Configuring LT groups
- Configure LTs so that they share a common workload of outbound FIN messages
- Configure application LTs, that is, LTs that applications use, but that are independent of the LTs that are used for FIN sessions
- If the sessions for all the LTs in the group are active, each LT can process any message addressed to any of the LTs in the group.
- If any of the LTs fails, the remaining LTs continue to processes messages, including those addressed to a failed LT.
To avoid confusion, include the BIC8 of the destination in the LT group name. For example, if two LTs with the IDs XXXXUSNYA and XXXXUSNYB are to share a common outbound message workload, you might use an LT group with the name XXXXUSNY_GRP.
The LT group to which an LT belongs is determined by the value of the LTGroup attribute of the CO based on the CT DnfLT. If no such LTGroup attribute is defined, the LT is not part of an LT group.
- For the CO of type DnfLT that corresponds to the LT, specify only the LTGroup attribute.
- For the CO of type DNF_ILC_CMD that corresponds to the LT, specify only the RESET pseudo-attribute. This ensures that the application LT cannot be used to login or select a session.
By having your applications use only application LTs you can reconfigure the LTs used for FIN sessions without having to modify your applications. For example, two LTs with the ID XXXXUSNYA and XXXXUSNYB are members of an LT group with the name XXXXUSNY_GRP. Configure a new LT, XXXXUSNYX, which is never used to log in or select a session, and add this LT to the LT group XXXXUSNY_GRP. Messages from applications that specify LT XXXXUSNYX are processed whenever a session is active for either XXXXUSNYA or XXXXUSNYB.