6 - !!! Cannot restore two jobs a the same time that were
7 written simultaneously unless they were totally spooled.
8 - Document cleaning up the spool files:
9 db, pid, state, bsr, mail, conmsg, spool
10 - Document the multiple-drive-changer.txt script.
11 - Pruning with Admin job.
12 - Does WildFile match against full name? Doc.
13 - %d and %v only valid on Director, not for ClientRunBefore/After.
14 - During tests with the 260 char fix code, I found one problem:
15 if the system "sees" a long path once, it seems to forget it's
16 working drive (e.g. c:\), which will lead to a problem during
17 the next job (create bootstrap file will fail). Here is the
18 workaround: specify absolute working and pid directory in
19 bacula-fd.conf (e.g. c:\bacula\working instead of
21 - Document techniques for restoring large numbers of files.
22 - Document setting my.cnf to big file usage.
23 - Add example of proper index output to doc. show index from File;
24 - Correct the Include syntax in the m4.xxx files in examples/conf
25 - Document JobStatus and Termination codes.
26 - Fix the error with the "DVI file can't be opened" while
27 building the French PDF.
28 - Document more DVD stuff
36 - Document all the little details of setting up certificates for
37 the Bacula data encryption code.
38 - Document more precisely how to use master keys -- especially
39 for disaster recovery.
42 - Migration from other vendors
46 - Backup conf/exe (all daemons)
47 - Backup up system state
48 - Detect state change of system (verify)
51 - How does restore JobId=nnn work? (Dirk)
52 - What does: restore select fileset="TestSet" client="workplay-fd" pool="Default"
53 > storage="File2" before="2007-02-05 23:05:04" do? (Dirk)
55 - Implement update jobid=
56 - Mention Eric and Marc's work + Marc's doc.
58 - When Pool specifies Storage command override does not work.
59 - Implement wait_for_sysop() message display in wait_for_device(), which
60 now prints warnings too often.
62 - Ensure that each device in an Autochanger has a different
64 - Add Catalog = to Pool resource so that pools will exist
65 in only one catalog -- currently Pools are "global".
66 - Look at sg_logs -a /dev/sg0 for getting soft errors.
67 - btape "test" command with Offline on Unmount = yes
69 This test is essential to Bacula.
71 I'm going to write one record in file 0,
72 two records in file 1,
73 and three records in file 2
75 02-Feb 11:00 btape: ABORTING due to ERROR in dev.c:715
76 dev.c:714 Bad call to rewind. Device "LTO" (/dev/nst0) not open
77 02-Feb 11:00 btape: Fatal Error because: Bacula interrupted by signal 11: Segmentation violation
78 Kaboom! btape, btape got signal 11. Attempting traceback.
80 - Ensure that moving a purged Volume in ua_purge.c to the RecyclePool
82 - Why doesn't @"xxx abc" work in a conf file?
83 - Figure out some way to "automatically" backup conf changes.
84 - Add the OS version back to the Win32 client info.
85 - Restarted jobs have a NULL in the from field.
86 - Modify SD status command to indicate when the SD is writing
87 to a DVD (the device is not open -- see bug #732).
88 - Look at the possibility of adding "SET NAMES UTF8" for MySQL,
89 and possibly changing the blobs into varchar.
90 - Ensure that the SD re-reads the Media record if the JobFiles
91 does not match -- it may have been updated by another job.
92 - Look at moving the Storage directive from the Job to the
93 Pool in the default conf files.
95 - Test Volume compatibility between machine architectures
96 - Encryption documentation
97 - Wrong jobbytes with query 12 (todo)
98 - bacula-1.38.2-ssl.patch
99 - Bare-metal recovery Windows (todo)
103 - Average tape size from Eric
104 SELECT COALESCE(media_avg_size.volavg,0) * count(Media.MediaId) AS volmax, GROUP BY Media.MediaType, Media.PoolId, media_avg_size.volavg
105 count(Media.MediaId) AS volnum,
106 sum(Media.VolBytes) AS voltotal,
107 Media.PoolId AS PoolId,
108 Media.MediaType AS MediaType
110 LEFT JOIN (SELECT avg(Media.VolBytes) AS volavg,
111 Media.MediaType AS MediaType
113 WHERE Media.VolStatus = 'Full'
114 GROUP BY Media.MediaType
115 ) AS media_avg_size ON (Media.MediaType = media_avg_size.MediaType)
116 GROUP BY Media.MediaType, Media.PoolId, media_avg_size.volavg
120 - Add doc for bweb -- especially Installation
122 http://www.orangecrate.com/modules.php?name=News&file=article&sid=501
124 - FD-SD quick disconnect
125 - Despool attributes in separate thread
128 - Check why restore repeatedly sends Rechdrs between
129 each data chunk -- according to James Harper 9Jan07.
130 - Building the in memory restore tree is slow.
133 - Full at least once a month, ...
134 - Cancel Inc if Diff/Full running
135 - More intelligent re-run
136 - New/deleted file backup
138 - Incremental backup -- rsync, Stow
142 - Look at mondo/mindi
143 - Don't restore Solaris Door files:
144 #define S_IFDOOR in st_mode.
145 see: http://docs.sun.com/app/docs/doc/816-5173/6mbb8ae23?a=view#indexterm-360
146 - Make Bacula by default not backup tmpfs, procfs, sysfs, ...
147 - Fix hardlinked immutable files when linking a second file, the
148 immutable flag must be removed prior to trying to link it.
149 - Implement Python event for backing up/restoring a file.
150 - Change dbcheck to tell users to use native tools for fixing
151 broken databases, and to ensure they have the proper indexes.
152 - add udev rules for Bacula devices.
153 - If a job terminates, the DIR connection can close before the
154 Volume info is updated, leaving the File count wrong.
155 - Look at why SIGPIPE during connection can cause seg fault in
156 writing the daemon message, when Dir dropped to bacula:bacula
157 - Look at zlib 32 => 64 problems.
158 - Possibly turn on St. Bernard code.
159 - Fix bextract to restore ACLs, or better yet, use common routines.
160 - Do we migrate appendable Volumes?
161 - Remove queue.c code.
162 - Print warning message if LANG environment variable does not specify
164 - New dot commands from Arno.
165 .show device=xxx lists information from one storage device, including
166 devices (I'm not even sure that information exists in the DIR...)
167 .move eject device=xxx mostly the same as 'unmount xxx' but perhaps with
168 better machine-readable output like "Ok" or "Error busy"
169 .move eject device=xxx toslot=yyy the same as above, but with a new
170 target slot. The catalog should be updated accordingly.
171 .move transfer device=xxx fromslot=yyy toslot=zzz
174 - Article: http://www.heise.de/open/news/meldung/83231
175 - Article: http://www.golem.de/0701/49756.html
176 - Article: http://lwn.net/Articles/209809/
177 - Article: http://www.onlamp.com/pub/a/onlamp/2004/01/09/bacula.html
178 - Article: http://www.linuxdevcenter.com/pub/a/linux/2005/04/07/bacula.html
179 - Article: http://www.osreviews.net/reviews/admin/bacula
180 - Article: http://www.debianhelp.co.uk/baculaweb.htm
182 - It appears to me that you have run into some sort of race
183 condition where two threads want to use the same Volume and they
184 were both given access. Normally that is no problem. However,
185 one thread wanted the particular Volume in drive 0, but it was
186 loaded into drive 1 so it decided to unload it from drive 1 and
187 then loaded it into drive 0, while the second thread went on
188 thinking that the Volume could be used in drive 1 not realizing
189 that in between time, it was loaded in drive 0.
190 I'll look at the code to see if there is some way we can avoid
191 this kind of problem. Probably the best solution is to make the
192 first thread simply start using the Volume in drive 1 rather than
193 transferring it to drive 0.
194 - Fix re-read of last block to check if job has actually written
195 a block, and check if block was written by a different job
196 (i.e. multiple simultaneous jobs writing).
197 - Figure out how to configure query.sql. Suggestion to use m4:
198 == changequote.m4 ===
199 changequote(`[',`]')dnl
200 ==== query.sql.in ===
201 :List next 20 volumes to expire
203 Pool.Name AS PoolName,
208 [ FROM_UNIXTIME(UNIX_TIMESTAMP(Media.LastWritten) Media.VolRetention) AS Expire, ])dnl
210 [ media.lastwritten + interval '1 second' * media.volretention as expire, ])dnl
214 ON Media.PoolId=Pool.PoolId
215 WHERE Media.LastWritten>0
219 Command: m4 -DmySQL changequote.m4 query.sql.in >query.sql
221 The problem is that it requires m4, which is not present on all machines
223 - Given all the problems with FIFOs, I think the solution is to do something a
224 little different, though I will look at the code and see if there is not some
225 simple solution (i.e. some bug that was introduced). What might be a better
226 solution would be to use a FIFO as a sort of "key" to tell Bacula to read and
227 write data to a program rather than the FIFO. For example, suppose you
232 Then, I could imagine if you backup and restore this file with a direct
233 reference as is currently done for fifos, instead, during backup Bacula will
236 /home/kern/my-fifo.backup
238 and read the data that my-fifo.backup writes to stdout. For restore, Bacula
241 /home/kern/my-fifo.restore
243 and send the data backed up to stdout. These programs can either be an
244 executable or a shell script and they need only read/write to stdin/stdout.
