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