Timesten JDBC error or Cache agent error.

Question

When I run a query in timesten with cache connect containing cast date, i came across with a problem
The query is :
SELECT ID, VALUE FROM Table WHERE (TRADE_DT BETWEEN cast ? as Date) - NUMTODSINTERVAL((90+1),'DAY') AND ?
Both parameters are set to Date.
When I use cache connect connecting
Oracle9i Enterprise Edition Release 9.2.0.6.0 - Production
There is a problem, the problem is:
[TimesTen][TimesTen 7.0.3.0.0 ODBC Driver][TimesTen]TT5107: Oracle(OCI) error in OCIStmtExecute(): ORA-25137: Data value out of range rc = -1 -- file "bdbStmt.c", lineno 1535, procedure "getOraOutTypesNLengths()"
I excute this query using JDBC, and set the parameters at runtime. It seems when passing this query to Oracle(I use cache connect), this query cannont execute successfully.
But when I use cache connect connecting
Oracle Database 10g Enterprise Edition Release 10.2.0.3.0 - 64bi
These is no such problem.
When I fill the ? field in this query with to_date('2007-10-01','yyyy-mm-dd') and execute directly in Oracle and using timesten command line, and even JDBC, all executions are successful.
I don't know why this happen.
Is it because of the oracle client and server version? Or the JDBC cast support?
Can anyone tell me? thanks very much.

Answer 1

I think the additional problem here is that the datastore invalidated. The "Could not start Cache agent..." message is unhelpful as it doesn't give you the PID of the Cache Agent process, but I think it's likely to be PID 6235. This says:
14:34:33.08 Warn: CAC: 6235: TT44023-6235--66038016-heduler01890: TTCACHE Cache agent was successfully able to connect to Oracle after 2 failed attempts
but then the datastore is invalidated here:
14:36:01.30 Warn: : 51448: 51452 exited while connected to data store '/var/opt/timesten/smsgw/info/ttcache' shm 470024242 count=11
14:36:01.30 Warn: : 51448: daRecovery: subdaemon 51452, managing data store, failed: invalidate (failcode=202)
It looks like a number of subdaemon processes were forcibly killed with kill -9 around this time, which would explain why 51452 exited uncleanly and the datastore had to be invalidated.
A couple of its threads then disconnect here:
14:36:02.40 Warn: : 51448: 6235/0x7fcf503b31f0: Forced Disconnect /var/opt/timesten/smsgw/info/ttcache
14:36:02.40 Warn: : 51448: 6235 ----------: Disconnecting from an old instance
14:36:02.90 Warn: : 51448: 6235/0x7fcf4c014810: Forced Disconnect /var/opt/timesten/smsgw/info/ttcache
14:36:02.90 Warn: : 51448: 6235 ----------: Disconnecting from an old instance
before this Warning is thrown:
14:36:03.30 Warn: : 51448: Could not start Cache agent for /var/opt/timesten/smsgw/info/ttcache (Cache agent already active for data store): setting to 'manual' policy
I haven't been able to find a specific reference to what triggers this warning to be issued, though there is an unfixed internal bug in this area:
Bug 9030650 : CACHE AGENT IS NOT ALWAYS RESTARTED AFTER INVALIDATION
Are you able to log an SR with a reference to this discussion, so it can be formally discussed with Development? If you do, please include the ttmesg.log files which cover this period as well.