246 I think this would give a lot of flexibility to the user without making any
247 significant changes to Bacula.
252 select FilenameId from Filename where Name='';
253 # Get list of all directories referenced in a Backup.
254 select Path.Path from Path,File where File.JobId=nnn and
255 File.FilenameId=(FilenameId-from-above) and File.PathId=Path.PathId
256 order by Path.Path ASC;
258 - Look into using Dart for testing
259 http://public.kitware.com/Dart/HTML/Index.shtml
261 - Look into replacing autotools with cmake
262 http://www.cmake.org/HTML/Index.html
264 === Migration from David ===
265 What I'd like to see:
268 Name = "<poolname>-migrate"
272 Migration Selection Type = LowestUtil | OldestVol | PoolOccupancy |
273 Client | PoolResidence | Volume | JobName | SQLquery
274 Migration Selection Pattern = "regexp"
275 Next Pool = <override>
278 There should be no need for a Level (migration is always Full, since you
279 don't calculate differential/incremental differences for migration),
280 Storage should be determined by the volume types in the pool, and Client
281 is really a selection issue. Migration should always occur to the
282 NextPool defined in the pool definition. If no nextpool is defined, the
283 job should end with a reason of "no place to go". If Next Pool statement
284 is present, we override the check in the pool definition and use the
287 Here's how I'd define Migration Selection Types:
290 Client -- Migrate data from selected client only. Migration Selection
291 Pattern regexp provides pattern to select client names, eg ^FS00* makes
292 all client names starting with FS00 eligible for migration.
294 Jobname -- Migration all jobs matching name. Migration Selection Pattern
295 regexp provides pattern to select jobnames existing in pool.
297 Volume -- Migrate all data on specified volumes. Migration Selection
298 Pattern regexp provides selection criteria for volumes to be migrated.
299 Volumes must exist in pool to be eligible for migration.
303 LowestUtil -- Identify the volume in the pool with the least data on it
304 and empty it. No Migration Selection Pattern required.
306 OldestVol -- Identify the LRU volume with data written, and empty it. No
307 Migration Selection Pattern required.
309 PoolOccupancy -- if pool occupancy exceeds <highmig>, migrate volumes
310 (starting with most full volumes) until pool occupancy drops below
311 <lowmig>. Pool highmig and lowmig values are in pool definition, no
312 Migration Selection Pattern required.
316 SQLQuery -- Migrate all jobuids returned by the supplied SQL query.
317 Migration Selection Pattern contains SQL query to execute; should return
318 a list of 1 or more jobuids to migrate.
320 PoolResidence -- Migrate data sitting in pool for longer than
321 PoolResidence value in pool definition. Migration Selection Pattern
322 optional; if specified, override value in pool definition (value in
326 [ possibly a Python event -- kes ]
328 - Mount on an Autochanger with no tape in the drive causes:
329 Automatically selected Storage: LTO-changer
330 Enter autochanger drive[0]: 0
331 3301 Issuing autochanger "loaded drive 0" command.
332 3302 Autochanger "loaded drive 0", result: nothing loaded.
333 3301 Issuing autochanger "loaded drive 0" command.
334 3302 Autochanger "loaded drive 0", result: nothing loaded.
335 3902 Cannot mount Volume on Storage Device "LTO-Drive1" (/dev/nst0) because:
336 Couldn't rewind device "LTO-Drive1" (/dev/nst0): ERR=dev.c:678 Rewind error on "LTO-Drive1" (/dev/nst0). ERR=No medium found.
337 3905 Device "LTO-Drive1" (/dev/nst0) open but no Bacula volume is mounted.
338 If this is not a blank tape, try unmounting and remounting the Volume.
339 - If Drive 0 is blocked, and drive 1 is set "Autoselect=no", drive 1 will
341 - Autochanger did not change volumes.
342 select * from Storage;
343 +-----------+-------------+-------------+
344 | StorageId | Name | AutoChanger |
345 +-----------+-------------+-------------+
346 | 1 | LTO-changer | 0 |
347 +-----------+-------------+-------------+
348 05-May 03:50 roxie-sd: 3302 Autochanger "loaded drive 0", result is Slot 11.
349 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Warning: Director wanted Volume "LT
350 Current Volume "LT0-002" not acceptable because:
351 1997 Volume "LT0-002" not in catalog.
352 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Error: Autochanger Volume "LT0-002"
353 Setting InChanger to zero in catalog.
354 05-May 03:50 roxie-dir: Tibs.2006-05-05_03.05.02 Error: Unable to get Media record
356 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: Error getting Volume i
357 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: Job 530 canceled.
358 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: spool.c:249 Fatal appe
359 05-May 03:49 Tibs: Tibs.2006-05-05_03.05.02 Fatal error: c:\cygwin\home\kern\bacula
368 FirstWritten: 2006-05-05 03:11:54
369 LastWritten: 2006-05-05 03:50:23
370 LabelDate: 2005-12-26 16:52:40
381 VolRetention: 31,536,000
393 Note VolStatus is blank!!!!!
400 FirstWritten: 0000-00-00 00:00:00
401 LastWritten: 0000-00-00 00:00:00
402 LabelDate: 2005-12-26 16:52:40
413 VolRetention: 31,536,000
426 Automatically selected Storage: LTO-changer
427 Enter autochanger drive[0]: 0
428 3301 Issuing autochanger "loaded drive 0" command.
429 3302 Autochanger "loaded drive 0", result: nothing loaded.
430 3301 Issuing autochanger "loaded drive 0" command.
431 3302 Autochanger "loaded drive 0", result: nothing loaded.
432 3902 Cannot mount Volume on Storage Device "LTO-Drive1" (/dev/nst0) because:
433 Couldn't rewind device "LTO-Drive1" (/dev/nst0): ERR=dev.c:678 Rewind error on "LTO-Drive1" (/dev/nst0). ERR=No medium found.
435 3905 Device "LTO-Drive1" (/dev/nst0) open but no Bacula volume is mounted.
436 If this is not a blank tape, try unmounting and remounting the Volume.
438 - http://www.dwheeler.com/essays/commercial-floss.html
439 - Add VolumeLock to prevent all but lock holder (SD) from updating
440 the Volume data (with the exception of VolumeState).
441 - The btape fill command does not seem to use the Autochanger
442 - Make Windows installer default to system disk drive.
443 - Look at using ioctl(FIOBMAP, ...) on Linux, and
444 DeviceIoControl(..., FSCTL_QUERY_ALLOCATED_RANGES, ...) on
445 Win32 for sparse files.
446 http://www.flexhex.com/docs/articles/sparse-files.phtml
447 http://www.informatik.uni-frankfurt.de/~loizides/reiserfs/fibmap.html
448 - Directive: at <event> "command"
449 - Command: pycmd "command" generates "command" event. How to
450 attach to a specific job?
451 - Integrate Christopher's St. Bernard code.
452 - run_cmd() returns int should return JobId_t
453 - get_next_jobid_from_list() returns int should return JobId_t
454 - Document export LDFLAGS=-L/usr/lib64
455 - Don't attempt to restore from "Disabled" Volumes.
456 - Network error on Win32 should set Win32 error code.
457 - What happens when you rename a Disk Volume?
458 - Job retention period in a Pool (and hence Volume). The job would
460 - Look at -D_FORTIFY_SOURCE=2
461 - Add Win32 FileSet definition somewhere
462 - Look at fixing restore status stats in SD.
463 - Look at using ioctl(FIMAP) and FIGETBSZ for sparse files.
464 http://www.informatik.uni-frankfurt.de/~loizides/reiserfs/fibmap.html
465 - Implement a mode that says when a hard read error is
466 encountered, read many times (as it currently does), and if the
467 block cannot be read, skip to the next block, and try again. If
468 that fails, skip to the next file and try again, ...
470 create table LevelType (LevelType binary(1), LevelTypeLong tinyblob);
471 insert into LevelType (LevelType,LevelTypeLong) values
475 - Show files/second in client status output.
476 - Add a recursive mark command (rmark) to restore.
