9 - Document cleaning up the spool files:
10 db, pid, state, bsr, mail, conmsg, spool
11 - Document the multiple-drive-changer.txt script.
12 - Pruning with Admin job.
13 - Does WildFile match against full name? Doc.
14 - %d and %v only valid on Director, not for ClientRunBefore/After.
15 - During tests with the 260 char fix code, I found one problem:
16 if the system "sees" a long path once, it seems to forget it's
17 working drive (e.g. c:\), which will lead to a problem during
18 the next job (create bootstrap file will fail). Here is the
19 workaround: specify absolute working and pid directory in
20 bacula-fd.conf (e.g. c:\bacula\working instead of
22 - Document techniques for restoring large numbers of files.
23 - Document setting my.cnf to big file usage.
24 - Add example of proper index output to doc. show index from File;
25 - Correct the Include syntax in the m4.xxx files in examples/conf
26 - Document JobStatus and Termination codes.
27 - Fix the error with the "DVI file can't be opened" while
28 building the French PDF.
29 - Document more DVD stuff
37 - Document all the little details of setting up certificates for
38 the Bacula data encryption code.
39 - Document more precisely how to use master keys -- especially
40 for disaster recovery.
44 - Ensure that the SD re-reads the Media record if the JobFiles
45 does not match -- it may have been updated by another job.
46 - Look at moving the Storage directive from the Job to the
47 Pool in the default conf files.
48 - Make sure the new level=Full syntax is used in all
49 example conf files (especially in the manual).
50 - Fix prog copyright (SD) all other files.
51 - Migration Volume span bug
54 - Test FIFO backup/restore -- make regression
56 - Add encryption regression tests
57 - Test Volume compatibility between machine architectures
58 - Encryption documentation
59 - Wrong jobbytes with query 12 (todo)
60 - bacula-1.38.2-ssl.patch
61 - Bare-metal recovery Windows (todo)
62 - Document need for UTF-8 format
67 - Implement Python event for backing up/restoring a file.
68 - Change dbcheck to tell users to use native tools for fixing
69 broken databases, and to ensure they have the proper indexes.
70 - add udev rules for Bacula devices.
71 - If a job terminates, the DIR connection can close before the
72 Volume info is updated, leaving the File count wrong.
73 - Look at why SIGPIPE during connection can cause seg fault in
74 writing the daemon message, when Dir dropped to bacula:bacula
75 - Look at zlib 32 => 64 problems.
76 - Try turning on disk seek code.
77 - Possibly turn on St. Bernard code.
78 - Fix bextract to restore ACLs, or better yet, use common routines.
79 - Do we migrate appendable Volumes?
80 - Remove queue.c code.
81 - Some users claim that they must do two prune commands to get a
82 Volume marked as purged.
83 - Print warning message if LANG environment variable does not specify
85 - New dot commands from Arno.
86 .show device=xxx lists information from one storage device, including
87 devices (I'm not even sure that information exists in the DIR...)
88 .move eject device=xxx mostly the same as 'unmount xxx' but perhaps with
89 better machine-readable output like "Ok" or "Error busy"
90 .move eject device=xxx toslot=yyy the same as above, but with a new
91 target slot. The catalog should be updated accordingly.
92 .move transfer device=xxx fromslot=yyy toslot=zzz
95 - After pruning, check to see if the Volume retention period has
97 - Check to see if jcr->stime is lost during rescheduling of
99 - Fix re-read of last block to check if job has actually written
100 a block, and check if block was written by a different job
101 (i.e. multiple simultaneous jobs writing).
102 - Figure out how to configure query.sql. Suggestion to use m4:
103 == changequote.m4 ===
104 changequote(`[',`]')dnl
105 ==== query.sql.in ===
106 :List next 20 volumes to expire
108 Pool.Name AS PoolName,
113 [ FROM_UNIXTIME(UNIX_TIMESTAMP(Media.LastWritten) Media.VolRetention) AS Expire, ])dnl
115 [ media.lastwritten + interval '1 second' * media.volretention as expire, ])dnl
119 ON Media.PoolId=Pool.PoolId
120 WHERE Media.LastWritten>0
124 Command: m4 -DmySQL changequote.m4 query.sql.in >query.sql
126 The problem is that it requires m4, which is not present on all machines
128 - Get Perl replacement for bregex.c
129 - Given all the problems with FIFOs, I think the solution is to do something a
130 little different, though I will look at the code and see if there is not some
131 simple solution (i.e. some bug that was introduced). What might be a better
132 solution would be to use a FIFO as a sort of "key" to tell Bacula to read and
133 write data to a program rather than the FIFO. For example, suppose you
138 Then, I could imagine if you backup and restore this file with a direct
139 reference as is currently done for fifos, instead, during backup Bacula will
142 /home/kern/my-fifo.backup
144 and read the data that my-fifo.backup writes to stdout. For restore, Bacula
147 /home/kern/my-fifo.restore
149 and send the data backed up to stdout. These programs can either be an
150 executable or a shell script and they need only read/write to stdin/stdout.
152 I think this would give a lot of flexibility to the user without making any
153 significant changes to Bacula.
158 select FilenameId from Filename where Name='';
159 # Get list of all directories referenced in a Backup.
160 select Path.Path from Path,File where File.JobId=nnn and
161 File.FilenameId=(FilenameId-from-above) and File.PathId=Path.PathId
162 order by Path.Path ASC;
164 - Look into using Dart for testing
165 http://public.kitware.com/Dart/HTML/Index.shtml
167 - Look into replacing autotools with cmake
168 http://www.cmake.org/HTML/Index.html
170 === Migration from David ===
171 What I'd like to see:
174 Name = "<poolname>-migrate"
178 Migration Selection Type = LowestUtil | OldestVol | PoolOccupancy |
179 Client | PoolResidence | Volume | JobName | SQLquery
180 Migration Selection Pattern = "regexp"
181 Next Pool = <override>
184 There should be no need for a Level (migration is always Full, since you
185 don't calculate differential/incremental differences for migration),
186 Storage should be determined by the volume types in the pool, and Client
187 is really a selection issue. Migration should always occur to the
188 NextPool defined in the pool definition. If no nextpool is defined, the
189 job should end with a reason of "no place to go". If Next Pool statement
190 is present, we override the check in the pool definition and use the
193 Here's how I'd define Migration Selection Types:
196 Client -- Migrate data from selected client only. Migration Selection
197 Pattern regexp provides pattern to select client names, eg ^FS00* makes
198 all client names starting with FS00 eligible for migration.
200 Jobname -- Migration all jobs matching name. Migration Selection Pattern
201 regexp provides pattern to select jobnames existing in pool.
203 Volume -- Migrate all data on specified volumes. Migration Selection
204 Pattern regexp provides selection criteria for volumes to be migrated.
205 Volumes must exist in pool to be eligible for migration.
209 LowestUtil -- Identify the volume in the pool with the least data on it
210 and empty it. No Migration Selection Pattern required.
212 OldestVol -- Identify the LRU volume with data written, and empty it. No
213 Migration Selection Pattern required.
215 PoolOccupancy -- if pool occupancy exceeds <highmig>, migrate volumes
216 (starting with most full volumes) until pool occupancy drops below
217 <lowmig>. Pool highmig and lowmig values are in pool definition, no
218 Migration Selection Pattern required.
222 SQLQuery -- Migrate all jobuids returned by the supplied SQL query.
223 Migration Selection Pattern contains SQL query to execute; should return
224 a list of 1 or more jobuids to migrate.
226 PoolResidence -- Migrate data sitting in pool for longer than
227 PoolResidence value in pool definition. Migration Selection Pattern
228 optional; if specified, override value in pool definition (value in
232 [ possibly a Python event -- kes ]
234 - Mount on an Autochanger with no tape in the drive causes:
235 Automatically selected Storage: LTO-changer
236 Enter autochanger drive[0]: 0
237 3301 Issuing autochanger "loaded drive 0" command.
238 3302 Autochanger "loaded drive 0", result: nothing loaded.
239 3301 Issuing autochanger "loaded drive 0" command.
240 3302 Autochanger "loaded drive 0", result: nothing loaded.
241 3902 Cannot mount Volume on Storage Device "LTO-Drive1" (/dev/nst0) because:
242 Couldn't rewind device "LTO-Drive1" (/dev/nst0): ERR=dev.c:678 Rewind error on "LTO-Drive1" (/dev/nst0). ERR=No medium found.
243 3905 Device "LTO-Drive1" (/dev/nst0) open but no Bacula volume is mounted.
