Allow restart operations when aborted or failed

[juarezr]

Operations in backup or restore can be interrupted by:

1. Errors in source docs:
   • Documents without required fields
   • Corrupted documents
   • Generally when user has changed schema after imported docs
2. Server/Cluster problems:
   • Server overload
   • Huge transaction log when restoring (avoided by using --commit hard)
   • Replication throubles
   • Network problems between cluster instances
   • Out of memory in JVM / Solr
3. Interruptions:
   • Power failure, computer restart, process killed
   • Lost ssh connection
   • can avoided using the screen command before running solrcopy

Todo:

1. Create a mecanism for recording checkpoints for backup/restore:
   • Backup can require using the --order for checkpointing
   • Restore already has --order using zip filename that could be extended for checkpointing
2. Create a new switch like --continue that could:
   • detect when the command failed before
   • restart from the point interrupted
   • can skip in case of invalid/corrupted docs