477 - "Minimum Job Interval = nnn" sets minimum interval between Jobs
478 of the same level and does not permit multiple simultaneous
479 running of that Job (i.e. lets any previous invocation finish
480 before doing Interval testing).
481 - Look at simplifying File exclusions.
482 - New directive "Delete purged Volumes"
483 - new pool XXX with ScratchPoolId = MyScratchPool's PoolId and
484 let it fill itself, and RecyclePoolId = XXX's PoolId so I can
485 see if it become stable and I just have to supervise
487 - If I want to remove this pool, I set RecyclePoolId = MyScratchPool's
488 PoolId, and when it is empty remove it.
489 - Figure out how to recycle Scratch volumes back to the Scratch Pool.
491 - Allow Check Labels to be used with Bacula labels.
492 - "Resuming" a failed backup (lost line for example) by using the
493 failed backup as a sort of "base" job.
495 - Email to the user when the tape is about to need changing x
496 days before it needs changing.
497 - Command to show next tape that will be used for a job even
498 if the job is not scheduled.
499 - From: Arunav Mandal <amandal@trolltech.com>
500 1. When jobs are running and bacula for some reason crashes or if I do a
501 restart it remembers and jobs it was running before it crashed or restarted
502 as of now I loose all jobs if I restart it.
504 2. When spooling and in the midway if client is disconnected for instance a
505 laptop bacula completely discard the spool. It will be nice if it can write
506 that spool to tape so there will be some backups for that client if not all.
508 3. We have around 150 clients machines it will be nice to have a option to
509 upgrade all the client machines bacula version automatically.
511 4. Atleast one connection should be reserved for the bconsole so at heavy load
512 I should connect to the director via bconsole which at sometimes I can't
514 5. Another most important feature that is missing, say at 10am I manually
515 started backup of client abc and it was a full backup since client abc has
516 no backup history and at 10.30am bacula again automatically started backup of
517 client abc as that was in the schedule. So now we have 2 multiple Full
518 backups of the same client and if we again try to start a full backup of
519 client backup abc bacula won't complain. That should be fixed.
521 - Fix bpipe.c so that it does not modify results pointer.
522 ***FIXME*** calling sequence should be changed.
523 - For Windows disaster recovery see http://unattended.sf.net/
524 - regardless of the retention period, Bacula will not prune the
525 last Full, Diff, or Inc File data until a month after the
526 retention period for the last Full backup that was done.
527 - update volume=xxx --- add status=Full
528 - Remove old spool files on startup.
529 - Exclude SD spool/working directory.
530 - Refuse to prune last valid Full backup. Same goes for Catalog.
532 - Make a callback when Rerun failed levels is called.
533 - Give Python program access to Scheduled jobs.
534 - Add setting Volume State via Python.
535 - Python script to save with Python, not save, save with Bacula.
536 - Python script to do backup.
538 - Change the Priority, Client, Storage, JobStatus (error)
539 at the start of a job.
540 - Why is SpoolDirectory = /home/bacula/spool; not reported
541 as an error when writing a DVD?
542 - Make bootstrap file handle multiple MediaTypes (SD)
543 - Remove all old Device resource code in Dir and code to pass it
544 back in SD -- better, rework it to pass back device statistics.
545 - Check locking of resources -- be sure to lock devices where previously
546 resources were locked.
547 - The last part is left in the spool dir.
550 - In restore don't compare byte count on a raw device -- directory
551 entry does not contain bytes.
555 MA = (last_MA * 3 + rate) / 4
556 rate = (bytes - last_bytes) / (runtime - last_runtime)
557 - Max Vols limit in Pool off by one?
558 - Implement Files/Bytes,... stats for restore job.
559 - Implement Total Bytes Written, ... for restore job.
560 - Despool attributes simultaneously with data in a separate
561 thread, rejoined at end of data spooling.
562 - 7. Implement new Console commands to allow offlining/reserving drives,
563 and possibly manipulating the autochanger (much asked for).
564 - Add start/end date editing in messages (%t %T, %e?) ...
565 - Add ClientDefs similar to JobDefs.
566 - Print more info when bextract -p accepts a bad block.
567 - Fix FD JobType to be set before RunBeforeJob in FD.
568 - Look at adding full Volume and Pool information to a Volume
569 label so that bscan can get *all* the info.
570 - If the user puts "Purge Oldest Volume = yes" or "Recycle Oldest Volume = yes"
571 and there is only one volume in the pool, refuse to do it -- otherwise
572 he fills the Volume, then immediately starts reusing it.
573 - Implement copies and stripes.
574 - Add history file to console.
575 - Each file on tape creates a JobMedia record. Peter has 4 million
576 files spread over 10000 tape files and four tapes. A restore takes
577 16 hours to build the restore list.
578 - Add and option to check if the file size changed during backup.
579 - Make sure SD deletes spool files on error exit.
580 - Delete old spool files when SD starts.
581 - When labeling tapes, if you enter 000026, Bacula uses
582 the tape index rather than the Volume name 000026.
583 - Add offline tape command to Bacula console.
585 Enter MediaId or Volume name: 32
586 Enter new Volume name: DLT-20Dec04
587 Automatically selected Pool: Default
588 Connecting to Storage daemon DLTDrive at 192.168.68.104:9103 ...
589 Sending relabel command from "DLT-28Jun03" to "DLT-20Dec04" ...
590 block.c:552 Write error at 0:0 on device /dev/nst0. ERR=Bad file descriptor.
591 Error writing final EOF to tape. This tape may not be readable.
592 dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
593 askdir.c:219 NULL Volume name. This shouldn't happen!!!
594 3912 Failed to label Volume: ERR=dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
595 Label command failed for Volume DLT-20Dec04.
596 Do not forget to mount the drive!!!
597 - Bug: if a job is manually scheduled to run later, it does not appear
598 in any status report and cannot be cancelled.
600 ==== Keeping track of deleted/new files ====
601 - To mark files as deleted, run essentially a Verify to disk, and
602 when a file is found missing (MarkId != JobId), then create
603 a new File record with FileIndex == -1. This could be done
604 by the FD at the same time as the backup.
606 My "trick" for keeping track of deletions is the following.
607 Assuming the user turns on this option, after all the files
608 have been backed up, but before the job has terminated, the
609 FD will make a pass through all the files and send their
610 names to the DIR (*exactly* the same as what a Verify job
611 currently does). This will probably be done at the same
612 time the files are being sent to the SD avoiding a second
613 pass. The DIR will then compare that to what is stored in
614 the catalog. Any files in the catalog but not in what the
615 FD sent will receive a catalog File entry that indicates
616 that at that point in time the file was deleted. This
617 either transmitted to the FD or simultaneously computed in
618 the FD, so that the FD can put a record on the tape that
619 indicates that the file has been deleted at this point.
620 A delete file entry could potentially be one with a FileIndex
621 of 0 or perhaps -1 (need to check if FileIndex is used for
622 some other thing as many of the Bacula fields are "overloaded"
625 During a restore, any file initially picked up by some
626 backup (Full, ...) then subsequently having a File entry
627 marked "delete" will be removed from the tree, so will not
628 be restored. If a file with the same name is later OK it
629 will be inserted in the tree -- this already happens. All
630 will be consistent except for possible changes during the
633 Since I'm on the subject, some of you may be wondering what
634 the utility of the in memory tree is if you are going to
635 restore everything (at least it comes up from time to time
636 on the list). Well, it is still *very* useful because it
637 allows only the last item found for a particular filename
638 (full path) to be entered into the tree, and thus if a file
639 is backed up 10 times, only the last copy will be restored.
640 I recently (last Friday) restored a complete directory, and
641 the Full and all the Differential and Incremental backups
642 spanned 3 Volumes. The first Volume was not even mounted
643 because all the files had been updated and hence backed up
644 since the Full backup was made. In this case, the tree
645 saved me a *lot* of time.
647 Make sure this information is stored on the tape too so
648 that it can be restored directly from the tape.
650 All the code (with the exception of formally generating and
651 saving the delete file entries) already exists in the Verify
652 Catalog command. It explicitly recognizes added/deleted files since
653 the last InitCatalog. It is more or less a "simple" matter of
654 taking that code and adapting it slightly to work for backups.
656 Comments from Martin Simmons (I think they are all covered):
657 Ok, that should cover the basics. There are few issues though:
659 - Restore will depend on the catalog. I think it is better to include the
660 extra data in the backup as well, so it can be seen by bscan and bextract.
662 - I'm not sure if it will preserve multiple hard links to the same inode. Or
663 maybe adding or removing links will cause the data to be dumped again?
665 - I'm not sure if it will handle renamed directories. Possibly it will work
666 by dumping the whole tree under a renamed directory?