244 If this is not a blank tape, try unmounting and remounting the Volume.
245 - If Drive 0 is blocked, and drive 1 is set "Autoselect=no", drive 1 will
247 - Autochanger did not change volumes.
248 select * from Storage;
249 +-----------+-------------+-------------+
250 | StorageId | Name | AutoChanger |
251 +-----------+-------------+-------------+
252 | 1 | LTO-changer | 0 |
253 +-----------+-------------+-------------+
254 05-May 03:50 roxie-sd: 3302 Autochanger "loaded drive 0", result is Slot 11.
255 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Warning: Director wanted Volume "LT
256 Current Volume "LT0-002" not acceptable because:
257 1997 Volume "LT0-002" not in catalog.
258 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Error: Autochanger Volume "LT0-002"
259 Setting InChanger to zero in catalog.
260 05-May 03:50 roxie-dir: Tibs.2006-05-05_03.05.02 Error: Unable to get Media record
262 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: Error getting Volume i
263 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: Job 530 canceled.
264 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: spool.c:249 Fatal appe
265 05-May 03:49 Tibs: Tibs.2006-05-05_03.05.02 Fatal error: c:\cygwin\home\kern\bacula
274 FirstWritten: 2006-05-05 03:11:54
275 LastWritten: 2006-05-05 03:50:23
276 LabelDate: 2005-12-26 16:52:40
287 VolRetention: 31,536,000
299 Note VolStatus is blank!!!!!
306 FirstWritten: 0000-00-00 00:00:00
307 LastWritten: 0000-00-00 00:00:00
308 LabelDate: 2005-12-26 16:52:40
319 VolRetention: 31,536,000
332 Automatically selected Storage: LTO-changer
333 Enter autochanger drive[0]: 0
334 3301 Issuing autochanger "loaded drive 0" command.
335 3302 Autochanger "loaded drive 0", result: nothing loaded.
336 3301 Issuing autochanger "loaded drive 0" command.
337 3302 Autochanger "loaded drive 0", result: nothing loaded.
338 3902 Cannot mount Volume on Storage Device "LTO-Drive1" (/dev/nst0) because:
339 Couldn't rewind device "LTO-Drive1" (/dev/nst0): ERR=dev.c:678 Rewind error on "LTO-Drive1" (/dev/nst0). ERR=No medium found.
341 3905 Device "LTO-Drive1" (/dev/nst0) open but no Bacula volume is mounted.
342 If this is not a blank tape, try unmounting and remounting the Volume.
344 - Add VolumeState (enable, disable, archive)
345 - Add VolumeLock to prevent all but lock holder (SD) from updating
346 the Volume data (with the exception of VolumeState).
347 - The btape fill command does not seem to use the Autochanger
348 - Make Windows installer default to system disk drive.
349 - Look at using ioctl(FIOBMAP, ...) on Linux, and
350 DeviceIoControl(..., FSCTL_QUERY_ALLOCATED_RANGES, ...) on
351 Win32 for sparse files.
352 http://www.flexhex.com/docs/articles/sparse-files.phtml
353 http://www.informatik.uni-frankfurt.de/~loizides/reiserfs/fibmap.html
354 - Directive: at <event> "command"
355 - Command: pycmd "command" generates "command" event. How to
356 attach to a specific job?
357 - Integrate Christopher's St. Bernard code.
358 - run_cmd() returns int should return JobId_t
359 - get_next_jobid_from_list() returns int should return JobId_t
360 - Document export LDFLAGS=-L/usr/lib64
361 - Don't attempt to restore from "Disabled" Volumes.
362 - Network error on Win32 should set Win32 error code.
363 - What happens when you rename a Disk Volume?
364 - Job retention period in a Pool (and hence Volume). The job would
366 - Detect resource deadlock in Migrate when same job wants to read
367 and write the same device.
368 - Queue warning/error messages during restore so that they
369 are reported at the end of the report rather than being
370 hidden in the file listing ...
371 - Look at -D_FORTIFY_SOURCE=2
372 - Add Win32 FileSet definition somewhere
373 - Look at fixing restore status stats in SD.
374 - Make selection of Database used in restore correspond to
376 - Look at using ioctl(FIMAP) and FIGETBSZ for sparse files.
377 http://www.informatik.uni-frankfurt.de/~loizides/reiserfs/fibmap.html
378 - Implement a mode that says when a hard read error is
379 encountered, read many times (as it currently does), and if the
380 block cannot be read, skip to the next block, and try again. If
381 that fails, skip to the next file and try again, ...
383 create table LevelType (LevelType binary(1), LevelTypeLong tinyblob);
384 insert into LevelType (LevelType,LevelTypeLong) values
388 - Add ACL to restore only to original location.
389 - Show files/second in client status output.
390 - Add a recursive mark command (rmark) to restore.
391 - "Minimum Job Interval = nnn" sets minimum interval between Jobs
392 of the same level and does not permit multiple simultaneous
393 running of that Job (i.e. lets any previous invocation finish
394 before doing Interval testing).
395 - Look at simplifying File exclusions.
396 - New directive "Delete purged Volumes"
397 - new pool XXX with ScratchPoolId = MyScratchPool's PoolId and
398 let it fill itself, and RecyclePoolId = XXX's PoolId so I can
399 see if it become stable and I just have to supervise
401 - If I want to remove this pool, I set RecyclePoolId = MyScratchPool's
402 PoolId, and when it is empty remove it.
403 - Figure out how to recycle Scratch volumes back to the Scratch Pool.
405 - Allow Check Labels to be used with Bacula labels.
406 - "Resuming" a failed backup (lost line for example) by using the
407 failed backup as a sort of "base" job.
409 - Email to the user when the tape is about to need changing x
410 days before it needs changing.
411 - Command to show next tape that will be used for a job even
412 if the job is not scheduled.
413 - From: Arunav Mandal <amandal@trolltech.com>
414 1. When jobs are running and bacula for some reason crashes or if I do a
415 restart it remembers and jobs it was running before it crashed or restarted
416 as of now I loose all jobs if I restart it.
418 2. When spooling and in the midway if client is disconnected for instance a
419 laptop bacula completely discard the spool. It will be nice if it can write
420 that spool to tape so there will be some backups for that client if not all.
422 3. We have around 150 clients machines it will be nice to have a option to
423 upgrade all the client machines bacula version automatically.
425 4. Atleast one connection should be reserved for the bconsole so at heavy load
426 I should connect to the director via bconsole which at sometimes I can't
428 5. Another most important feature that is missing, say at 10am I manually
429 started backup of client abc and it was a full backup since client abc has
430 no backup history and at 10.30am bacula again automatically started backup of
431 client abc as that was in the schedule. So now we have 2 multiple Full
432 backups of the same client and if we again try to start a full backup of
433 client backup abc bacula won't complain. That should be fixed.
435 - Fix bpipe.c so that it does not modify results pointer.
436 ***FIXME*** calling sequence should be changed.
437 - For Windows disaster recovery see http://unattended.sf.net/
438 - regardless of the retention period, Bacula will not prune the
439 last Full, Diff, or Inc File data until a month after the
440 retention period for the last Full backup that was done.
441 - update volume=xxx --- add status=Full
442 - Remove old spool files on startup.
443 - Exclude SD spool/working directory.
444 - Refuse to prune last valid Full backup. Same goes for Catalog.
446 - Make a callback when Rerun failed levels is called.
447 - Give Python program access to Scheduled jobs.
448 - Add setting Volume State via Python.
449 - Python script to save with Python, not save, save with Bacula.
450 - Python script to do backup.
452 - Change the Priority, Client, Storage, JobStatus (error)
453 at the start of a job.
454 - Why is SpoolDirectory = /home/bacula/spool; not reported
455 as an error when writing a DVD?
456 - Make bootstrap file handle multiple MediaTypes (SD)
457 - Remove all old Device resource code in Dir and code to pass it
458 back in SD -- better, rework it to pass back device statistics.
459 - Check locking of resources -- be sure to lock devices where previously
460 resources were locked.
461 - The last part is left in the spool dir.
464 - In restore don't compare byte count on a raw device -- directory
465 entry does not contain bytes.
469 MA = (last_MA * 3 + rate) / 4
470 rate = (bytes - last_bytes) / (runtime - last_runtime)
471 - Max Vols limit in Pool off by one?
472 - Implement Files/Bytes,... stats for restore job.
473 - Implement Total Bytes Written, ... for restore job.
474 - Despool attributes simultaneously with data in a separate
475 thread, rejoined at end of data spooling.
476 - 7. Implement new Console commands to allow offlining/reserving drives,
477 and possibly manipulating the autochanger (much asked for).
478 - Add start/end date editing in messages (%t %T, %e?) ...
