As far the scalability of Bladelogic server is concerned, the calculation for patching job should be same as any other job, since the patching jobs execute the tasks asynchronously. Depending on the size of your server, I am sure you would already have max jobs, max parallel work items, maxlightworkitemthreads etc. parameters set. I would recommend looking at: https://docs.bmc.com/docs/display/public/bsa86/Performance%2CScalability%2CandSizing?src=search <https://docs.bmc.com/docs/display/public/bsa86/Performance%2CScalability%2CandSizing?src=search&src=search> &src=search
However, the patching jobs are do transfer lot of payloads - both analysis as well as deploy. And if the targets machines to be patched are spread out across different locations, then you may want to plan your patch repositories and configure standard repeaters https://docs.bmc.com/docs/pages/releaseview.action?pageId=489032259 <https://docs.bmc.com/docs/pages/releaseview.action?pageId=489032259&src=search> &src=search, so that the payload transfers that would take the bulk of time as well as bandwidth can be reduced.
Just to add - Depending patch state of the machines that are part of same patch analysis job run, you may end up generating large number of deploy jobs that you will execute to deploy missing patches. Keeping MaxJobs to higher value will help with parallelism bringing down the total deploy time for all the servers that are apart of same maintenance windows.
Also to add, as I presume you are using advanced deploy jobs so you can simulate and stage to have all content staged on targets before commit so when patching window starts you only would run commit phase. If you are not doing it this way you should consider using it.
Thanks All, I will take a look at all suggestions.
Unfortunately I am a complete novice to the world of BSA and I think this is going to take some time to understand.
So would I be looking at multi site Patching repositories and pre-stageing the targets for best multi site performance?