668 - It remains to be seen how the backup performance of the DIR's will be
669 affected when comparing the catalog for a large filesystem.
673 How about introducing a Type = MgmtPolicy job type? That job type would
674 be responsible for scanning the Bacula environment looking for specific
675 conditions, and submitting the appropriate jobs for implementing said
679 Name = "Migration-Policy"
681 Policy Selection Job Type = Migrate
682 Scope = "<keyword> <operator> <regexp>"
683 Threshold = "<keyword> <operator> <regexp>"
684 Job Template = <template-name>
687 Where <keyword> is any legal job keyword, <operator> is a comparison
688 operator (=,<,>,!=, logical operators AND/OR/NOT) and <regexp> is a
689 appropriate regexp. I could see an argument for Scope and Threshold
690 being SQL queries if we want to support full flexibility. The
691 Migration-Policy job would then get scheduled as frequently as a site
692 felt necessary (suggested default: every 15 minutes).
697 Name = "Migration-Policy"
699 Policy Selection Job Type = Migration
701 Threshold = "Migration Selection Type = LowestUtil"
702 Job Template = "MigrationTemplate"
705 would select all pools for examination and generate a job based on
706 MigrationTemplate to automatically select the volume with the lowest
707 usage and migrate it's contents to the nextpool defined for that pool.
709 This policy abstraction would be really handy for adjusting the behavior
710 of Bacula according to site-selectable criteria (one thing that pops
711 into mind is Amanda's ability to automatically adjust backup levels
712 depending on various criteria).
718 - Add Pool/Storage override regression test.
719 - Add delete JobId to regression.
720 - Add a regression test for dbcheck.
721 - New test to add bscan to four-concurrent-jobs regression,
722 i.e. after the four-concurrent jobs zap the
723 database as is done in the bscan-test, then use bscan to
724 restore the database, do a restore and compare with the
726 - Add restore of specific JobId to regression (item 3
727 on the restore prompt)
728 - Add IPv6 to regression
729 - Add database test to regression. Test each function like delete,
732 - AntiVir can slow down backups on Win32 systems.
733 - Win32 systems with FAT32 can be much slower than NTFS for
734 more than 1000 files per directory.
738 - A HOLD command to stop all jobs from starting.
739 - A PAUSE command to pause all running jobs ==> release the
741 - Media Type = LTO,LTO-2,LTO-3
742 Media Type Read = LTO,LTO2,LTO3
743 Media Type Write = LTO2, LTO3
745 === From Carsten Menke <bootsy52@gmx.net>
747 Following is a list of what I think in the situations where I'm faced with,
748 could be a usefull enhancement to bacula, which I'm certain other users will
749 benefit from as well.
751 1. NextJob/NextJobs Directive within a Job Resource in the form of
752 NextJobs = job1,job2.
755 I currently solved the problem with running multiple jobs each after each
756 by setting the Max Wait Time for a job to 8 hours, and give
757 the jobs different Priorities. However, there scenarios where
758 1 Job is directly depending on another job, so if the former job fails,
759 the job after it needn't to be run
760 while maybe other jobs should run despite of that
763 A Backup Job and a Verify job, if the backup job fails there is no need to run
764 the verify job, as the backup job already failed. However, one may like
765 to backup the Catalog to disk despite of that the main backup job failed.
768 I see that this is related to the Event Handlers which are on the ToDo
769 list, also it is maybe a good idea to check for the return value and
770 execute different actions based on the return value
773 3. offline capability to bconsole
776 Currently I use a script which I execute within the last Job via the
777 RunAfterJob Directive, to release and eject the tape.
778 So I have to call bconsole "release=Storage-Name" and afterwards
779 mt -f /dev/nst0 eject to get the tape out.
781 If I have multiple Storage Devices, than these may not be /dev/nst0 and
782 I have to modify the script or call it with parameters etc.
783 This would actually not be needed, as everything is already defined
784 in bacula-sd.conf and if I can invoke bconsole with the
785 storage name via $1 in the script than I'm done and information is
788 4. %s for Storage Name added to the chars being substituted in "RunAfterJob"
792 For the reason mentioned in 3. to have the ability to call a
793 script with /scripts/foobar %s and in the script use $1
794 to pass the Storage Name to bconsole
796 5. Setting Volume State within a Job Resource
799 Instead of using "Maximum Volume Jobs" in the Pool Resource,
800 I would have the possibilty to define
801 in a Job Resource that after this certain job is run, the Volume State
802 should be set to "Volume State = Used", this give more flexibility (IMHO).
804 6. Localization of Bacula Messages
807 Unfortunatley many,many people I work with don't speak english very well.
808 So if at least the Reporting messages would be localized then they
809 would understand that they have to change the tape,etc. etc.
811 I volunteer to do the german translations, and if I can convince my wife also
812 french and Morre (western african language).
814 7. OK, this is evil, probably bound to security risks and maybe not possible
815 due to the design of bacula.
817 Implementation of Backtics ( `command` ) for shell comand execution to
818 the "Label Format" Directive.
822 Currently I have defined BACULA_DAY_OF_WEEK="day1|day2..." resulting in
823 Label Format = "HolyBackup-${BACULA_DAY_OF_WEEK[${WeekDay}]}". If I could
824 use backticks than I could use "Label Format = HolyBackup-`date +%A` to have
825 the localized name for the day of the week appended to the
826 format string. Then I have the tape labeled automatically with weekday
827 name in the correct language.
829 - Yes, that is surely the case. I probably should turn those into Warning
830 errors. In addition, you just made me think that it might not be bad to
831 add an option to check the file size after backing up the file and
832 report if it changes. This would be done as an option because it would
835 Kern, good idea. If you do do that, mention in the output: file
836 shrunk, or file expanded, just to make it obvious to the user
837 (without having to the refer to file size), just how the file size
840 Would this option be for all file, or just one file? Or a fileset?
841 - Make output from status use html table tags for nicely
842 presenting in a browser.
843 - Can one write tapes faster with 8192 byte block sizes?
844 - Document security problems with the same password for everyone in
845 rpm and Win32 releases.
846 - Browse generations of files.
847 - I've seen an error when my catalog's File table fills up. I
848 then have to recreate the File table with a larger maximum row
849 size. Relevant information is at
850 http://dev.mysql.com/doc/mysql/en/Full_table.html ; I think the
851 "Installing and Configuring MySQL" chapter should talk a bit
852 about this potential problem, and recommend a solution.
853 - For Solaris must use POSIX awk.
854 - Want speed of writing to tape while despooling.
855 - Supported autochanger:
863 Wangtek 6525ES (SCSI-1 QIC drive, 525MB), under Linux 2.4.something,
864 bacula 1.36.0/1 works with blocksize 16k INSIDE bacula-sd.conf.
865 - Add regex from http://www.pcre.org to Bacula for Win32.
866 - Use only shell tools no make in CDROM package.
867 - Include within include does it work?
868 - Implement a Pool of type Cleaning?
869 - Implement VolReadTime and VolWriteTime in SD
870 - Modify Backing up Your Database to include a bootstrap file.
871 - Think about making certain database errors fatal.
872 - Look at correcting the time jump in the scheduler for daylight
873 savings time changes.
874 - Add a "real" timer to network connections.
875 - Promote to Full = Time period
876 - Check dates entered by user for correctness (month/day/... ranges)
877 - Compress restore Volume listing by date and first file.
878 - Look at patches/bacula_db.b2z postgresql that loops during restore.
880 - Perhaps add read/write programs and/or plugins to FileSets.
881 - How to handle backing up portables ...
882 - Add some sort of guaranteed Interval for upgrading jobs.
883 - Can we write the state file after every job terminates? On Win32
884 the system crashes and the state file is not updated.
887 Documentation to do: (any release a little bit at a time)
888 - Doc to do unmount before removing magazine.
889 - Alternative to static linking "ldd prog" save all binaries listed,
890 restore them and point LD_LIBRARY_PATH to them.
891 - Document add "</dev/null >/dev/null 2>&1" to the bacula-fd command line
892 - Document query file format.
893 - Add more documentation for bsr files.
894 - Document problems with Verify and pruning.
895 - Document how to use multiple databases.
896 - VXA drives have a "cleaning required"
897 indicator, but Exabyte recommends preventive cleaning after every 75
900 In this context, it should be noted that Exabyte has a command-line
901 vxatool utility available for free download. (The current version is
902 vxatool-3.72.) It can get diagnostic info, read, write and erase tapes,
903 test the drive, unload tapes, change drive settings, flash new firmware,
905 Of particular interest in this context is that vxatool <device> -i will
906 report, among other details, the time since last cleaning in tape motion
907 minutes. This information can be retrieved (and settings changed, for
908 that matter) through the generic-SCSI device even when Bacula has the
909 regular tape device locked. (Needless to say, I don't recommend
910 changing tape settings while a job is running.)