479 - Add ClientDefs similar to JobDefs.
480 - Print more info when bextract -p accepts a bad block.
481 - Fix FD JobType to be set before RunBeforeJob in FD.
482 - Look at adding full Volume and Pool information to a Volume
483 label so that bscan can get *all* the info.
484 - If the user puts "Purge Oldest Volume = yes" or "Recycle Oldest Volume = yes"
485 and there is only one volume in the pool, refuse to do it -- otherwise
486 he fills the Volume, then immediately starts reusing it.
487 - Implement copies and stripes.
488 - Add history file to console.
489 - Each file on tape creates a JobMedia record. Peter has 4 million
490 files spread over 10000 tape files and four tapes. A restore takes
491 16 hours to build the restore list.
492 - Add and option to check if the file size changed during backup.
493 - Make sure SD deletes spool files on error exit.
494 - Delete old spool files when SD starts.
495 - When labeling tapes, if you enter 000026, Bacula uses
496 the tape index rather than the Volume name 000026.
497 - Add offline tape command to Bacula console.
499 Enter MediaId or Volume name: 32
500 Enter new Volume name: DLT-20Dec04
501 Automatically selected Pool: Default
502 Connecting to Storage daemon DLTDrive at 192.168.68.104:9103 ...
503 Sending relabel command from "DLT-28Jun03" to "DLT-20Dec04" ...
504 block.c:552 Write error at 0:0 on device /dev/nst0. ERR=Bad file descriptor.
505 Error writing final EOF to tape. This tape may not be readable.
506 dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
507 askdir.c:219 NULL Volume name. This shouldn't happen!!!
508 3912 Failed to label Volume: ERR=dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
509 Label command failed for Volume DLT-20Dec04.
510 Do not forget to mount the drive!!!
511 - Bug: if a job is manually scheduled to run later, it does not appear
512 in any status report and cannot be cancelled.
514 ==== Keeping track of deleted/new files ====
515 - To mark files as deleted, run essentially a Verify to disk, and
516 when a file is found missing (MarkId != JobId), then create
517 a new File record with FileIndex == -1. This could be done
518 by the FD at the same time as the backup.
520 My "trick" for keeping track of deletions is the following.
521 Assuming the user turns on this option, after all the files
522 have been backed up, but before the job has terminated, the
523 FD will make a pass through all the files and send their
524 names to the DIR (*exactly* the same as what a Verify job
525 currently does). This will probably be done at the same
526 time the files are being sent to the SD avoiding a second
527 pass. The DIR will then compare that to what is stored in
528 the catalog. Any files in the catalog but not in what the
529 FD sent will receive a catalog File entry that indicates
530 that at that point in time the file was deleted. This
531 either transmitted to the FD or simultaneously computed in
532 the FD, so that the FD can put a record on the tape that
533 indicates that the file has been deleted at this point.
534 A delete file entry could potentially be one with a FileIndex
535 of 0 or perhaps -1 (need to check if FileIndex is used for
536 some other thing as many of the Bacula fields are "overloaded"
539 During a restore, any file initially picked up by some
540 backup (Full, ...) then subsequently having a File entry
541 marked "delete" will be removed from the tree, so will not
542 be restored. If a file with the same name is later OK it
543 will be inserted in the tree -- this already happens. All
544 will be consistent except for possible changes during the
547 Since I'm on the subject, some of you may be wondering what
548 the utility of the in memory tree is if you are going to
549 restore everything (at least it comes up from time to time
550 on the list). Well, it is still *very* useful because it
551 allows only the last item found for a particular filename
552 (full path) to be entered into the tree, and thus if a file
553 is backed up 10 times, only the last copy will be restored.
554 I recently (last Friday) restored a complete directory, and
555 the Full and all the Differential and Incremental backups
556 spanned 3 Volumes. The first Volume was not even mounted
557 because all the files had been updated and hence backed up
558 since the Full backup was made. In this case, the tree
559 saved me a *lot* of time.
561 Make sure this information is stored on the tape too so
562 that it can be restored directly from the tape.
564 All the code (with the exception of formally generating and
565 saving the delete file entries) already exists in the Verify
566 Catalog command. It explicitly recognizes added/deleted files since
567 the last InitCatalog. It is more or less a "simple" matter of
568 taking that code and adapting it slightly to work for backups.
570 Comments from Martin Simmons (I think they are all covered):
571 Ok, that should cover the basics. There are few issues though:
573 - Restore will depend on the catalog. I think it is better to include the
574 extra data in the backup as well, so it can be seen by bscan and bextract.
576 - I'm not sure if it will preserve multiple hard links to the same inode. Or
577 maybe adding or removing links will cause the data to be dumped again?
579 - I'm not sure if it will handle renamed directories. Possibly it will work
580 by dumping the whole tree under a renamed directory?
582 - It remains to be seen how the backup performance of the DIR's will be
583 affected when comparing the catalog for a large filesystem.
587 How about introducing a Type = MgmtPolicy job type? That job type would
588 be responsible for scanning the Bacula environment looking for specific
589 conditions, and submitting the appropriate jobs for implementing said
593 Name = "Migration-Policy"
595 Policy Selection Job Type = Migrate
596 Scope = "<keyword> <operator> <regexp>"
597 Threshold = "<keyword> <operator> <regexp>"
598 Job Template = <template-name>
601 Where <keyword> is any legal job keyword, <operator> is a comparison
602 operator (=,<,>,!=, logical operators AND/OR/NOT) and <regexp> is a
603 appropriate regexp. I could see an argument for Scope and Threshold
604 being SQL queries if we want to support full flexibility. The
605 Migration-Policy job would then get scheduled as frequently as a site
606 felt necessary (suggested default: every 15 minutes).
611 Name = "Migration-Policy"
613 Policy Selection Job Type = Migration
615 Threshold = "Migration Selection Type = LowestUtil"
616 Job Template = "MigrationTemplate"
619 would select all pools for examination and generate a job based on
620 MigrationTemplate to automatically select the volume with the lowest
621 usage and migrate it's contents to the nextpool defined for that pool.
623 This policy abstraction would be really handy for adjusting the behavior
624 of Bacula according to site-selectable criteria (one thing that pops
625 into mind is Amanda's ability to automatically adjust backup levels
626 depending on various criteria).
632 - Add Pool/Storage override regression test.
633 - Add delete JobId to regression.
634 - Add a regression test for dbcheck.
635 - New test to add bscan to four-concurrent-jobs regression,
636 i.e. after the four-concurrent jobs zap the
637 database as is done in the bscan-test, then use bscan to
638 restore the database, do a restore and compare with the
640 - Add restore of specific JobId to regression (item 3
641 on the restore prompt)
642 - Add IPv6 to regression
643 - Add database test to regression. Test each function like delete,
646 - AntiVir can slow down backups on Win32 systems.
647 - Win32 systems with FAT32 can be much slower than NTFS for
648 more than 1000 files per directory.
652 - A HOLD command to stop all jobs from starting.
653 - A PAUSE command to pause all running jobs ==> release the
655 - Media Type = LTO,LTO-2,LTO-3
656 Media Type Read = LTO,LTO2,LTO3
657 Media Type Write = LTO2, LTO3
659 === From Carsten Menke <bootsy52@gmx.net>
661 Following is a list of what I think in the situations where I'm faced with,
662 could be a usefull enhancement to bacula, which I'm certain other users will
663 benefit from as well.
665 1. NextJob/NextJobs Directive within a Job Resource in the form of
666 NextJobs = job1,job2.
669 I currently solved the problem with running multiple jobs each after each
670 by setting the Max Wait Time for a job to 8 hours, and give
671 the jobs different Priorities. However, there scenarios where
672 1 Job is directly depending on another job, so if the former job fails,
673 the job after it needn't to be run
674 while maybe other jobs should run despite of that
677 A Backup Job and a Verify job, if the backup job fails there is no need to run
678 the verify job, as the backup job already failed. However, one may like
679 to backup the Catalog to disk despite of that the main backup job failed.
682 I see that this is related to the Event Handlers which are on the ToDo
683 list, also it is maybe a good idea to check for the return value and
684 execute different actions based on the return value
687 3. offline capability to bconsole
690 Currently I use a script which I execute within the last Job via the
691 RunAfterJob Directive, to release and eject the tape.
692 So I have to call bconsole "release=Storage-Name" and afterwards
693 mt -f /dev/nst0 eject to get the tape out.
