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