I've been using NSH a lot these days and I would like to have some brainstorming here.
Once I do some report out of NSH script, say I have CPU, disks usage and server FQDN and want to save this out to some CSV.
Running this kind of script using macros and "for" loop on 5.000 servers is tooo slow for me. Running this as runscript is fine, but the report never contains proper data since many servers may try to write at one time.
A solution might be run a script using macros (2nd type of script) and call another script via blcli against targets (parallelly) so we have fast results. Each server would write its own report and then the 1st script would collect all the results.
Have anyone wondered about it?
yes, that's generally how you would do this.
one type 1 script to run in parallel and collect the data.
one type 2 script that aggregates the results.
these could be in a batch job.