695 If I have multiple Storage Devices, than these may not be /dev/nst0 and
696 I have to modify the script or call it with parameters etc.
697 This would actually not be needed, as everything is already defined
698 in bacula-sd.conf and if I can invoke bconsole with the
699 storage name via $1 in the script than I'm done and information is
702 4. %s for Storage Name added to the chars being substituted in "RunAfterJob"
706 For the reason mentioned in 3. to have the ability to call a
707 script with /scripts/foobar %s and in the script use $1
708 to pass the Storage Name to bconsole
710 5. Setting Volume State within a Job Resource
713 Instead of using "Maximum Volume Jobs" in the Pool Resource,
714 I would have the possibilty to define
715 in a Job Resource that after this certain job is run, the Volume State
716 should be set to "Volume State = Used", this give more flexibility (IMHO).
718 6. Localization of Bacula Messages
721 Unfortunatley many,many people I work with don't speak english very well.
722 So if at least the Reporting messages would be localized then they
723 would understand that they have to change the tape,etc. etc.
725 I volunteer to do the german translations, and if I can convince my wife also
726 french and Morre (western african language).
728 7. OK, this is evil, probably bound to security risks and maybe not possible
729 due to the design of bacula.
731 Implementation of Backtics ( `command` ) for shell comand execution to
732 the "Label Format" Directive.
736 Currently I have defined BACULA_DAY_OF_WEEK="day1|day2..." resulting in
737 Label Format = "HolyBackup-${BACULA_DAY_OF_WEEK[${WeekDay}]}". If I could
738 use backticks than I could use "Label Format = HolyBackup-`date +%A` to have
739 the localized name for the day of the week appended to the
740 format string. Then I have the tape labeled automatically with weekday
741 name in the correct language.
743 - Yes, that is surely the case. I probably should turn those into Warning
744 errors. In addition, you just made me think that it might not be bad to
745 add an option to check the file size after backing up the file and
746 report if it changes. This would be done as an option because it would
749 Kern, good idea. If you do do that, mention in the output: file
750 shrunk, or file expanded, just to make it obvious to the user
751 (without having to the refer to file size), just how the file size
754 Would this option be for all file, or just one file? Or a fileset?
755 - Make output from status use html table tags for nicely
756 presenting in a browser.
757 - Can one write tapes faster with 8192 byte block sizes?
758 - Document security problems with the same password for everyone in
759 rpm and Win32 releases.
760 - Browse generations of files.
761 - I've seen an error when my catalog's File table fills up. I
762 then have to recreate the File table with a larger maximum row
763 size. Relevant information is at
764 http://dev.mysql.com/doc/mysql/en/Full_table.html ; I think the
765 "Installing and Configuring MySQL" chapter should talk a bit
766 about this potential problem, and recommend a solution.
767 - For Solaris must use POSIX awk.
768 - Want speed of writing to tape while despooling.
769 - Supported autochanger:
777 Wangtek 6525ES (SCSI-1 QIC drive, 525MB), under Linux 2.4.something,
778 bacula 1.36.0/1 works with blocksize 16k INSIDE bacula-sd.conf.
779 - Add regex from http://www.pcre.org to Bacula for Win32.
780 - Use only shell tools no make in CDROM package.
781 - Include within include does it work?
782 - Implement a Pool of type Cleaning?
783 - Implement VolReadTime and VolWriteTime in SD
784 - Modify Backing up Your Database to include a bootstrap file.
785 - Think about making certain database errors fatal.
786 - Look at correcting the time jump in the scheduler for daylight
787 savings time changes.
788 - Add a "real" timer to network connections.
789 - Promote to Full = Time period
790 - Check dates entered by user for correctness (month/day/... ranges)
791 - Compress restore Volume listing by date and first file.
792 - Look at patches/bacula_db.b2z postgresql that loops during restore.
794 - Perhaps add read/write programs and/or plugins to FileSets.
795 - How to handle backing up portables ...
796 - Add some sort of guaranteed Interval for upgrading jobs.
797 - Can we write the state file after every job terminates? On Win32
798 the system crashes and the state file is not updated.
801 Documentation to do: (any release a little bit at a time)
802 - Doc to do unmount before removing magazine.
803 - Alternative to static linking "ldd prog" save all binaries listed,
804 restore them and point LD_LIBRARY_PATH to them.
805 - Document add "</dev/null >/dev/null 2>&1" to the bacula-fd command line
806 - Document query file format.
807 - Add more documentation for bsr files.
808 - Document problems with Verify and pruning.
809 - Document how to use multiple databases.
810 - VXA drives have a "cleaning required"
811 indicator, but Exabyte recommends preventive cleaning after every 75
814 In this context, it should be noted that Exabyte has a command-line
815 vxatool utility available for free download. (The current version is
816 vxatool-3.72.) It can get diagnostic info, read, write and erase tapes,
817 test the drive, unload tapes, change drive settings, flash new firmware,
819 Of particular interest in this context is that vxatool <device> -i will
820 report, among other details, the time since last cleaning in tape motion
821 minutes. This information can be retrieved (and settings changed, for
822 that matter) through the generic-SCSI device even when Bacula has the
823 regular tape device locked. (Needless to say, I don't recommend
824 changing tape settings while a job is running.)
825 - Lookup HP cleaning recommendations.
826 - Lookup HP tape replacement recommendations (see trouble shooting autochanger)
827 - Document doing table repair
830 ===================================
831 - Add macro expansions in JobDefs.
832 Run Before Job = "SomeFile %{Level} %{Client}"
833 Write Bootstrap="/some/dir/%{JobName}_%{Client}.bsr"
834 - Use non-blocking network I/O but if no data is available, use
836 - Use gather write() for network I/O.
837 - Autorestart on crash.
838 - Add bandwidth limiting.
839 - Add acks every once and a while from the SD to keep
840 the line from timing out.
841 - When an error in input occurs and conio beeps, you can back
842 up through the prompt.
843 - Detect fixed tape block mode during positioning by looking at
844 block numbers in btape "test". Possibly adjust in Bacula.
845 - Fix list volumes to output volume retention in some other
846 units, perhaps via a directive.
847 - If opening a tape in read/write mode fails attempt to open
848 it in read-only mode, and mark the tape for read only.
849 - Allow Simultaneous Priorities = yes => run up to Max concurrent jobs even
850 with multiple priorities.
851 - If you use restore replace=never, the directory attributes for
852 non-existent directories will not be restored properly.
854 - see lzma401.zip in others directory for new compression
856 - Minimal autochanger handling in Bacula and in btape.
857 - Look into how tar does not save sockets and the possiblity of
858 not saving them in Bacula (Martin Simmons reported this).
859 - Fix restore jobs so that multiple jobs can run if they
860 are not using the same tape(s).
861 - Allow the user to select JobType for manual pruning/purging.
862 - bscan does not put first of two volumes back with all info in
864 - Implement the FreeBSD nodump flag in chflags.
865 - Figure out how to make named console messages go only to that
866 console and to the non-restricted console (new console class?).
867 - Make restricted console prompt for password if *ask* is set or
868 perhaps if password is undefined.
869 - Implement "from ISO-date/time every x hours/days/weeks/months" in
872 ==== from Marc Schoechlin
873 - the help-command should be more verbose
874 (it should explain the paramters of the different
876 -> its time-comsuming to consult the manual anytime
877 you need a special parameter
878 -> maybe its more easy to maintain this, if the
879 descriptions of that commands are outsourced to
881 - the cd-command should allow complete paths
882 i.e. cd /foo/bar/foo/bar
883 -> if a customer mails me the path to a certain file,
884 its faster to enter the specified directory
885 - if the password is not configured in bconsole.conf
886 you should be asked for it.
887 -> sometimes you like to do restore on a customer-machine
888 which shouldnt know the password for bacula.
889 -> adding the password to the file favours admins
890 to forget to remove the password after usage
892 the protection of that file is less important
893 - long-listed-output of commands should be scrollable
894 like the unix more/less-command does
895 -> if someone runs 200 and more machines, the lists could
896 be a little long and complex
897 - command-output should be shown column by column
898 to reduce scrolling and to increase clarity
900 - lsmark should list the selected files with full
902 - wildcards for selecting and file and directories would be nice
903 - any actions should be interuptable with STRG+C
904 - command-expansion would be pretty cool
906 - When the replace Never option is set, new directory permissions
907 are not restored. See bug 213. To fix this requires creating a
908 list of newly restored directories so that those directory
909 permissions *can* be restored.
910 - Compaction of Disk space by "migrating" Volumes that have pruned
911 Jobs (what criteria? size, #jobs, time).
912 - Add prune all command
913 - Document fact that purge can destroy a part of a restore by purging
914 one volume while others remain valid -- perhaps mark Jobs.
915 - Add multiple-media-types.txt
916 - look at mxt-changer.html
917 - Make ? do a help command (no return needed).
