HP-UX SNAplus2 R7 Administration Guide

Appendix DUsing SNAplus2 in a High Availability Environment
Notice that this shell script uses the return code of snapmon to determine what action to take. If snapmon returns
0, the TRLS link station was active at some point while snapmon was running. If that is the case, it makes sense
to simply attempt to restart the LS. If snapmon returns with a non-zero value, however, the LS failed to start for
the entire initialization period (about 30 seconds). If that happens, the script quits trying to perform a local restart of
the Token Ring LS, and attempts to start the SDLC LS. When that LS fails, the script exits, to allow ServiceGuard
to perform a remote failover.
Notice also that this solution does not provide the best level of application transparency. Since two different
SNAplus2 LSs are being used on this server, each application must have the ability to access LUs that are con
figured
for each LS. For 3270, 3179G, and LUA, LU pools are often used for this purpose. APPC transaction programs
would need to be designed and coded to be able to communicate with a remote TP over multiple LUs.
Local recovery is an excellent complement to the remote failover functions provided by ServiceGuard. By adding
local restart and local failover commands to your customized Package Control Script, you equip your SNAplus2
server for maximum uptime in a high availability environment.
227