IA Database setup

[irshanuddin]

We are setting up our IA environment and I am unable to create a database connection for IA to use for analysis. The IADB is an Oracle database on RAC. I believe RAC setups have no SIDs. I mentioned IADB(Name of database) in "Database Name" and it said it couldnt find the database. I tried using the ServiceName in "Database Name" and it errored saying this SID was not recognized by the Oracle Server. I have created the entries in .odbc.ini and uvconfig files in both AnalyzerProject and DSEngine. Any ideas are appreciated.

Here's my TNS:
XX11PD.MTDPRODUCTS.COM =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = chfoh770.xxx.com)
(PORT = 1521))
(LOAD_BALANCE = YES)
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = XX11PD11G.xxxPRODUCTS.COM)
(FAILOVER_MODE =
(TYPE = SELECT)(METHOD = BASIC)(RETRIES = 180)(DELAY = 5)
)))

My odbc.ini entry:
[IADB]
Driver=/opt/IBM/InformationServer/Server/branded_odbc/lib/VMora24.so
Description=DataDirect Oracle Wire Protocol driver
ApplicationUsingThreads=1
ArraySize=60000
CachedCursorLimit=32
CachedDescLimit=0
CatalogIncludesSynonyms=1
CatalogOptions=0
DefaultLongDataBuffLen=1024
DescribeAtPrepare=0
EnableDescribeParam=0
EnableNcharSupport=0
EnableScrollableCursors=1
EnableStaticCursorsForLongData=0
EnableTimestampWithTimeZone=0
HostName=chfoh770.xxx.com
LocalTimeZoneOffset=
LockTimeOut=-1
LogonID=iadb
Password=iadb
PortNumber=1521
ProcedureRetResults=0
ServiceName=XX11PD11G.xxxPRODUCTS.COM
UseCurrentSchema=1

Thanks

[irshanuddin]

I found out that when database is a cluster, I can provide the instance name for one node and the database name and it worked.

Hope this helps someone!

[stuartjvnorton]

This question was bugging me a bit, so I had a look in the Planning, Installation and Configuration guide (p177) and came across a section on xmeta, etc on an Oracle RAC. Just about the first thing it said was:

IBM InfoSphere Information Server does not support Oracle RAC configurations of
the IBM InfoSphere Information Analyzer analysis database or the IBM InfoSphere
QualityStage Match Designer database.

By picking 1 node I think you've bypassed the RAC and all of the benefits of using the it.

That said, this pick 1 node workaround is used to allow you to put xmeta on a RAC in the first place, but you still need to fix it up later (later on in that same section of the doco).