918 - Implement restore directory.
919 - Document streams and how to implement them.
920 - Try not to re-backup a file if a new hard link is added.
921 - Add feature to backup hard links only, but not the data.
922 - Fix stream handling to be simpler.
923 - Add Priority and Bootstrap to Run a Job.
924 - Eliminate Restore "Run Restore Job" prompt by allowing new "run command
926 - Remove View FileSet button from Run a Job dialog.
927 - Handle prompt for restore job at end of Restore command.
928 - Add display of total selected files to Restore window.
929 - Add tree pane to left of window.
930 - Add progress meter.
931 - Max wait time or max run time causes seg fault -- see runtime-bug.txt
932 - Document writing to a CD/DVD with Bacula.
933 - Add a "base" package to the window installer for pthreadsVCE.dll
934 which is needed by all packages.
935 - Add message to user to check for fixed block size when the forward
936 space test fails in btape.
937 - When unmarking a directory check if all files below are unmarked and
938 then remove the + flag -- in the restore tree.
939 - Possibly implement: Action = Unmount Device="TapeDrive1" in Admin jobs.
940 - Setup lrrd graphs: (http://www.linpro.no/projects/lrrd/) Mike Acar.
941 - Revisit the question of multiple Volumes (disk) on a single device.
942 - Add a block copy option to bcopy.
943 - Investigate adding Mac Resource Forks.
944 - Finish work on Gnome restore GUI.
945 - Fix "llist jobid=xx" where no fileset or client exists.
946 - For each job type (Admin, Restore, ...) require only the really necessary
947 fields.- Pass Director resource name as an option to the Console.
948 - Add a "batch" mode to the Console (no unsolicited queries, ...).
949 - Add a .list all files in the restore tree (probably also a list all files)
950 Do both a long and short form.
951 - Allow browsing the catalog to see all versions of a file (with
952 stat data on each file).
953 - Restore attributes of directory if replace=never set but directory
955 - Use SHA1 on authentication if possible.
956 - See comtest-xxx.zip for Windows code to talk to USB.
957 - Add John's appended files:
958 Appended = { /files/server/logs/http/*log }
959 and such files would be treated as follows.On a FULL backup, they would
960 be backed up like any other file.On an INCREMENTAL backup, where a
961 previous INCREMENTAL or FULL was already in thecatalogue and the length
962 of the file wasgreater than the length of the last backup, only thedata
963 added since the last backup will be dumped.On an INCREMENTAL backup, if
964 the length of the file is less than thelength of the file with the same
965 name last backed up, the completefile is dumped.On Windows systems, with
966 creation date of files, we can be evensmarter about this and not count
967 entirely upon the length.On a restore, the full and all incrementals
968 since it will beapplied in sequence to restore the file.
969 - Check new HAVE_WIN32 open bits.
970 - Check if the tape has moved before writing.
971 - Handling removable disks -- see below:
972 - Keep track of tape use time, and report when cleaning is necessary.
973 - Add FromClient and ToClient keywords on restore command (or
974 BackupClient RestoreClient).
975 - Implement a JobSet, which groups any number of jobs. If the
976 JobSet is started, all the jobs are started together.
977 Allow Pool, Level, and Schedule overrides.
978 - Enhance cancel to timeout BSOCK packets after a specific delay.
979 - Do scheduling by UTC using gmtime_r() in run_conf, scheduler, and
980 ua_status.!!! Thanks to Alan Brown for this tip.
981 - Look at updating Volume Jobs so that Max Volume Jobs = 1 will work
982 correctly for multiple simultaneous jobs.
983 - Correct code so that FileSet MD5 is calculated for < and | filename
985 - Implement the Media record flag that indicates that the Volume does disk
987 - Implement VolAddr, which is used when Volume is addressed like a disk,
988 and form it from VolFile and VolBlock.
989 - Make multiple restore jobs for multiple media types specifying
990 the proper storage type.
991 - Fix fast block rejection (stored/read_record.c:118). It passes a null
992 pointer (rec) to try_repositioning().
993 - Look at extracting Win data from BackupRead.
994 - Implement RestoreJobRetention? Maybe better "JobRetention" in a Job,
995 which would take precidence over the Catalog "JobRetention".
996 - Implement Label Format in Add and Label console commands.
997 - Possibly up network buffers to 65K. Put on variable.
998 - Put email tape request delays on one or more variables. User wants
999 to cancel the job after a certain time interval. Maximum Mount Wait?
1000 - Job, Client, Device, Pool, or Volume?
1001 Is it possible to make this a directive which is *optional* in multiple
1002 resources, like Level? If so, I think I'd make it an optional directive
1003 in Job, Client, and Pool, with precedence such that Job overrides Client
1004 which in turn overrides Pool.
1006 - New Storage specifications:
1007 - Want to write to multiple storage devices simultaneously
1008 - Want to write to multiple storage devices sequentially (in one job)
1009 - Want to read/write simultaneously
1010 - Key is MediaType -- it must match
1012 Passed to SD as a sort of BSR record called Storage Specification
1016 MediaType -> Next MediaType
1018 Device -> Next Device
1020 Allow multiple Storage specifications
1028 Allow Multiple Pool specifications (note, Pool currently
1030 Allow Multiple MediaType specifications in Dir conf
1031 Allow Multiple Device specifications in Dir conf
1032 Perhaps keep this in a single SSR
1033 Tie a Volume to a specific device by using a MediaType that
1034 is contained in only one device.
1035 In SD allow Device to have Multiple MediaTypes
1037 - Ideas from Jerry Scharf:
1038 First let's point out some big pluses that bacula has for this
1040 more importantly it's active. Thank you so much for that
1041 even more important, it's not flaky
1042 it has an open access catalog, opening many possibilities
1043 it's pushing toward heterogeneous systems capability
1045 Macintosh file client
1046 macs are an interesting niche, but I fear a server is a rathole
1047 working bare iron recovery for windows
1048 the option for inc/diff backups not reset on fileset revision
1049 a) use both change and inode update time against base time
1050 b) do the full catalog check (expensive but accurate)
1051 sizing guide (how much system is needed to back up N systems/files)
1052 consultants on using bacula in building a disaster recovery system
1053 an integration guide
1054 or how to get at fancy things that one could do with bacula
1055 logwatch code for bacula logs (or similar)
1056 linux distro inclusion of bacula (brings good and bad, but necessary)
1057 win2k/XP server capability (icky but you asked)
1058 support for Oracle database ??
1060 - Look at adding SQL server and Exchange support for Windows.
1061 - Make dev->file and dev->block_num signed integers so that -1 can
1062 be an invalid value which happens with BSR.
1063 - Create VolAddr for disk files in place of VolFile and VolBlock. This
1064 is needed to properly specify ranges.
1065 - Add progress of files/bytes to SD and FD.
1066 - Print warning message if FileId > 4 billion
1067 - do a "messages" before the first prompt in Console
1068 - Client does not show busy during Estimate command.
1069 - Implement Console mtx commands.
1070 - Implement a Mount Command and an Unmount Command where
1071 the users could specify a system command to be performed
1072 to do the mount, after which Bacula could attempt to
1073 read the device. This is for Removeable media such as a CDROM.
1074 - Most likely, this mount command would be invoked explicitly
1075 by the user using the current Console "mount" and "unmount"
1076 commands -- the Storage Daemon would do the right thing
1077 depending on the exact nature of the device.
1078 - As with tape drives, when Bacula wanted a new removable
1079 disk mounted, it would unmount the old one, and send a message
1080 to the user, who would then use "mount" as described above
1081 once he had actually inserted the disk.
1082 - Implement dump/print label to UA
1083 - Spool to disk only when the tape is full, then when a tape is hung move
1085 - bextract is sending everything to the log file ****FIXME****
1086 - Allow multiple Storage specifications (or multiple names on
1087 a single Storage specification) in the Job record. Thus a job
1088 can be backed up to a number of storage devices.
1089 - Implement some way for the File daemon to contact the Director
1090 to start a job or pass its DHCP obtained IP number.
1091 - Implement a query tape prompt/replace feature for a console
1092 - Copy console @ code to gnome2-console
1093 - Make AES the only encryption algorithm see
1094 http://csrc.nist.gov/CryptoToolkit/aes/). It's
1095 an officially adopted standard, has survived peer
1096 review, and provides keys up to 256 bits.
1097 - Take a careful look at SetACL http://setacl.sourceforge.net
1098 - Make tree walk routines like cd, ls, ... more user friendly
1099 by handling spaces better.
1100 - Make sure that Bacula rechecks the tape after the 20 min wait.
