summaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2012-12-16Prepare release version 1.3obnam-1.3Lars Wirzenius4-5/+6
2012-12-16Note less-functional fsckLars Wirzenius1-1/+3
2012-12-16Add dependencies to READMELars Wirzenius1-0/+9
2012-12-15Update progress reporting during file backupLars Wirzenius1-0/+4
This makes the progress report, especially the elapsed time, update even when backing up a large file that has had very few changes.
2012-12-07Reduce info level logging and bugfixLars Wirzenius3-25/+39
2012-12-07Fix remove_old_roots to handle a backup root that does not end in /Lars Wirzenius1-21/+35
Also, reduce INFO level logging, and report backed up files in addition to total number of files found.
2012-12-07Remove unnecessary info level loggingLars Wirzenius2-4/+4
2012-12-07More detailed progress reporting for backupLars Wirzenius2-1/+27
2012-12-03Merge fsck improvementsLars Wirzenius2-44/+88
2012-12-03Update NEWS about fsck improvementsLars Wirzenius1-0/+3
2012-12-03Complain if no backup roots have been givenLars Wirzenius2-2/+4
2012-12-03Add a small helper script to show meliae dumps usefullyLars Wirzenius1-0/+19
2012-12-02Fix creation of OSError instancesLars Wirzenius2-4/+4
2012-11-25Put work items from current one into queue in right placeLars Wirzenius1-10/+10
If we have a queue [A, B, C], and we start processing A, and A generates A' and A'', we want to process them before B. Thus we put A' and A'' to the head of the queue. We also (now) remove the head of the queue rather than the tail. The old code tried to do all of these things, but was majorly confused.
2012-11-25Exclude from test coverageLars Wirzenius1-1/+1
2012-11-24Do not restore setuid/setgid unless root or owner of fileLars Wirzenius2-1/+8
2012-11-24Restore setuid/setgid bits correctly as rootLars Wirzenius2-4/+7
Reported by Pavel Kokolemin.
2012-11-24Backup files in inode number orderLars Wirzenius2-1/+3
Patch by Christophe Vu-Brugier.
2012-11-24Obnam diff fix and manpage update for diffLars Wirzenius3-0/+12
2012-11-24Update NEWS about bug fixLars Wirzenius1-0/+2
2012-11-24Fix bug: set variable name correctlyLars Wirzenius1-0/+1
Patch from Peter Valdemar Mørch
2012-11-24Remove --fsck-skip-b-trees, add --fsck-skip-shared-b-treesLars Wirzenius1-7/+5
2012-11-24Check per-client B-tree only if client is not ignoredLars Wirzenius1-3/+4
2012-11-24Open forest in write-mode if fixingLars Wirzenius1-3/+3
2012-11-24Log fsck errors and warningsLars Wirzenius1-0/+2
2012-11-23Remove flushing of ttystatusLars Wirzenius1-2/+2
It's disturbing and makes things go slower.
2012-11-23Add --fsck-skip-files optionLars Wirzenius1-1/+7
2012-11-20Add --fsck-skip-per-client-b-trees settingLars Wirzenius1-1/+6
2012-11-19Add --fsck-skip-files optionLars Wirzenius1-1/+4
2012-11-19Optimise away a file metadata lookupLars Wirzenius1-6/+7
We always know the metadata when creating the class, so we can pass it in, rather than look it up again.
2012-11-19Fix how OSError gets raisedLars Wirzenius3-4/+6
2012-11-19Handle shared B-tree lookup errors during backupsLars Wirzenius6-7/+349
2012-11-19Update NEWS about bug fixLars Wirzenius1-0/+4
2012-11-19Revert log file timestamp changeLars Wirzenius1-39/+0
It's mainly useful for debugging, and otherwise just bloats the sizes of the log file.
2012-11-18Add --fsck-last-generation-only optionLars Wirzenius1-0/+5
2012-11-18Add missing import for larchLars Wirzenius1-0/+1
2012-11-17Improve work item naming for progress reportingLars Wirzenius1-2/+2
2012-11-17Fix counting of items in progress reportingLars Wirzenius1-6/+7
2012-11-17Add --fsck-ignore-client settingLars Wirzenius1-2/+8
2012-11-17Clarify name of generation checking taskLars Wirzenius1-1/+1
2012-11-17Add --fsck-skip-b-trees settingLars Wirzenius1-7/+12
2012-11-17Optimisation: don't retrieve generation list twiceLars Wirzenius1-3/+3
2012-11-17Put larch fsck work items into queueLars Wirzenius1-7/+8
2012-11-17Add fsck --fsck-ignore-chunks settingLars Wirzenius1-2/+9
This is useful to allow checking integrity of a repository without having to check all chunks, which is quite time consuming. I am going to be using this to find out why there are missing nodes in one of my repositories.
2012-11-17Fix handling of value-less extended attributesLars Wirzenius3-2/+32
2012-11-17Update NEWS with bug fixedLars Wirzenius1-0/+2
2012-11-17Make Obnam handle value-less extended attributesLars Wirzenius2-2/+30
What happens: * llistxattr returns a list of names * lgetxattr for one or more names returns ENODATA (ENOATTR) In other words, the name exists, but has no value. This is different from it having an empty value. This happens at least sometimes on a btrfs filesystem, but in my opinion should not be happening at all. However, since it does, I have changed Obnam to deal with this by pretending the name does not exist. It logs a warning, but otherwise it is as if the name had not been returned by llistxattr. It would not be acceptable to pretend the value is empty, I think. An empty value may have some significance that is different from a value-less name. (I'm happy to change this if it turns out to be the wrong decision.) I was unable to reproduce this until Vladimir made me a btrfs disk image, after which this was easy to debug. While debugging this, I also realised that lgetxattr in LocalFS creates an OSError wrongly. Fixed that as well. Reported-by: Vladimir Elisseev
2012-11-17Ignore shared B-tree lookup errors during backupLars Wirzenius1-2/+10
Because of inherent race conditions in Obnam's chosen locking model, the lookup may fail due to another Obnam instance updating the B-trees while we are looking up (or after we last opened the B-tree). This should not be a fatal error, so we ignore such errors.
2012-11-11Check node read/remove/rename eventsLars Wirzenius1-43/+131
2012-11-11Configure logging to use sub-second timestampsLars Wirzenius1-0/+39
generated by cgit v1.2.1 (git 2.18.0) at 2024-05-14 02:02:08 +0000