Can not start JDBC Sender Communication Channel

[daixiong_jiang3]

Hi Guys,

Some very strange things is happenning in MY PI 7.1.

Normally, when I stop then start the corresponding JDBC Sender CC in the CC monitor, the CC will try to get data from the data source, and the message text such as '3/10/11 2:35:27 PM Processing started' will be shown there under the 'cluster node' tab, but there are 4 of the JDBC SENDER CCs do not work, which means when I stop and start the CC, there is no interaction and no message shown.

I have 2 other JDBC SENDER CCs connected to the same data source and they work well. I don't know why these 4 CCs can not be started.

I have tried to test the 4 CCs in my development enviroment using the same connection parameters, they works fine in the development environment.

it seems very strange,so could any guys tell me any possible reasons? thanks very much.

Dario.

[Former Member]

Hi ,

What you can Do is check the following places...

1. Check if Communication channel is not locked.

29.How to delete manually hanging Enqueue-Locks

Q: During the OOM error the J2ee engine goes down while the lock in the enqueue server still persists.When the channel comes back during restart as the effect of OOM error it will not able to get the cluster wide enqueue lock as the previous lock remains there in the enqueue server.How do I resolve this?

A: The existing (hanging) enqueue server locks for the corresponding JDBC channels(JDBC Lock Arguments) can be deleted manually by following given below stated procedure:

1. Stop the corresponding JDBC channels.

2. Open Visual Administrator.

3. Route to correspondig server.

4. Expand services and go to Locking adapter.

5. In the locking adapter service go to Display locks tab and

click on refesh button then you will able to see the list

of all the available locks.

6. Select the existing (hanging) enqueue lock with JDBC Argument from the list.

7. Click on "Delete Selected Locks" button.

Follow the instructions and delete the lock related to the stopped communication channel. For the 7.1 versions, please note that the Lock Manager is under the /nwa/locks, or on the location "Availability and Performance Management -> Locks".

2. Check if Query Written in Select at Sender Channel is working fine and picking up any data .

Regards

Prabhat Sharma.

[daixiong_jiang3]

Hi Prabhat Sharma.

thanks for your reply.

I am using PI 7.1, and I can not see any locks related to JDBC in the NWA.

I think restart the server will solve the issue, which should be the last choice especially when we are in the production environment.

regards.

Dario

[Former Member]

Before Restarting the Server.

Try One more thing ..

Create a new CC with same parameter. Check if that is working ......

enclose it in ur scenario.

Regards

PS.

[daixiong_jiang3]

Hi,

I just deleted the scenario and create a scenario with new name, and finally i can see the message:

3/11/11 10:51:50 PM 86935d10-943a-40f7-17db-be93ae61a03f Polling interval started. Length: 21600.0 seconds

3/11/11 10:51:50 PM 86935d10-943a-40f7-17db-be93ae61a03f Processing finished successfully

3/11/11 10:51:50 PM 86935d10-943a-40f7-17db-be93ae61a03f Message with ID 86935d10-943a-40f7-17db- be93ae61a03f processed

3/11/11 10:50:12 PM Processing started

it works!!

the most important point is that I created the new scenario with different name,^-

thanks,everyone

dario

[Former Member]

Hi,

See Note 1483974 - File and JDBC sender adapter's retry not working after error

REgards,

Carme

[daixiong_jiang3]

Hi Carme,

Thanks for your reply but I don't think we need to apply this note now.

Because some other CCs also had errors occured before and they can be started manually or as scheduled.

But for the 4 CCs I mentioned above, I even can not start them manually, which might be another issue.

Just now I deleted one configuration scenario and then rebuilt it, I can see the CC in the monitor, no error information shown, but it still can not be started.

thanks.

Dario

[naveen_chichili]

Hi Dario,

Could you please confirm if the channel has not at all started or channel started and not able to poll the data?

If you are facing the problem with second case...the table which you are trying to access might be opened or some other might be using the same table . Could you please cross check these and let me know the status.

Cheers!!!!

Naveen.

[daixiong_jiang3]

Hi Naveen,

1. The table is OK because I can get data using the same connection parameters in my development

environment.

2. The status icon of the communication channel is green in the CC monitor, and the 'short log' column says 'Functioning'.

But when I manually start the CC, there is no response and no message, which means i can't see anything under the

title"Processing Details for Cluster Node Server 00 00_12345"

3. I have tried CPA refresh, restart the JDBC adapter, delete the configuration scenario and then rebuild it, but none of them

works.

4. the other 2 CCs connected to the same table work well, they can get data by manually start or as scheduled.

Regards

Dario

[naveen_chichili]

Hi Dario,

The reason why I pointed regarding the table is I have faced the same problem recently and struggled a lot to resolve this simple issue. the mistake I have done was I have opened the table in TOAD application and tried to poll the data from the table which I opened.

The channel was in green- functioning and was unable to poll the data .

Cheers!!!

Naveen.

[aashish_sinha]

Hi,

just a check, Open any of the sender JDBC in ID and go to change mode, and in description add a space, and save ,activate.

Clear SLD Cache and restart the Communication channel in Communication channel monitoring. Before Doing change in JDBC adapter stop that channel in CC monitoring.

Its funny but it works sometime. It happens coz of cache is not reflecting to the adapter. So small change and activation can refresh.

Regards

Aashish Sinha

Edited by: Aashish Sinha on Mar 10, 2011 7:44 AM

[daixiong_jiang3]

Hi,

thanks for your reply.

but it does not work, actually i have ever tried both the CPA cache refresh and restart the JDBC adapter, but the CC still can not be started.

thanks.

Dario