1101 - Set IO_NOWAIT on Bacula TCP/IP packets.
1102 - Try doing a raw partition backup and restore by mounting a
1104 - From Lars Kellers:
1105 Yes, it would allow to highly automatic the request for new tapes. If a
1106 tape is empty, bacula reads the barcodes (native or simulated), and if
1107 an unused tape is found, it runs the label command with all the
1108 necessary parameters.
1110 By the way can bacula automatically "move" an empty/purged volume say
1111 in the "short" pool to the "long" pool if this pool runs out of volume
1113 - What to do about "list files job=xxx".
1114 - Get and test MySQL 4.0
1115 - Look at how fuser works and /proc/PID/fd that is how Nic found the
1116 file descriptor leak in Bacula.
1117 - Implement WrapCounters in Counters.
1118 - Add heartbeat from FD to SD if hb interval expires.
1119 - Can we dynamically change FileSets?
1120 - If pool specified to label command and Label Format is specified,
1121 automatically generate the Volume name.
1122 - Why can't SQL do the filename sort for restore?
1123 - Add ExhautiveRestoreSearch
1124 - Look at the possibility of loading only the necessary
1125 data into the restore tree (i.e. do it one directory at a
1126 time as the user walks through the tree).
1127 - Possibly use the hash code if the user selects all for a restore command.
1128 - Fix "restore all" to bypass building the tree.
1129 - Prohibit backing up archive device (findlib/find_one.c:128)
1130 - Implement Release Device in the Job resource to unmount a drive.
1131 - Implement Acquire Device in the Job resource to mount a drive,
1132 be sure this works with admin jobs so that the user can get
1133 prompted to insert the correct tape. Possibly some way to say to
1134 run the job but don't save the files.
1135 - Make things like list where a file is saved case independent for
1138 - Use autochanger to handle multiple devices.
1139 - On Windows with very long path names, it may be impossible to create
1140 a file (and thus restore it) because the total length is too long.
1141 We must cd into the directory then create the file without the
1143 - Implement a Recycle command
1144 - Test a second language e.g. french.
1145 - Start working on Base jobs.
1146 - Implement UnsavedFiles DB record.
1147 - From Phil Stracchino:
1148 It would probably be a per-client option, and would be called
1149 something like, say, "Automatically purge obsoleted jobs". What it
1150 would do is, when you successfully complete a Differential backup of a
1151 client, it would automatically purge all Incremental backups for that
1152 client that are rendered redundant by that Differential. Likewise,
1153 when a Full backup on a client completed, it would automatically purge
1154 all Differential and Incremental jobs obsoleted by that Full backup.
1155 This would let people minimize the number of tapes they're keeping on
1156 hand without having to master the art of retention times.
1157 - When doing a Backup send all attributes back to the Director, who
1158 would then figure out what files have been deleted.
1159 - Currently in mount.c:236 the SD simply creates a Volume. It should have
1160 explicit permission to do so. It should also mark the tape in error
1161 if there is an error.
1162 - Cancel waiting for Client connect in SD if FD goes away.
1164 - Implement timeout in response() when it should come quickly.
1165 - Implement a Slot priority (loaded/not loaded).
1166 - Implement "vacation" Incremental only saves.
1167 - Implement create "FileSet"?
1168 - Add prefixlinks to where or not where absolute links to FD.
1169 - Issue message to mount a new tape before the rewind.
1170 - Simplified client job initiation for portables.
1171 - If SD cannot open a drive, make it periodically retry.
1172 - Add more of the config info to the tape label.
1174 - If tape is marked read-only, then try opening it read-only rather than
1175 failing, and remember that it cannot be written.
1176 - Refine SD waiting output:
1177 Device is being positioned
1178 > Device is being positioned for append
1179 > Device is being positioned to file x
1181 - Figure out some way to estimate output size and to avoid splitting
1182 a backup across two Volumes -- this could be useful for writing CDROMs
1183 where you really prefer not to have it split -- not serious.
1184 - Have SD compute MD5 or SHA1 and compare to what FD computes.
1185 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
1186 actual data on the Volume and compare it.
1187 - Implement Bacula plugins -- design API
1188 - Make bcopy read through bad tape records.
1189 - Program files (i.e. execute a program to read/write files).
1190 Pass read date of last backup, size of file last time.
1191 - Add Signature type to File DB record.
1192 - CD into subdirectory when open()ing files for backup to
1193 speed up things. Test with testfind().
1194 - Priority job to go to top of list.
1195 - Why are save/restore of device different sizes (sparse?) Yup! Fix it.
1196 - Implement some way for the Console to dynamically create a job.
1197 - Solaris -I on tar for include list
1198 - Need a verbose mode in restore, perhaps to bsr.
1199 - bscan without -v is too quiet -- perhaps show jobs.
1200 - Add code to reject whole blocks if not wanted on restore.
1201 - Check if we can increase Bacula FD priorty in Win2000
1202 - Make sure the MaxVolFiles is fully implemented in SD
1203 - Check if both CatalogFiles and UseCatalog are set to SD.
1204 - Possibly add email to Watchdog if drive is unmounted too
1205 long and a job is waiting on the drive.
1206 - After unmount, if restore job started, ask to mount.
1207 - Add UA rc and history files.
1208 - put termcap (used by console) in ./configure and
1209 allow -with-termcap-dir.
1210 - Fix Autoprune for Volumes to respect need for full save.
1211 - Compare tape to Client files (attributes, or attributes and data)
1212 - Make all database Ids 64 bit.
1213 - Allow console commands to detach or run in background.
1214 - Fix status delay on storage daemon during rewind.
1215 - Add SD message variables to control operator wait time
1216 - Maximum Operator Wait
1217 - Minimum Message Interval
1218 - Maximum Message Interval
1219 - Send Operator message when cannot read tape label.
1220 - Verify level=Volume (scan only), level=Data (compare of data to file).
1221 Verify level=Catalog, level=InitCatalog
1223 - Add keyword search to show command in Console.
1224 - Events : tape has more than xxx bytes.
1225 - Complete code in Bacula Resources -- this will permit
1226 reading a new config file at any time.
1227 - Handle ctl-c in Console
1228 - Implement script driven addition of File daemon to config files.
1229 - Think about how to make Bacula work better with File (non-tape) archives.
1230 - Write Unix emulator for Windows.
1231 - Put memory utilization in Status output of each daemon
1232 if full status requested or if some level of debug on.
1233 - Make database type selectable by .conf files i.e. at runtime
1234 - Set flag for uname -a. Add to Volume label.
1235 - Restore files modified after date
1236 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
1237 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
1238 - Timout a job or terminate if link goes down, or reopen link and query.
1239 - Concept of precious tapes (cannot be reused).
1240 - Make bcopy copy with a single tape drive.
1241 - Permit changing ownership during restore.
1244 > My suggestion: Add a feature on the systray menu-icon menu to request
1245 > an immediate backup now. This would be useful for laptop users who may
1246 > not be on the network when the regular scheduled backup is run.
1248 > My wife's suggestion: Add a setting to the win32 client to allow it to
1249 > shut down the machine after backup is complete (after, of course,
1250 > displaying a "System will shut down in one minute, click here to cancel"
1251 > warning dialog). This would be useful for sites that want user
1252 > woorkstations to be shut down overnight to save power.
1255 - Autolabel should be specified by DIR instead of SD.
1257 - Add media capacity
1258 - AutoScan (check checksum of tape)
1259 - Format command = "format /dev/nst0"
1263 - Seek resolution (usually corresponds to buffer size)
1264 - EODErrorCode=ENOSPC or code
1265 - Partial Read error code
1266 - Partial write error code
1267 - Nonformatted read error
1268 - Nonformatted write error
1269 - WriteProtected error
1273 - IgnoreCloseErrors=yes
1283 - FD sends unsaved file list to Director at end of job (see
1285 - File daemon should build list of files skipped, and then
1286 at end of save retry and report any errors.
1287 - Write a Storage daemon that uses pipes and
1288 standard Unix programs to write to the tape.
1290 - Need something that monitors the JCR queue and
1291 times out jobs by asking the deamons where they are.
1292 - Enhance Jmsg code to permit buffering and saving to disk.
1293 - device driver = "xxxx" for drives.
1294 - Verify from Volume
1295 - Ensure that /dev/null works
1296 - Need report class for messages. Perhaps
1297 report resource where report=group of messages
1298 - enhance scan_attrib and rename scan_jobtype, and
1299 fill in code for "since" option
1300 - Director needs a time after which the report status is sent
1301 anyway -- or better yet, a retry time for the job.
1302 - Don't reschedule a job if previous incarnation is still running.
