Async Processing and Bulk Import

Thumbnails

Added dedicated new factory to generate PDF thumbnails on the fly and not store it.

However:

• I did not go through the conversion service nor add cache
• adding caching should be easy but adds IO

We probably want to disable the listeners related to Thumbnails, starting with CheckBlobUpdateListener:

• did not find a clean way to do this
  • globally disable the lister: but we want the thing to still run for pictures or other files
  • there is a context flag map, but not easy to used from Stream Importer

Indexing

Need to choose between 2 approaches:

• Run import with all indexing disabled and then ReIndex
• Run import and index at the same time

Even if the first approach is likely to deliver the best throughput for import, I would expect the second approach to be more efficient and more realistic.

Other listeners

Are there other listeners we should consider tweaking for the initial import ?