Knapsack 1.3.2.0

• new: support for Elasticsearch 1.3.2
• new: all knapsack actions are reimplemented as Java API transport actions
• new: Elasticsearch bulk format support
• new: byte progress watcher, splitting into more than one archive file by byte size
• new: _push action for copying indices
• new: _pull endoint for fetching indices
• new: index aliases recorded in archive file
• new: archive codec API
• removed S3 support - use ES snapshot/restore for this
• added numerous junit tests
• switch to bzip2 implementation of https://code.google.com/p/jbzip2/
• switch to JDK ZIP archive implementation
• cleaned up tar implementation
• _state action overhaul
• _abort action overhaul

1.0.0.RC2.1

• update to Elasticsearch 1.0.0.RC2
• plugin version support
• bug fix: TransportClient now initializes only with elasticsearch-support plugin and ignores others
• added abort command to kill all ongoing imports or exports from REST

0.90.9.1

Changes:

• more archive support (ZIP, TAR, CPIO) and compression codecs (gzip, bzip2, lzf, xz)
• ES queries can be used to select content for archiving
• 'target' parameter renamed to 'path'
• new 'map' parameter for mapping index names and index/type names
• support for document meta fields (_parent, _routing, _version, _timestamp, _source)
• direct copy to local or remote cluster (endpoints _export/copy, _import/copy)
• optional AWS S3 support (endpoints _export/s3, _import/s3)
• archive entry names have four components: index/type/id/fieldname, where fieldname can contain arbitray (stored) fields, not only _source