6H1 (4proc, 6gb mem), both connected to SSA enclosure via
dual paths, both running AIX 4.3.3 (ml10). The 6H1 has an expansion cab
for local storage when in a dev mode, whereas in an HACMP status it sees
the SSA. We are having a problem stopping the development instance and
getting the production db started. We believe the HACMP process is
complete and functioning, prior to adding the Oracle pieces.
Anticipated failover process. (if prod 6M1 is down, SSA up, 6H1 takes
over)
1. HACMP does its thing with IP addresses, Volume groups
2. Stop Oracle listener on 6H1 (dev).
3. Shutdown dev instance on 6H1
4. copy oratab.prod to oratab, profile.prod to .profile
5. issue lsnrctl start
6. issue dbstart
7. send notifications
HACMP question:
2. Due to the memory differences, what other Oracle files do I need
to be concerned with?
Non-related question:
1. One of the DBAs is trying to configure Oracle Intelligent Agent to
capture some performance statistics and is getting messages about snmp
as follows: NMS-00308: Failed to listen on address: another agent may
be running from dbsnmpc.log and NMS-00001: Warning: dbsnmp unable to
connect to SNMP master agent from dbsnmpw.log. We are not running any
Openview type app, just HACMP using snmp to my knowledge.
Are there any special configurations that need to be defined for
Oracle's IA app?
Tia
Tom Wood
Rezlink International
This archive was generated by hypermail 2.1.7 : Wed Apr 09 2008 - 22:16:19 EDT