Webmethods| JDBC Pools | Connection Towards Oracle OCI PDB database

,
1

We are Trying to connect to a oracle database in OCI. below is the connection string configured in Settings → JDBC pool → pool Alias

image
image1534×795 57.5 KB

Issue:

When we do a test connection for Pool alias, sometime we see that the Test is successful and sometime we notice the below error

Test of test_aliasforstore failed [SoftwareAG][Oracle JDBC Driver]null

The Stack trace details are as follows:

com.wm.dd.jdbc.oraclebase.ft: [SoftwareAG][Oracle JDBC Driver]null
at com.wm.dd.jdbc.oraclebase.BaseExceptions.b(BaseExceptions.java:1014)
at com.wm.dd.jdbc.oraclebase.BaseExceptions.a(BaseExceptions.java:977)
at com.wm.dd.jdbc.oraclebase.BaseExceptions.b(BaseExceptions.java:1132)
at com.wm.dd.jdbc.oraclebase.BaseExceptions.b(BaseExceptions.java:800)
at com.wm.dd.jdbc.oracle.OracleImplConnection.open(OracleImplConnection.java:2551)
at com.wm.dd.jdbc.oraclebase.BaseConnection.b(BaseConnection.java:5086)
at com.wm.dd.jdbc.oraclebase.BaseConnection.connect(BaseConnection.java:1564)
at com.wm.dd.jdbc.oraclebase.ag.run(BaseConnectionStartup.java:247)

Observation when debugging with DB team:

      When we see the NULL error in webmethods during the test of the pool alias. We see the below error in alert.log of the database.

Fatal NI connect error 12637, connecting to:
(ADDRESS=(PROTOCOL=tcp)(HOST= ip address )(PORT=*****)(FIREWALL=OFF))

VERSION INFORMATION:
TNS for Linux: Version 19.0.0.0.0 - Production
Oracle Bequeath NT Protocol Adapter for Linux: Version 19.0.0.0.0 - Production
TCP/IP NT Protocol Adapter for Linux: Version 19.0.0.0.0 - Production
Version 19.25.0.0.0
Time: 11-MAR-2025 10:31:04
Tracing not turned on.
Tns error struct:
ns main err code: 12637

TNS-12637: Packet receive failed
ns secondary err code: 12532
nt main err code: 0
nt secondary err code: 0
nt OS err code: 0
2025-03-11T10:31:04.428345+01:00
opiodr aborting process unknown ospid (258759) as a result of ORA-609

Could it be a driver issue which is causing this intermittent issue? or does it have to do something with the database?

Version Details:

webMethods Integration Server 10.5:

Oracle Version : 19c

DataDirect Oracle driver version: com.wm.dd.jdbc.oracle;version=“6.0.0”

spy(Thread-850)(2025-03-11T09:18:23.809)>> Driver Name = Oracle
spy(Thread-850)(2025-03-11T09:18:23.809)>> Driver Version = 6.0.0.000304 (F000924.U000449)
spy(Thread-850)(2025-03-11T09:18:23.809)>> Database Name = Oracle
spy(Thread-850)(2025-03-11T09:18:23.809)>> Database Version = 19.0.0.0.0

2

Is it causing any problems other then test function? Do you see the same log every now and then?

3

we observed that after changing the connection string of the pool, the auditing and other internal functions(iscoreaudit etc…) of the IS were not working as expected. These Functions were continously going into fail fast mode, and then exiting fail fast…

so we did repetitive testing of the pool alias , and noticed that sometimes it successful and the sometimes we get the NULL error. And in database alert.log we see the TNS-12637 error which i started above.

4

Did you move the database or recreate it? What was the change? Try catalog function in DB configurator and see if there is anything missing. Also check if you are missing any patches or not. You are using an old version. It’s hard to tell if it is a bug or a missing patch with these information only. Do you have extended support agreement?

5

Hi Sandeep,

if I am not completely mistaken I remember that the DataDirect Driver is a Type 4 (Thin-Mode) Driver only and does not have the OCI-Mode (Type 2) available.
If you need to connect with OCI-Mode you will have to switch to the driver provided by Oracle.
This mode will require an additional native library, which will lead to the fact that the IS needs to be stopped and restarted to reload the driver. A simple reload of the WmJDBCAdapter package is no longer possible then.

IS this a JDBC Adapter 10.1 or 10.3?
Any Fixes applied to IS, ART, JDBC Adapter?

Regards,
Holger

6

Hi @Holger_von_Thomsen ,

The dd-cjdbc.jar in the 10.5 Integration server instance, where the intermittent issue is observed has the Version = 6.0.0.000304

We tried the similar test from the jdbc pools to OCI database from of one our 10.15 integration instance and the Connection there is stable. This instance has the datadirect driver version: 6.0.0.0112-0001 . we tried this in our lower environment

So we pulled this datadirect jar from 10.15 and installed it in 10.5 instance where the issue was occuring . Post reboot of the 10.5 instance loaded with new jar, the connection test were stable without any issues. so now we understand it was a driver issue.

Due to some complexities we are not able to immediately move from 10.5 to 10.15 but we would like to use this latest jar in 10.5

Since we have replaced the datadirect jar in 10.5 with latest one pulled from 10.15 and it is working, I would like to understand if there would be any issues in doing so any (like dependancy issues , or compatibility issues ) etc… or is the latest datadirect jar independant of the integration server instance ?

7

Hi Sandeep,

please check for the latest DDJ_10.5_DataDirect and TPP_10.5_DataDirect Fixes as well as for 10.15 respectively.

In the Readmes for these Fixes the Version of driver should be mentioned.
Both are using a DataDirect Driver Version 6.0.0.x, but 10.15 is using a much newer build than 10.5.
The version can also be retrieved via inspecting the META-INF/manifest.mf file in the Jar with any ZIP-compatible tool.

Regards,
Holger