2 - Initial support for direct backups to remote storage. A sample
3 script is provided for backing up to Amazon S3. Other scripts
4 should be simple to write.
5 - Allow selected files/directories to be restored, not simply an
6 entire snapshot. Additionally, restore files in an order that
7 should optimize performance (restore files based on how they are
8 grouped into segments, instead of lexicographic order).
11 - SNAPSHOT FORMAT CHANGE: A few minor tweaks have been made to the
12 snapshot format. There is nothing substantial--this is primarily
13 a cleaning up of the format before use becomes more widespread.
14 An efficient and explicit means for representing sparse files has
15 been added to the format. Updated tools will be needed to
16 properly understand the new snapshots. Tools will continue to be
17 able to read the old snapshot format.
18 - LOCAL DATABASE CHANGE: The local database has been changed so that
19 segment utilization data for past snapshots is stored more
20 efficiently. The script in contrib/upgrade0.6-localdb.sql must be
21 run on a database to upgrade it prior to running backups with the
23 - The statcache implementation has been reworked, and the format
24 changed. The first backup will run more slowly since the old
25 statcache data will be ignored. New statcache data will be
26 written to statcache2. The old statcache file can be deleted.
27 - Metadata is shared between snapshots where possible. The
28 --full-metadata option can be used to disable this (completely
29 writing out all metadata with the new snapshot).
30 - Add an "intent" field to snapshots, which specifies informally how
31 long a snapshot is intended to be kept. For example, 1 can be
32 used for daily snapshots, and 7 for weekly snapshots. Segment
33 cleaning is now partly guided by intent values, but tuning of the
34 cleaning algorithms is not yet finished.
37 - Have the lbs-util tool check the version number when reading a
38 snapshot, and signal an error if the format is not recognized.
39 - Include a sample script, contrib/parity-gen, for creating
40 RAID-like parity sets to recover from some segment corruption.
41 The par2 command (http://parchive.sourceforge.net/) is used to
42 actually generate the parity sets; the parity-gen script simply
43 automates maintaining the parity sets.
46 - Much improved Python interface for accessing and manipulating LBS
47 archives and local database information. The interface should not
48 yet be considered completely stable.
49 - Python implementation of lbs-util now includes most of the
50 features of the Perl implementation, plus some other new features.
51 The Perl library and utility are deprecated and have been removed
53 - Preliminary snapshot restore support in lbs-util. This is still
54 not yet extensively tested.
57 - Documentation improvements: a getting started README, and a
58 description of some of the implementation details.
59 - Include an example filter script, lbs-filter-gpg, for encrypting
60 segment data as it is written.
61 - Add support for signed snapshot descriptor files (via the
62 --signature-filter option). Due to the checksums contained in the
63 snapshot descriptor, the signature covers the entire contents of
65 - Preview release of lbs-util.py, a Python-based program for
66 managing LBS snapshots. It currently implements a very simple
67 form of automatic segment cleaning, though this isn't much tested.
68 - Improved segment repacking support; after cleaning segments,
69 objects can be written out in different groups based on age. This
70 may help with long-term segment cleaning efficiency, but probably
74 - LOCAL DATABASE CHANGE: A checksums file is now written out along
75 with the snapshot descriptor that contains checksums of segments
76 used in a snapshot. This will allow quick integrity checks of a
77 snapshot, without needing to decompress or decrypt the segment
78 files. The local database schema was modified to store the
79 checksums for segments.
80 - Build dependence on libtar has been dropped. All necessary
81 support for writing TAR files is now directly included with the
83 - The snapshot format is now documented, at least in part. See
85 - Snapshots now include link count and inode number for files with
86 multiple hard links, so that a restore program could determine
87 which files should be hard linked on restore. The reference
88 restore.pl script does not use this.
89 - Bugfix: Print a help message instead of crashing if no files are
91 - Bugfix: File descriptors of files being backed up were closed
92 twice. In some cases this might have led to an unrelated file
93 being closed (if the file descriptor was quickly re-used).
94 - Preview of a new lbs-util command for maintaining snapshots.
95 Functionality is currently limited.
98 - Update reference restore.pl script to handle octal/hexadecimal
99 (format change introduced in 0.2).
100 - Better support for mixing multiple backup schemes: the statcache
101 file is now per-scheme, so performance should be better in cases
102 where multiple different backups (of different source directories)
103 are made, all sharing the same local database directory.
106 - Bugfix: Do not print an error when only one directory is specified
110 - SNAPSHOT FORMAT CHANGE: Metadata logs in backups can now contain
111 octal and hexadecimal values. Old restore programs may not be
112 able to read these new backups. The reference restore.pl script
113 has not yet been updated.
114 - LOCAL DATABASE CHANGE: Backups may be assigned a name with the
115 --scheme= command-line option, to allow different backup sets to
116 be mixed in the same directory. The database schema for the local
118 - Improved selection (inclusion/exclusion) of files to be included
120 - Device major/minor numbers are dumped for block and character