1303 - Some way to automatically backup everything is needed????
1304 - Need a structure for pending actions:
1306 - termination status (part of buffered msgs?)
1308 Read, Write, Clean, Delete
1309 - Login to Bacula; Bacula users with different permissions:
1310 owner, group, user, quotas
1311 - Store info on each file system type (probably in the job header on tape.
1312 This could be the output of df; or perhaps some sort of /etc/mtab record.
1314 ========= ideas ===============
1315 From: "Jerry K. Schieffer" <jerry@skylinetechnology.com>
1316 To: <kern@sibbald.com>
1317 Subject: RE: [Bacula-users] future large programming jobs
1318 Date: Thu, 26 Feb 2004 11:34:54 -0600
1320 I noticed the subject thread and thought I would offer the following
1321 merely as sources of ideas, i.e. something to think about, not even as
1322 strong as a request. In my former life (before retiring) I often
1323 dealt with backups and storage management issues/products as a
1324 developer and as a consultant. I am currently migrating my personal
1325 network from amanda to bacula specifically because of the ability to
1326 cross media boundaries during storing backups.
1327 Are you familiar with the commercial product called ADSM (I think IBM
1328 now sells it under the Tivoli label)? It has a couple of interesting
1329 ideas that may apply to the following topics.
1331 1. Migration: Consider that when you need to restore a system, there
1332 may be pressure to hurry. If all the information for a single client
1333 can eventually end up on the same media (and in chronological order),
1334 the restore is facillitated by not having to search past information
1335 from other clients. ADSM has the concept of "client affinity" that
1336 may be associated with it's storage pools. It seems to me that this
1337 concept (as an optional feature) might fit in your architecture for
1340 ADSM also has the concept of defining one or more storage pools as
1341 "copy pools" (almost mirrors, but only in the sense of contents).
1342 These pools provide the ability to have duplicte data stored both
1343 onsite and offsite. The copy process can be scheduled to be handled
1344 by their storage manager during periods when there is no backup
1345 activity. Again, the migration process might be a place to consider
1346 implementing something like this.
1349 > It strikes me that it would be very nice to be able to do things
1351 > have the Job(s) backing up the machines run, and once they have all
1352 > completed, start a migration job to copy the data from disks Volumes
1354 > a tape library and then to offsite storage. Maybe this can already
1356 > done with some careful scheduling and Job prioritzation; the events
1357 > mechanism described below would probably make it very easy.
1359 This is the goal. In the first step (before events), you simply
1361 the Migration to tape later.
1363 2. Base jobs: In ADSM, each copy of each stored file is tracked in
1364 the database. Once a file (unique by path and metadata such as dates,
1365 size, ownership, etc.) is in a copy pool, no more copies are made. In
1366 other words, when you start ADSM, it begins like your concept of a
1367 base job. After that it is in the "incremental" mode. You can
1368 configure the number of "generations" of files to be retained, plus a
1369 retention date after which even old generations are purged. The
1370 database tracks the contents of media and projects the percentage of
1371 each volume that is valid. When the valid content of a volume drops
1372 below a configured percentage, the valid data are migrated to another
1373 volume and the old volume is marked as empty. Note, this requires
1374 ADSM to have an idea of the contents of a client, i.e. marking the
1375 database when an existing file was deleted, but this would solve your
1376 issue of restoring a client without restoring deleted files.
1378 This is pretty far from what bacula now does, but if you are going to
1379 rip things up for Base jobs,.....
1380 Also, the benefits of this are huge for very large shops, especially
1381 with media robots, but are a pain for shops with manual media
1385 > Base jobs sound pretty useful, but I'm not dying for them.
1387 Nobody is dying for them, but when you see what it does, you will die
1390 3. Restoring deleted files: Since I think my comments in (2) above
1391 have low probability of implementation, I'll also suggest that you
1392 could approach the issue of deleted files by a mechanism of having the
1393 fd report to the dir, a list of all files on the client for every
1394 backup job. The dir could note in the database entry for each file
1395 the date that the file was seen. Then if a restore as of date X takes
1396 place, only files that exist from before X until after X would be
1397 restored. Probably the major cost here is the extra date container in
1398 each row of the files table.
1400 Thanks for "listening". I hope some of this helps. If you want to
1401 contact me, please send me an email - I read some but not all of the
1402 mailing list traffic and might miss a reply there.
1404 Please accept my compliments for bacula. It is doing a great job for
1405 me!! I sympathize with you in the need to wrestle with excelence in
1406 execution vs. excelence in feature inclusion.
1411 ==============================
1414 - Design at hierarchial storage for Bacula. Migration and Clone.
1415 - Implement FSM (File System Modules).
1416 - Audit M_ error codes to ensure they are correct and consistent.
1417 - Add variable break characters to lex analyzer.
1418 Either a bit mask or a string of chars so that
1419 the caller can change the break characters.
1420 - Make a single T_BREAK to replace T_COMMA, etc.
1421 - Ensure that File daemon and Storage daemon can
1422 continue a save if the Director goes down (this
1423 is NOT currently the case). Must detect socket error,
1424 buffer messages for later.
1425 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
1426 - Add ability to backup to two Storage devices (two SD sessions) at
1427 the same time -- e.g. onsite, offsite.
1428 - Add the ability to consolidate old backup sets (basically do a restore
1429 to tape and appropriately update the catalog). Compress Volume sets.
1430 Might need to spool via file is only one drive is available.
1431 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
1432 someway to do so with every volume that has less than x% valid
1436 Migration: Move a backup from one Volume to another
1437 Clone: Copy a backup -- two Volumes
1439 Bacula Migration is based on Jobs (apparently Networker is file by file).
1441 Migration triggered by:
1445 Highwater mark (keep total size)
1450 ======================================================
1452 It is somewhat like a Full save becomes an incremental since
1453 the Base job (or jobs) plus other non-base files.
1455 - A Base backup is same as Full backup, just different type.
1456 - New BaseFiles table that contains:
1458 BaseJobId - Base JobId referenced for this FileId (needed ???)
1459 JobId - JobId currently running
1460 FileId - File not backed up, exists in Base Job
1461 FileIndex - FileIndex from Base Job.
1462 i.e. for each base file that exists but is not saved because
1463 it has not changed, the File daemon sends the JobId, BaseId,
1464 FileId, FileIndex back to the Director who creates the DB entry.
1465 - To initiate a Base save, the Director sends the FD
1466 the FileId, and full filename for each file in the Base.
1467 - When the FD finds a Base file, he requests the Director to
1468 send him the full File entry (stat packet plus MD5), or
1469 conversely, the FD sends it to the Director and the Director
1470 says yes or no. This can be quite rapid if the FileId is kept
1471 by the FD for each Base Filename.
1472 - It is probably better to have the comparison done by the FD
1473 despite the fact that the File entry must be sent across the
1475 - An alternative would be to send the FD the whole File entry
1476 from the start. The disadvantage is that it requires a lot of
1477 space. The advantage is that it requires less communications
1479 - The Job record must be updated to indicate that one or more
1481 - At end of Job, FD returns:
1482 1. Count of base files/bytes not written to tape (i.e. matches)
1483 2. Count of base file that were saved i.e. had changed.
1484 - No tape record would be written for a Base file that matches, in the
1485 same way that no tape record is written for Incremental jobs where
1486 the file is not saved because it is unchanged.
1487 - On a restore, all the Base file records must explicitly be
1488 found from the BaseFile tape. I.e. for each Full save that is marked
1489 to have one or more Base Jobs, search the BaseFile for all occurrences
1491 - An optimization might be to make the BaseFile have:
1497 This would avoid the need to explicitly fetch each File record for
1498 the Base job. The Base Job record will be fetched to get the
1499 VolSessionId and VolSessionTime.
1500 =========================================================
1503 ==========================================================
1505 For each Incremental job that is run, there may be files that
1506 were found but not saved because they were locked (this applies
1507 only to Windows). Such a system could send back to the Director
1508 a list of Unsaved files.
1510 - New UnSavedFiles table that contains:
1514 - Then in the next Incremental job, the list of Unsaved Files will be
1515 feed to the FD, who will ensure that they are explicitly chosen even
1516 if standard date/time check would not have selected them.
1517 =============================================================
1521 Multiple drive autochanger data: see Alan Brown
1522 mtx -f xxx unloadStorage Element 1 is Already Full(drive 0 was empty)
1523 Unloading Data Transfer Element into Storage Element 1...source Element
1524 Address 480 is Empty
1526 (drive 0 was empty and so was slot 1)
1527 > mtx -f xxx load 15 0
1528 no response, just returns to the command prompt when complete.