911 - Lookup HP cleaning recommendations.
912 - Lookup HP tape replacement recommendations (see trouble shooting autochanger)
913 - Document doing table repair
916 ===================================
917 - Add macro expansions in JobDefs.
918 Run Before Job = "SomeFile %{Level} %{Client}"
919 Write Bootstrap="/some/dir/%{JobName}_%{Client}.bsr"
920 - Use non-blocking network I/O but if no data is available, use
922 - Use gather write() for network I/O.
923 - Autorestart on crash.
924 - Add bandwidth limiting.
925 - Add acks every once and a while from the SD to keep
926 the line from timing out.
927 - When an error in input occurs and conio beeps, you can back
928 up through the prompt.
929 - Detect fixed tape block mode during positioning by looking at
930 block numbers in btape "test". Possibly adjust in Bacula.
931 - Fix list volumes to output volume retention in some other
932 units, perhaps via a directive.
933 - Allow Simultaneous Priorities = yes => run up to Max concurrent jobs even
934 with multiple priorities.
935 - If you use restore replace=never, the directory attributes for
936 non-existent directories will not be restored properly.
938 - see lzma401.zip in others directory for new compression
940 - Allow the user to select JobType for manual pruning/purging.
941 - bscan does not put first of two volumes back with all info in
943 - Implement the FreeBSD nodump flag in chflags.
944 - Figure out how to make named console messages go only to that
945 console and to the non-restricted console (new console class?).
946 - Make restricted console prompt for password if *ask* is set or
947 perhaps if password is undefined.
948 - Implement "from ISO-date/time every x hours/days/weeks/months" in
951 ==== from Marc Schoechlin
952 - the help-command should be more verbose
953 (it should explain the paramters of the different
955 -> its time-comsuming to consult the manual anytime
956 you need a special parameter
957 -> maybe its more easy to maintain this, if the
958 descriptions of that commands are outsourced to
960 - the cd-command should allow complete paths
961 i.e. cd /foo/bar/foo/bar
962 -> if a customer mails me the path to a certain file,
963 its faster to enter the specified directory
964 - if the password is not configured in bconsole.conf
965 you should be asked for it.
966 -> sometimes you like to do restore on a customer-machine
967 which shouldnt know the password for bacula.
968 -> adding the password to the file favours admins
969 to forget to remove the password after usage
971 the protection of that file is less important
972 - long-listed-output of commands should be scrollable
973 like the unix more/less-command does
974 -> if someone runs 200 and more machines, the lists could
975 be a little long and complex
976 - command-output should be shown column by column
977 to reduce scrolling and to increase clarity
979 - lsmark should list the selected files with full
981 - wildcards for selecting and file and directories would be nice
982 - any actions should be interuptable with STRG+C
983 - command-expansion would be pretty cool
985 - When the replace Never option is set, new directory permissions
986 are not restored. See bug 213. To fix this requires creating a
987 list of newly restored directories so that those directory
988 permissions *can* be restored.
989 - Add prune all command
990 - Document fact that purge can destroy a part of a restore by purging
991 one volume while others remain valid -- perhaps mark Jobs.
992 - Add multiple-media-types.txt
993 - look at mxt-changer.html
994 - Make ? do a help command (no return needed).
995 - Implement restore directory.
996 - Document streams and how to implement them.
997 - Try not to re-backup a file if a new hard link is added.
998 - Add feature to backup hard links only, but not the data.
999 - Fix stream handling to be simpler.
1000 - Add Priority and Bootstrap to Run a Job.
1001 - Eliminate Restore "Run Restore Job" prompt by allowing new "run command
1003 - Remove View FileSet button from Run a Job dialog.
1004 - Handle prompt for restore job at end of Restore command.
1005 - Add display of total selected files to Restore window.
1006 - Add tree pane to left of window.
1007 - Add progress meter.
1008 - Max wait time or max run time causes seg fault -- see runtime-bug.txt
1009 - Add message to user to check for fixed block size when the forward
1010 space test fails in btape.
1011 - When unmarking a directory check if all files below are unmarked and
1012 then remove the + flag -- in the restore tree.
1013 - Possibly implement: Action = Unmount Device="TapeDrive1" in Admin jobs.
1014 - Setup lrrd graphs: (http://www.linpro.no/projects/lrrd/) Mike Acar.
1015 - Revisit the question of multiple Volumes (disk) on a single device.
1016 - Add a block copy option to bcopy.
1017 - Finish work on Gnome restore GUI.
1018 - Fix "llist jobid=xx" where no fileset or client exists.
1019 - For each job type (Admin, Restore, ...) require only the really necessary
1020 fields.- Pass Director resource name as an option to the Console.
1021 - Add a "batch" mode to the Console (no unsolicited queries, ...).
1022 - Add a .list all files in the restore tree (probably also a list all files)
1023 Do both a long and short form.
1024 - Allow browsing the catalog to see all versions of a file (with
1025 stat data on each file).
1026 - Restore attributes of directory if replace=never set but directory
1028 - Use SHA1 on authentication if possible.
1029 - See comtest-xxx.zip for Windows code to talk to USB.
1030 - Add John's appended files:
1031 Appended = { /files/server/logs/http/*log }
1032 and such files would be treated as follows.On a FULL backup, they would
1033 be backed up like any other file.On an INCREMENTAL backup, where a
1034 previous INCREMENTAL or FULL was already in thecatalogue and the length
1035 of the file wasgreater than the length of the last backup, only thedata
1036 added since the last backup will be dumped.On an INCREMENTAL backup, if
1037 the length of the file is less than thelength of the file with the same
1038 name last backed up, the completefile is dumped.On Windows systems, with
1039 creation date of files, we can be evensmarter about this and not count
1040 entirely upon the length.On a restore, the full and all incrementals
1041 since it will beapplied in sequence to restore the file.
1042 - Check new HAVE_WIN32 open bits.
1043 - Check if the tape has moved before writing.
1044 - Handling removable disks -- see below:
1045 - Keep track of tape use time, and report when cleaning is necessary.
1046 - Add FromClient and ToClient keywords on restore command (or
1047 BackupClient RestoreClient).
1048 - Implement a JobSet, which groups any number of jobs. If the
1049 JobSet is started, all the jobs are started together.
1050 Allow Pool, Level, and Schedule overrides.
1051 - Enhance cancel to timeout BSOCK packets after a specific delay.
1052 - Do scheduling by UTC using gmtime_r() in run_conf, scheduler, and
1053 ua_status.!!! Thanks to Alan Brown for this tip.
1054 - Look at updating Volume Jobs so that Max Volume Jobs = 1 will work
1055 correctly for multiple simultaneous jobs.
1056 - Correct code so that FileSet MD5 is calculated for < and | filename
1058 - Implement the Media record flag that indicates that the Volume does disk
1060 - Implement VolAddr, which is used when Volume is addressed like a disk,
1061 and form it from VolFile and VolBlock.
1062 - Make multiple restore jobs for multiple media types specifying
1063 the proper storage type.
1064 - Fix fast block rejection (stored/read_record.c:118). It passes a null
1065 pointer (rec) to try_repositioning().
1066 - Look at extracting Win data from BackupRead.
1067 - Implement RestoreJobRetention? Maybe better "JobRetention" in a Job,
1068 which would take precidence over the Catalog "JobRetention".
1069 - Implement Label Format in Add and Label console commands.
1070 - Possibly up network buffers to 65K. Put on variable.
1071 - Put email tape request delays on one or more variables. User wants
1072 to cancel the job after a certain time interval. Maximum Mount Wait?
1073 - Job, Client, Device, Pool, or Volume?
1074 Is it possible to make this a directive which is *optional* in multiple
1075 resources, like Level? If so, I think I'd make it an optional directive
1076 in Job, Client, and Pool, with precedence such that Job overrides Client
1077 which in turn overrides Pool.
1079 - New Storage specifications:
1080 - Want to write to multiple storage devices simultaneously
1081 - Want to write to multiple storage devices sequentially (in one job)
1082 - Want to read/write simultaneously
1083 - Key is MediaType -- it must match
1085 Passed to SD as a sort of BSR record called Storage Specification
1089 MediaType -> Next MediaType
1091 Device -> Next Device
1093 Allow multiple Storage specifications
1101 Allow Multiple Pool specifications (note, Pool currently
1103 Allow Multiple MediaType specifications in Dir conf
1104 Allow Multiple Device specifications in Dir conf
1105 Perhaps keep this in a single SSR
1106 Tie a Volume to a specific device by using a MediaType that
1107 is contained in only one device.
