IBM Tivoli Enterprise Console (TEC)
適用於: Operations Manager 2007 R2
This section covers problems specific to the IBM Tivoli Connector.
RDBMS Connect Error on TEC-UNIX
Run any W-COMMAND on the UNIX-based TEC provider computer. If you get an error stating “Cannot connect to RDBMS, check that the RDBMS is running.
To determine the RDBMS name and whether it is running, enter the command ODADMIN ODLIST.
The important part of the command output will be similar to the following:
[scxsun02]:# odadmin odlist
RegionDisp Flags Port IPaddr Hostname(s)
14694898491 ct- 94 10.195.172.38 scxsun02.opsmgr.lan,scxsun02
2 ct- 94 10.195.172.26 opsmgr-sql.opsmgr.lan,opsmgr-sql
The RDBMS system is opsmgr-sql. The first entry in the Flags column can contain one of the following codes:
c – connection is active- – connection is inactive? – connection status is unknown
If the connection to the RDBMS is Inactive, restart the system or check the system status. Rerun the ODADMIN ODLIST command to make sure the connection is active.
Default Port Setting for Windows Provider
You need to change the EMS port for Windows provider for TEC to a default setting. Open the Connector service user interface and change the EMS port from blank to 5529, and then restart the Connector service.
Tivoli Enterprise Console Alert Exceeds Maximum Field Size
When you include Product Knowledge or Context in a TEC alert, in the TEC Connector Configuration dialog box, on the Operations Manager tab, reduce the Maximum field size setting to a lower number, such as 1,000. If these fields exceed 1,000 bytes, one of the following scenarios could occur:
The alert is forwarded to TEC.
TEC receives the alert.
Because the alert is too large, it will not pass the TEC rules.
The alert eventually reaches a MAX Retries status.
or
The alert is forwarded to TEC.
Because the alert is too large, it is dropped and, therefore, not created in TEC.
The alert eventually reaches a MAX Retries status.
To determine whether the alert size is the cause of a forwarding failure, do the following: On the TEC server, use the Tivoli command, wtdumprl, to determine whether the alert was received and PROCESSED. If the alert is too large, the TEC server might have dropped it.
If the event was created in TEC, but the alert TicketID property was never updated with the TEC event ID, do the following:
Examine the Tec_rule file for an error similar to:
Apr 8 17:41:11.326091 tec_rule[421948] ERR tec_msg.c:130: Packet size too small to contain message
The Tec_rule file is in the /Tmp directory on UNIX-based systems and in the /$DBDIR/Tmp directory on Windows-based computers.
Change the Field Size for Product Knowledge and Context
Open the TEC Connector Configuration dialog box.
Click the Operations Manager tab.
Set the setting in the Maximum field size box to 1,000 or less.