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