1108 In SD allow Device to have Multiple MediaTypes
1110 - Ideas from Jerry Scharf:
1111 First let's point out some big pluses that bacula has for this
1113 more importantly it's active. Thank you so much for that
1114 even more important, it's not flaky
1115 it has an open access catalog, opening many possibilities
1116 it's pushing toward heterogeneous systems capability
1118 Macintosh file client
1119 macs are an interesting niche, but I fear a server is a rathole
1120 working bare iron recovery for windows
1121 the option for inc/diff backups not reset on fileset revision
1122 a) use both change and inode update time against base time
1123 b) do the full catalog check (expensive but accurate)
1124 sizing guide (how much system is needed to back up N systems/files)
1125 consultants on using bacula in building a disaster recovery system
1126 an integration guide
1127 or how to get at fancy things that one could do with bacula
1128 logwatch code for bacula logs (or similar)
1129 linux distro inclusion of bacula (brings good and bad, but necessary)
1130 win2k/XP server capability (icky but you asked)
1131 support for Oracle database ??
1133 - Look at adding SQL server and Exchange support for Windows.
1134 - Make dev->file and dev->block_num signed integers so that -1 can
1135 be an invalid value which happens with BSR.
1136 - Create VolAddr for disk files in place of VolFile and VolBlock. This
1137 is needed to properly specify ranges.
1138 - Add progress of files/bytes to SD and FD.
1139 - Print warning message if FileId > 4 billion
1140 - do a "messages" before the first prompt in Console
1141 - Client does not show busy during Estimate command.
1142 - Implement Console mtx commands.
1143 - Implement a Mount Command and an Unmount Command where
1144 the users could specify a system command to be performed
1145 to do the mount, after which Bacula could attempt to
1146 read the device. This is for Removeable media such as a CDROM.
1147 - Most likely, this mount command would be invoked explicitly
1148 by the user using the current Console "mount" and "unmount"
1149 commands -- the Storage Daemon would do the right thing
1150 depending on the exact nature of the device.
1151 - As with tape drives, when Bacula wanted a new removable
1152 disk mounted, it would unmount the old one, and send a message
1153 to the user, who would then use "mount" as described above
1154 once he had actually inserted the disk.
1155 - Implement dump/print label to UA
1156 - Spool to disk only when the tape is full, then when a tape is hung move
1158 - bextract is sending everything to the log file ****FIXME****
1159 - Allow multiple Storage specifications (or multiple names on
1160 a single Storage specification) in the Job record. Thus a job
1161 can be backed up to a number of storage devices.
1162 - Implement some way for the File daemon to contact the Director
1163 to start a job or pass its DHCP obtained IP number.
1164 - Implement a query tape prompt/replace feature for a console
1165 - Copy console @ code to gnome2-console
1166 - Make tree walk routines like cd, ls, ... more user friendly
1167 by handling spaces better.
1168 - Make sure that Bacula rechecks the tape after the 20 min wait.
1169 - Set IO_NOWAIT on Bacula TCP/IP packets.
1170 - Try doing a raw partition backup and restore by mounting a
1172 - From Lars Kellers:
1173 Yes, it would allow to highly automatic the request for new tapes. If a
1174 tape is empty, bacula reads the barcodes (native or simulated), and if
1175 an unused tape is found, it runs the label command with all the
1176 necessary parameters.
1178 By the way can bacula automatically "move" an empty/purged volume say
1179 in the "short" pool to the "long" pool if this pool runs out of volume
1181 - What to do about "list files job=xxx".
1182 - Look at how fuser works and /proc/PID/fd that is how Nic found the
1183 file descriptor leak in Bacula.
1184 - Implement WrapCounters in Counters.
1185 - Add heartbeat from FD to SD if hb interval expires.
1186 - Can we dynamically change FileSets?
1187 - If pool specified to label command and Label Format is specified,
1188 automatically generate the Volume name.
1189 - Why can't SQL do the filename sort for restore?
1190 - Add ExhautiveRestoreSearch
1191 - Look at the possibility of loading only the necessary
1192 data into the restore tree (i.e. do it one directory at a
1193 time as the user walks through the tree).
1194 - Possibly use the hash code if the user selects all for a restore command.
1195 - Fix "restore all" to bypass building the tree.
1196 - Prohibit backing up archive device (findlib/find_one.c:128)
1197 - Implement Release Device in the Job resource to unmount a drive.
1198 - Implement Acquire Device in the Job resource to mount a drive,
1199 be sure this works with admin jobs so that the user can get
1200 prompted to insert the correct tape. Possibly some way to say to
1201 run the job but don't save the files.
1202 - Make things like list where a file is saved case independent for
1204 - Use autochanger to handle multiple devices.
1205 - Implement a Recycle command
1206 - Start working on Base jobs.
1207 - Implement UnsavedFiles DB record.
1208 - From Phil Stracchino:
1209 It would probably be a per-client option, and would be called
1210 something like, say, "Automatically purge obsoleted jobs". What it
1211 would do is, when you successfully complete a Differential backup of a
1212 client, it would automatically purge all Incremental backups for that
1213 client that are rendered redundant by that Differential. Likewise,
1214 when a Full backup on a client completed, it would automatically purge
1215 all Differential and Incremental jobs obsoleted by that Full backup.
1216 This would let people minimize the number of tapes they're keeping on
1217 hand without having to master the art of retention times.
1218 - When doing a Backup send all attributes back to the Director, who
1219 would then figure out what files have been deleted.
1220 - Currently in mount.c:236 the SD simply creates a Volume. It should have
1221 explicit permission to do so. It should also mark the tape in error
1222 if there is an error.
1223 - Cancel waiting for Client connect in SD if FD goes away.
1225 - Implement timeout in response() when it should come quickly.
1226 - Implement a Slot priority (loaded/not loaded).
1227 - Implement "vacation" Incremental only saves.
1228 - Implement create "FileSet"?
1229 - Add prefixlinks to where or not where absolute links to FD.
1230 - Issue message to mount a new tape before the rewind.
1231 - Simplified client job initiation for portables.
1232 - If SD cannot open a drive, make it periodically retry.
1233 - Add more of the config info to the tape label.
1235 - Refine SD waiting output:
1236 Device is being positioned
1237 > Device is being positioned for append
1238 > Device is being positioned to file x
1240 - Figure out some way to estimate output size and to avoid splitting
1241 a backup across two Volumes -- this could be useful for writing CDROMs
1242 where you really prefer not to have it split -- not serious.
1243 - Have SD compute MD5 or SHA1 and compare to what FD computes.
1244 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
1245 actual data on the Volume and compare it.
1246 - Implement Bacula plugins -- design API
1247 - Make bcopy read through bad tape records.
1248 - Program files (i.e. execute a program to read/write files).
1249 Pass read date of last backup, size of file last time.
1250 - Add Signature type to File DB record.
1251 - CD into subdirectory when open()ing files for backup to
1252 speed up things. Test with testfind().
1253 - Priority job to go to top of list.
1254 - Why are save/restore of device different sizes (sparse?) Yup! Fix it.
1255 - Implement some way for the Console to dynamically create a job.
1256 - Solaris -I on tar for include list
1257 - Need a verbose mode in restore, perhaps to bsr.
1258 - bscan without -v is too quiet -- perhaps show jobs.
1259 - Add code to reject whole blocks if not wanted on restore.
1260 - Check if we can increase Bacula FD priorty in Win2000
1261 - Make sure the MaxVolFiles is fully implemented in SD
1262 - Check if both CatalogFiles and UseCatalog are set to SD.
1263 - Possibly add email to Watchdog if drive is unmounted too
1264 long and a job is waiting on the drive.
1265 - After unmount, if restore job started, ask to mount.
1266 - Add UA rc and history files.
1267 - put termcap (used by console) in ./configure and
1268 allow -with-termcap-dir.
1269 - Fix Autoprune for Volumes to respect need for full save.
1270 - Compare tape to Client files (attributes, or attributes and data)
1271 - Make all database Ids 64 bit.
1272 - Allow console commands to detach or run in background.
1273 - Add SD message variables to control operator wait time
1274 - Maximum Operator Wait
1275 - Minimum Message Interval
1276 - Maximum Message Interval
1277 - Send Operator message when cannot read tape label.
1278 - Verify level=Volume (scan only), level=Data (compare of data to file).
1279 Verify level=Catalog, level=InitCatalog
1281 - Add keyword search to show command in Console.
1282 - Events : tape has more than xxx bytes.
1283 - Complete code in Bacula Resources -- this will permit
1284 reading a new config file at any time.
1285 - Handle ctl-c in Console
1286 - Implement script driven addition of File daemon to config files.
1287 - Think about how to make Bacula work better with File (non-tape) archives.
1288 - Write Unix emulator for Windows.
