Lizhong Li wrote: > Anurag, > > I got some odd results while debuging the test cases. It seems like that > 'destroy -a' doesn't work correctly in command-file mode. > > Here is the script file using for 'nwamcfg -f ...' > > :exp$ cat nwamadm_create_all.s > destroy -a > create enm myenm1 > set start="/var/tmp/start_myenm1" > set stop="/var/tmp/stop_myenm1" > set activation-mode=manual > end > create enm myenm2 > set start="/var/tmp/start_myenm2" > set stop="/var/tmp/stop_myenm2" > set activation-mode=manual > end > > firering:exp# nwamcfg -f nwamadm_create_all.s > Configuration read. > firering:exp# nwamadm list > TYPE PROFILE STATE > ncp User inactive > ncp Automatic active > loc myloc1 inactive > enm myenm1 inactive > enm myenm2 inactive > firering:exp# > firering:exp# nwamadm enable myenm1 > Enabling enm 'myenm1' > firering:exp# nwamadm list > TYPE PROFILE STATE > ncp User inactive > ncp Automatic active > loc myloc1 inactive > enm myenm2 inactive > enm myenm1 active > > These are right. > > firering:exp# nwamcfg -f nwamadm_create_all.s > Configuration read. > firering:exp# nwamadm list > TYPE PROFILE STATE > ncp User inactive > ncp Automatic active > loc myloc1 inactive > enm myenm1 active > enm myenm2 inactive > > Now the issue is here, it seems like 'destroy -a' in > nwamadm_create_all.s is not effective, but after about 10 secs the > 'myenm1' turns into 'inactive', why ? > Took me a while to realize that myloc1 should be showing up in "nwamadm list". I've known about this issue. When objects are destroyed, nwamd isn't sent the request. This task is already in the list. When objects are destroyed, nwamd should (and soon will) _fini() and remove that object.
Another task for nwamd is when objects are renamed. Altough rename isn't available in CLI, the functionality exists in the library. Darren, I think the GUI can rename objects right? Isn't that how you found that renames don't work at MPK? Anurag