1529 > mtx -f xxx status Storage Changer /dev/changer:2 Drives, 60 Slots ( 2 Import/Export )
1530 Data Transfer Element 0:Full (Storage Element 15 Loaded):VolumeTag = HX001
1531 Data Transfer Element 1:Empty
1532 Storage Element 1:Empty
1533 Storage Element 2:Full :VolumeTag=HX002
1534 Storage Element 3:Full :VolumeTag=HX003
1535 Storage Element 4:Full :VolumeTag=HX004
1536 Storage Element 5:Full :VolumeTag=HX005
1537 Storage Element 6:Full :VolumeTag=HX006
1538 Storage Element 7:Full :VolumeTag=HX007
1539 Storage Element 8:Full :VolumeTag=HX008
1540 Storage Element 9:Full :VolumeTag=HX009
1541 Storage Element 10:Full :VolumeTag=HX010
1542 Storage Element 11:Empty
1543 Storage Element 12:Empty
1544 Storage Element 13:Empty
1545 Storage Element 14:Empty
1546 Storage Element 15:Empty
1547 Storage Element 16:Empty....
1548 Storage Element 28:Empty
1549 Storage Element 29:Full :VolumeTag=CLNU01L1
1550 Storage Element 30:Empty....
1551 Storage Element 57:Empty
1552 Storage Element 58:Full :VolumeTag=NEX261L2
1553 Storage Element 59 IMPORT/EXPORT:Empty
1554 Storage Element 60 IMPORT/EXPORT:Empty
1556 Unloading Data Transfer Element into Storage Element 15...done
1558 (just to verify it remembers where it came from, however it can be
1559 overrriden with mtx unload {slotnumber} to go to any storage slot.)
1561 There needs to be a table of drive # to devices somewhere - If there are
1562 multiple changers or drives there may not be a 1:1 correspondance between
1563 changer drive number and system device name - and depending on the way the
1564 drives are hooked up to scsi busses, they may not be linearly numbered
1565 from an offset point either.something like
1567 Autochanger drives = 2
1568 Autochanger drive 0 = /dev/nst1
1569 Autochanger drive 1 = /dev/nst2
1570 IMHO, it would be _safest_ to use explicit mtx unload commands at all
1571 times, not just for multidrive changers. For a 1 drive changer, that's
1577 MTX's manpage (1.2.15):
1578 unload [<slotnum>] [ <drivenum> ]
1579 Unloads media from drive <drivenum> into slot
1580 <slotnum>. If <drivenum> is omitted, defaults to
1581 drive 0 (as do all commands). If <slotnum> is
1582 omitted, defaults to the slot that the drive was
1583 loaded from. Note that there's currently no way
1584 to say 'unload drive 1's media to the slot it
1585 came from', other than to explicitly use that
1586 slot number as the destination.AB
1592 undef# camcontrol devlist
1593 <WANGTEK 51000 SCSI M74H 12B3> at scbus0 target 2 lun 0 (pass0,sa0)
1594 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 0 (pass1,sa1)
1595 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 1 (pass2)
1597 tapeinfo -f /dev/sg0 with a bad tape in drive 1:
1598 [kern@rufus mtx-1.2.17kes]$ ./tapeinfo -f /dev/sg0
1599 Product Type: Tape Drive
1601 Product ID: 'C5713A '
1603 Attached Changer: No
1604 TapeAlert[3]: Hard Error: Uncorrectable read/write error.
1605 TapeAlert[20]: Clean Now: The tape drive neads cleaning NOW.
1612 Medium Type: Not Loaded
1615 DataCompEnabled: yes
1616 DataCompCapable: yes
1617 DataDeCompEnabled: yes
1624 Handling removable disks
1626 From: Karl Cunningham <karlc@keckec.com>
1628 My backups are only to hard disk these days, in removable bays. This is my
1629 idea of how a backup to hard disk would work more smoothly. Some of these
1630 things Bacula does already, but I mention them for completeness. If others
1631 have better ways to do this, I'd like to hear about it.
1633 1. Accommodate several disks, rotated similar to how tapes are. Identified
1634 by partition volume ID or perhaps by the name of a subdirectory.
1635 2. Abort & notify the admin if the wrong disk is in the bay.
1636 3. Write backups to different subdirectories for each machine to be backed
1638 4. Volumes (files) get created as needed in the proper subdirectory, one
1640 5. When a disk is recycled, remove or zero all old backup files. This is
1641 important as the disk being recycled may be close to full. This may be
1642 better done manually since the backup files for many machines may be
1643 scattered in many subdirectories.
1648 - Make sure that all do_prompt() calls in Dir check for
1649 -1 (error) and -2 (cancel) returns.
1650 - Fix foreach_jcr() to have free_jcr() inside next().
1651 jcr=jcr_walk_start();
1652 for ( ; jcr; (jcr=jcr_walk_next(jcr)) )
1655 - A Volume taken from Scratch should take on the retention period
1657 - Correct doc for Maximum Changer Wait (and others) accepting only
1659 - Implement status that shows why a job is being held in reserve, or
1660 rather why none of the drives are suitable.
1661 - Implement a way to disable a drive (so you can use the second
1662 drive of an autochanger, and the first one will not be used or
1664 - Make sure Maximum Volumes is respected in Pools when adding
1665 Volumes (e.g. when pulling a Scratch volume).
1666 - Keep same dcr when switching device ...
1667 - Implement code that makes the Dir aware that a drive is an
1668 autochanger (so the user doesn't need to use the Autochanger = yes
1670 - Make catalog respect ACL.
1671 - Add recycle count to Media record.
1672 - Add initial write date to Media record.
1673 - Fix store_yesno to be store_bitmask.
1674 --- create_file.c.orig Fri Jul 8 12:13:05 2005
1675 +++ create_file.c Fri Jul 8 12:13:07 2005
1677 attr->ofname, be.strerror());
1680 + } else if(S_ISSOCK(attr->statp.st_mode)) {
1681 + Dmsg1(200, "Skipping socket: %s\n", attr->ofname);
1683 Dmsg1(200, "Restore node: %s\n", attr->ofname);
1684 if (mknod(attr->ofname, attr->statp.st_mode, attr->statp.st_rdev) != 0 && errno != EEXIST) {
1685 - Add true/false to conf same as yes/no
1686 - Reserve blocks other restore jobs when first cannot connect to SD.
1687 - Fix Maximum Changer Wait, Maximum Open Wait, Maximum Rewind Wait to
1688 accept time qualifiers.
1689 - Does ClientRunAfterJob fail the job on a bad return code?
1690 - Make hardlink code at line 240 of find_one.c use binary search.
1691 - Add ACL error messages in src/filed/acl.c.
1692 - Make authentication failures single threaded.
1693 - Make Dir and SD authentication errors single threaded.
1694 - Fix catreq.c digestbuf at line 411 in src/dird/catreq.c
1695 - Make base64.c (bin_to_base64) take a buffer length
1696 argument to avoid overruns.
1697 and verify that other buffers cannot overrun.
1698 - Implement VolumeState as discussed with Arno.
1699 - Add LocationId to update volume
1707 - Add Comment to Media record
1708 - Fix auth compatibility with 1.38
1709 - Update dbcheck to include Log table
1710 - Update llist to include new fields.
1711 - Make unmount unload autochanger. Make mount load slot.
1712 - Fix bscan to report the JobType when restoring a job.
1713 - Fix wx-console scanning problem with commas in names.
1714 - Add manpages to the list of directories for make install. Notify
1716 - Add bconsole option to use stdin/out instead of conio.
1717 - Fix ClientRunBefore/AfterJob compatibility.
1718 - Ensure that connection to daemon failure always indicates what
1719 daemon it was trying to connect to.
1720 - Freespace on DVD requested over and over even with no intervening
1722 - .update volume [enabled|disabled|*see below]
1723 > However, I could easily imagine an option to "update slots" that says
1724 > "enable=yes|no" that would automatically enable or disable all the Volumes
1725 > found in the autochanger. This will permit the user to optionally mark all
1726 > the Volumes in the magazine disabled prior to taking them offsite, and mark
1727 > them all enabled when bringing them back on site. Coupled with the options
1728 > to the slots keyword, you can apply the enable/disable to any or all volumes.
1729 - Restricted consoles start in the Default catalog even if it
1731 - When reading through parts on the DVD, the DVD is mounted and
1732 unmounted for each part.
1733 - Make sure that the restore options don't permit "seeing" other
1735 - Restore of a raw drive should not try to check the volume size.
1736 - Lock tape drive door when open()
1737 - Make release unload any autochanger.
1738 - Arno's reservation deadlock.