1289 - Put memory utilization in Status output of each daemon
1290 if full status requested or if some level of debug on.
1291 - Make database type selectable by .conf files i.e. at runtime
1292 - Set flag for uname -a. Add to Volume label.
1293 - Restore files modified after date
1294 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
1295 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
1296 - Timout a job or terminate if link goes down, or reopen link and query.
1297 - Concept of precious tapes (cannot be reused).
1298 - Make bcopy copy with a single tape drive.
1299 - Permit changing ownership during restore.
1302 > My suggestion: Add a feature on the systray menu-icon menu to request
1303 > an immediate backup now. This would be useful for laptop users who may
1304 > not be on the network when the regular scheduled backup is run.
1306 > My wife's suggestion: Add a setting to the win32 client to allow it to
1307 > shut down the machine after backup is complete (after, of course,
1308 > displaying a "System will shut down in one minute, click here to cancel"
1309 > warning dialog). This would be useful for sites that want user
1310 > woorkstations to be shut down overnight to save power.
1313 - Autolabel should be specified by DIR instead of SD.
1315 - Add media capacity
1316 - AutoScan (check checksum of tape)
1317 - Format command = "format /dev/nst0"
1321 - Seek resolution (usually corresponds to buffer size)
1322 - EODErrorCode=ENOSPC or code
1323 - Partial Read error code
1324 - Partial write error code
1325 - Nonformatted read error
1326 - Nonformatted write error
1327 - WriteProtected error
1331 - IgnoreCloseErrors=yes
1341 - FD sends unsaved file list to Director at end of job (see
1343 - File daemon should build list of files skipped, and then
1344 at end of save retry and report any errors.
1345 - Write a Storage daemon that uses pipes and
1346 standard Unix programs to write to the tape.
1348 - Need something that monitors the JCR queue and
1349 times out jobs by asking the deamons where they are.
1350 - Enhance Jmsg code to permit buffering and saving to disk.
1351 - device driver = "xxxx" for drives.
1352 - Verify from Volume
1353 - Ensure that /dev/null works
1354 - Need report class for messages. Perhaps
1355 report resource where report=group of messages
1356 - enhance scan_attrib and rename scan_jobtype, and
1357 fill in code for "since" option
1358 - Director needs a time after which the report status is sent
1359 anyway -- or better yet, a retry time for the job.
1360 - Don't reschedule a job if previous incarnation is still running.
1361 - Some way to automatically backup everything is needed????
1362 - Need a structure for pending actions:
1364 - termination status (part of buffered msgs?)
1366 Read, Write, Clean, Delete
1367 - Login to Bacula; Bacula users with different permissions:
1368 owner, group, user, quotas
1369 - Store info on each file system type (probably in the job header on tape.
1370 This could be the output of df; or perhaps some sort of /etc/mtab record.
1372 ========= ideas ===============
1373 From: "Jerry K. Schieffer" <jerry@skylinetechnology.com>
1374 To: <kern@sibbald.com>
1375 Subject: RE: [Bacula-users] future large programming jobs
1376 Date: Thu, 26 Feb 2004 11:34:54 -0600
1378 I noticed the subject thread and thought I would offer the following
1379 merely as sources of ideas, i.e. something to think about, not even as
1380 strong as a request. In my former life (before retiring) I often
1381 dealt with backups and storage management issues/products as a
1382 developer and as a consultant. I am currently migrating my personal
1383 network from amanda to bacula specifically because of the ability to
1384 cross media boundaries during storing backups.
1385 Are you familiar with the commercial product called ADSM (I think IBM
1386 now sells it under the Tivoli label)? It has a couple of interesting
1387 ideas that may apply to the following topics.
1389 1. Migration: Consider that when you need to restore a system, there
1390 may be pressure to hurry. If all the information for a single client
1391 can eventually end up on the same media (and in chronological order),
1392 the restore is facillitated by not having to search past information
1393 from other clients. ADSM has the concept of "client affinity" that
1394 may be associated with it's storage pools. It seems to me that this
1395 concept (as an optional feature) might fit in your architecture for
1398 ADSM also has the concept of defining one or more storage pools as
1399 "copy pools" (almost mirrors, but only in the sense of contents).
1400 These pools provide the ability to have duplicte data stored both
1401 onsite and offsite. The copy process can be scheduled to be handled
1402 by their storage manager during periods when there is no backup
1403 activity. Again, the migration process might be a place to consider
1404 implementing something like this.
1407 > It strikes me that it would be very nice to be able to do things
1409 > have the Job(s) backing up the machines run, and once they have all
1410 > completed, start a migration job to copy the data from disks Volumes
1412 > a tape library and then to offsite storage. Maybe this can already
1414 > done with some careful scheduling and Job prioritzation; the events
1415 > mechanism described below would probably make it very easy.
1417 This is the goal. In the first step (before events), you simply
1419 the Migration to tape later.
1421 2. Base jobs: In ADSM, each copy of each stored file is tracked in
1422 the database. Once a file (unique by path and metadata such as dates,
1423 size, ownership, etc.) is in a copy pool, no more copies are made. In
1424 other words, when you start ADSM, it begins like your concept of a
1425 base job. After that it is in the "incremental" mode. You can
1426 configure the number of "generations" of files to be retained, plus a
1427 retention date after which even old generations are purged. The
1428 database tracks the contents of media and projects the percentage of
1429 each volume that is valid. When the valid content of a volume drops
1430 below a configured percentage, the valid data are migrated to another
1431 volume and the old volume is marked as empty. Note, this requires
1432 ADSM to have an idea of the contents of a client, i.e. marking the
1433 database when an existing file was deleted, but this would solve your
1434 issue of restoring a client without restoring deleted files.
1436 This is pretty far from what bacula now does, but if you are going to
1437 rip things up for Base jobs,.....
1438 Also, the benefits of this are huge for very large shops, especially
1439 with media robots, but are a pain for shops with manual media
1443 > Base jobs sound pretty useful, but I'm not dying for them.
1445 Nobody is dying for them, but when you see what it does, you will die
1448 3. Restoring deleted files: Since I think my comments in (2) above
1449 have low probability of implementation, I'll also suggest that you
1450 could approach the issue of deleted files by a mechanism of having the
1451 fd report to the dir, a list of all files on the client for every
1452 backup job. The dir could note in the database entry for each file
1453 the date that the file was seen. Then if a restore as of date X takes
1454 place, only files that exist from before X until after X would be
1455 restored. Probably the major cost here is the extra date container in
1456 each row of the files table.
1458 Thanks for "listening". I hope some of this helps. If you want to
1459 contact me, please send me an email - I read some but not all of the
1460 mailing list traffic and might miss a reply there.
1462 Please accept my compliments for bacula. It is doing a great job for
1463 me!! I sympathize with you in the need to wrestle with excelence in
1464 execution vs. excelence in feature inclusion.
1469 ==============================
1472 - Design at hierarchial storage for Bacula. Migration and Clone.
1473 - Implement FSM (File System Modules).
1474 - Audit M_ error codes to ensure they are correct and consistent.
1475 - Add variable break characters to lex analyzer.
1476 Either a bit mask or a string of chars so that
1477 the caller can change the break characters.
1478 - Make a single T_BREAK to replace T_COMMA, etc.
1479 - Ensure that File daemon and Storage daemon can
1480 continue a save if the Director goes down (this
1481 is NOT currently the case). Must detect socket error,
1482 buffer messages for later.
1483 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
1484 - Add ability to backup to two Storage devices (two SD sessions) at
1485 the same time -- e.g. onsite, offsite.
1486 - Add the ability to consolidate old backup sets (basically do a restore
1487 to tape and appropriately update the catalog). Compress Volume sets.
1488 Might need to spool via file is only one drive is available.
1489 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
1490 someway to do so with every volume that has less than x% valid
1494 Migration: Move a backup from one Volume to another
1495 Clone: Copy a backup -- two Volumes
1498 ======================================================
1500 It is somewhat like a Full save becomes an incremental since
1501 the Base job (or jobs) plus other non-base files.
1503 - A Base backup is same as Full backup, just different type.
1504 - New BaseFiles table that contains:
1506 BaseJobId - Base JobId referenced for this FileId (needed ???)
1507 JobId - JobId currently running
1508 FileId - File not backed up, exists in Base Job
1509 FileIndex - FileIndex from Base Job.
1510 i.e. for each base file that exists but is not saved because
1511 it has not changed, the File daemon sends the JobId, BaseId,
1512 FileId, FileIndex back to the Director who creates the DB entry.
1513 - To initiate a Base save, the Director sends the FD
1514 the FileId, and full filename for each file in the Base.
