Node Data Integrity Problem

Support staff receives email notification that "file is older than 15 minutes".

A ps -efw | grep vwpoint shows that very few if any of the agent processes are running. However, the jobInitiater and serviceQueue processes are still running.

First, examine the jobInit.log file under $VIEWPOINT/log. (If logging is not enabled this file may not exist or it may be left over from the past. In either case, goto the bin folder and edit the startNetWatchNode.sh script to enable it, then restart the node software.

If you see that any of the quarter hour check cycles are not completing then this typically indicates a data integrity problem in the database.

SQL> select * from servicelist where serviceid=20799;

 SERVICEID SERVICETYPE     CHANGE D M
---------- ----------- ---------- - -
     20799           4          0 N N

SQL> select * from servicenode where serviceid=20799;       

 SERVICEID     NODEID EXPIREDAT   INTERVAL
---------- ---------- --------- ----------
     20799          2 01-JAN-25         15
     20799          4 01-JAN-25         15
     20799          9 01-JAN-25         15
     20799         32 01-JAN-25         15
     20799         33 01-JAN-25         15

SQL> delete from servicenode where serviceid=20799;

5 rows deleted.

SQL> delete from servicelist where serviceid=20799;

1 row deleted.

SQL> commit;

Commit complete.

Here's an example of how to cleanup a HTTP Transaction

SQL> delete from tservicelist where serviceid=22224;

9 rows deleted.

SQL> delete from httptservice where serviceid=22224;

1 row deleted.

SQL> delete from servicenode where serviceid=22224;

5 rows deleted.

SQL> delete from servicelist where serviceid=22224;

1 row deleted.

SQL> commit;

Commit complete.

SQL> quit

After cleaning up all data integrity problems you will need to restart the jobInitiater.

gnwstop

gnwstart

Node Data Integrity Problem (last edited 2006-10-05 16:44:45 by mark)