Hi, >In general the repository shutdown should IMO *never* block for any >background tasks. A "kill -9" on a JVM process should IMHO be considered a >valid way to shutdown a repository running inside the JVM, so there's >nothing that the repository should expect a background task to finish in >time for shutdown.
I agree. Unless if we want to optimize for the 'quick startup' case (I don't think we should at this stage). For a quick startup, it does make sense to wait until the background threads, for example indexing, have finished. If not, it will (has to) still work, but startup is slower because the indexing has to be done at that time, and at that time the nodes are no longer in the cache. But I think we should wait with such optimizations until we have a good reason. Regards, Thomas