1515 - When the FD finds a Base file, he requests the Director to
1516 send him the full File entry (stat packet plus MD5), or
1517 conversely, the FD sends it to the Director and the Director
1518 says yes or no. This can be quite rapid if the FileId is kept
1519 by the FD for each Base Filename.
1520 - It is probably better to have the comparison done by the FD
1521 despite the fact that the File entry must be sent across the
1523 - An alternative would be to send the FD the whole File entry
1524 from the start. The disadvantage is that it requires a lot of
1525 space. The advantage is that it requires less communications
1527 - The Job record must be updated to indicate that one or more
1529 - At end of Job, FD returns:
1530 1. Count of base files/bytes not written to tape (i.e. matches)
1531 2. Count of base file that were saved i.e. had changed.
1532 - No tape record would be written for a Base file that matches, in the
1533 same way that no tape record is written for Incremental jobs where
1534 the file is not saved because it is unchanged.
1535 - On a restore, all the Base file records must explicitly be
1536 found from the BaseFile tape. I.e. for each Full save that is marked
1537 to have one or more Base Jobs, search the BaseFile for all occurrences
1539 - An optimization might be to make the BaseFile have:
1545 This would avoid the need to explicitly fetch each File record for
1546 the Base job. The Base Job record will be fetched to get the
1547 VolSessionId and VolSessionTime.
1548 =========================================================
1551 ==========================================================
1553 For each Incremental job that is run, there may be files that
1554 were found but not saved because they were locked (this applies
1555 only to Windows). Such a system could send back to the Director
1556 a list of Unsaved files.
1558 - New UnSavedFiles table that contains:
1562 - Then in the next Incremental job, the list of Unsaved Files will be
1563 feed to the FD, who will ensure that they are explicitly chosen even
1564 if standard date/time check would not have selected them.
1565 =============================================================
1569 Multiple drive autochanger data: see Alan Brown
1570 mtx -f xxx unloadStorage Element 1 is Already Full(drive 0 was empty)
1571 Unloading Data Transfer Element into Storage Element 1...source Element
1572 Address 480 is Empty
1574 (drive 0 was empty and so was slot 1)
1575 > mtx -f xxx load 15 0
1576 no response, just returns to the command prompt when complete.
1577 > mtx -f xxx status Storage Changer /dev/changer:2 Drives, 60 Slots ( 2 Import/Export )
1578 Data Transfer Element 0:Full (Storage Element 15 Loaded):VolumeTag = HX001
1579 Data Transfer Element 1:Empty
1580 Storage Element 1:Empty
1581 Storage Element 2:Full :VolumeTag=HX002
1582 Storage Element 3:Full :VolumeTag=HX003
1583 Storage Element 4:Full :VolumeTag=HX004
1584 Storage Element 5:Full :VolumeTag=HX005
1585 Storage Element 6:Full :VolumeTag=HX006
1586 Storage Element 7:Full :VolumeTag=HX007
1587 Storage Element 8:Full :VolumeTag=HX008
1588 Storage Element 9:Full :VolumeTag=HX009
1589 Storage Element 10:Full :VolumeTag=HX010
1590 Storage Element 11:Empty
1591 Storage Element 12:Empty
1592 Storage Element 13:Empty
1593 Storage Element 14:Empty
1594 Storage Element 15:Empty
1595 Storage Element 16:Empty....
1596 Storage Element 28:Empty
1597 Storage Element 29:Full :VolumeTag=CLNU01L1
1598 Storage Element 30:Empty....
1599 Storage Element 57:Empty
1600 Storage Element 58:Full :VolumeTag=NEX261L2
1601 Storage Element 59 IMPORT/EXPORT:Empty
1602 Storage Element 60 IMPORT/EXPORT:Empty
1604 Unloading Data Transfer Element into Storage Element 15...done
1606 (just to verify it remembers where it came from, however it can be
1607 overrriden with mtx unload {slotnumber} to go to any storage slot.)
1609 There needs to be a table of drive # to devices somewhere - If there are
1610 multiple changers or drives there may not be a 1:1 correspondance between
1611 changer drive number and system device name - and depending on the way the
1612 drives are hooked up to scsi busses, they may not be linearly numbered
1613 from an offset point either.something like
1615 Autochanger drives = 2
1616 Autochanger drive 0 = /dev/nst1
1617 Autochanger drive 1 = /dev/nst2
1618 IMHO, it would be _safest_ to use explicit mtx unload commands at all
1619 times, not just for multidrive changers. For a 1 drive changer, that's
1625 MTX's manpage (1.2.15):
1626 unload [<slotnum>] [ <drivenum> ]
1627 Unloads media from drive <drivenum> into slot
1628 <slotnum>. If <drivenum> is omitted, defaults to
1629 drive 0 (as do all commands). If <slotnum> is
1630 omitted, defaults to the slot that the drive was
1631 loaded from. Note that there's currently no way
1632 to say 'unload drive 1's media to the slot it
1633 came from', other than to explicitly use that
1634 slot number as the destination.AB
1640 undef# camcontrol devlist
1641 <WANGTEK 51000 SCSI M74H 12B3> at scbus0 target 2 lun 0 (pass0,sa0)
1642 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 0 (pass1,sa1)
1643 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 1 (pass2)
1645 tapeinfo -f /dev/sg0 with a bad tape in drive 1:
1646 [kern@rufus mtx-1.2.17kes]$ ./tapeinfo -f /dev/sg0
1647 Product Type: Tape Drive
1649 Product ID: 'C5713A '
1651 Attached Changer: No
1652 TapeAlert[3]: Hard Error: Uncorrectable read/write error.
1653 TapeAlert[20]: Clean Now: The tape drive neads cleaning NOW.
1660 Medium Type: Not Loaded
1663 DataCompEnabled: yes
1664 DataCompCapable: yes
1665 DataDeCompEnabled: yes
1672 Handling removable disks
1674 From: Karl Cunningham <karlc@keckec.com>
1676 My backups are only to hard disk these days, in removable bays. This is my
1677 idea of how a backup to hard disk would work more smoothly. Some of these
1678 things Bacula does already, but I mention them for completeness. If others
1679 have better ways to do this, I'd like to hear about it.
1681 1. Accommodate several disks, rotated similar to how tapes are. Identified
1682 by partition volume ID or perhaps by the name of a subdirectory.
1683 2. Abort & notify the admin if the wrong disk is in the bay.
1684 3. Write backups to different subdirectories for each machine to be backed
1686 4. Volumes (files) get created as needed in the proper subdirectory, one
1688 5. When a disk is recycled, remove or zero all old backup files. This is
1689 important as the disk being recycled may be close to full. This may be
1690 better done manually since the backup files for many machines may be
1691 scattered in many subdirectories.
1696 - Why the heck doesn't bacula drop root priviledges before connecting to
1698 - Look at using posix_fadvise(2) for backups -- see bug #751.
1699 Possibly add the code at findlib/bfile.c:795
1700 /* TCP socket options */
1701 #define TCP_KEEPIDLE 4 /* Start keeplives after this period */
1702 - Fix bnet_connect() code to set a timer and to use time to
1704 - Implement 4th argument to make_catalog_backup that passes hostname.
1705 - Test FIFO backup/restore -- make regression
1706 - Please mount volume "xxx" on Storage device ... should also list
1707 Pool and MediaType in case user needs to create a new volume.
1708 - On restore add Restore Client, Original Client.
1709 01-Apr 00:42 rufus-dir: Start Backup JobId 55, Job=kernsave.2007-04-01_00.42.48
1710 01-Apr 00:42 rufus-sd: Python SD JobStart: JobId=55 Client=Rufus
1711 01-Apr 00:42 rufus-dir: Created new Volume "Full0001" in catalog.
1712 01-Apr 00:42 rufus-dir: Using Device "File"
1713 01-Apr 00:42 rufus-sd: kernsave.2007-04-01_00.42.48 Warning: Device "File" (/tmp) not configured to autolabel Volumes.
1714 01-Apr 00:42 rufus-sd: kernsave.2007-04-01_00.42.48 Warning: Device "File" (/tmp) not configured to autolabel Volumes.
1715 01-Apr 00:42 rufus-sd: Please mount Volume "Full0001" on Storage Device "File" (/tmp) for Job kernsave.2007-04-01_00.42.48
1716 01-Apr 00:44 rufus-sd: Wrote label to prelabeled Volume "Full0001" on device "File" (/tmp)
1717 - Check if gnome-console works with TLS.
1718 - the director seg faulted when I omitted the pool directive from a
1719 job resource. I was experimenting and thought it redundant that I had
1720 specified Pool, Full Backup Pool. and Differential Backup Pool. but
1721 apparently not. This happened when I removed the pool directive and
1722 started the director.
1723 - Add Where: client:/.... to restore job report.