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 - Look at moving the Storage directive from the Job to the
45 Pool in the default conf files.
46 - Make sure the new level=Full syntax is used in all
47 example conf files (especially in the manual).
48 - Fix prog copyright (SD) all other files.
49 - Migration Volume span bug
52 - Test FIFO backup/restore -- make regression
54 - Add encryption regression tests
55 - Test Volume compatibility between machine architectures
56 - Encryption documentation
57 - Wrong jobbytes with query 12 (todo)
58 - bacula-1.38.2-ssl.patch
59 - Bare-metal recovery Windows (todo)
60 - Document need for UTF-8 format
65 - Implement Python event for backing up/restoring a file.
66 - Change dbcheck to tell users to use native tools for fixing
67 broken databases, and to ensure they have the proper indexes.
68 - add udev rules for Bacula devices.
69 - If a job terminates, the DIR connection can close before the
70 Volume info is updated, leaving the File count wrong.
71 - Look at why SIGPIPE during connection can cause seg fault in
72 writing the daemon message, when Dir dropped to bacula:bacula
73 - Look at zlib 32 => 64 problems.
74 - Try turning on disk seek code.
75 - Possibly turn on St. Bernard code.
76 - Fix bextract to restore ACLs, or better yet, use common routines.
77 - Do we migrate appendable Volumes?
78 - Remove queue.c code.
79 - Some users claim that they must do two prune commands to get a
80 Volume marked as purged.
81 - Print warning message if LANG environment variable does not specify
83 - New dot commands from Arno.
84 .show device=xxx lists information from one storage device, including
85 devices (I'm not even sure that information exists in the DIR...)
86 .move eject device=xxx mostly the same as 'unmount xxx' but perhaps with
87 better machine-readable output like "Ok" or "Error busy"
88 .move eject device=xxx toslot=yyy the same as above, but with a new
89 target slot. The catalog should be updated accordingly.
90 .move transfer device=xxx fromslot=yyy toslot=zzz
93 - Check to see if jcr->stime is lost during rescheduling of
95 - Fix re-read of last block to check if job has actually written
96 a block, and check if block was written by a different job
97 (i.e. multiple simultaneous jobs writing).
98 - Figure out how to configure query.sql. Suggestion to use m4:
100 changequote(`[',`]')dnl
101 ==== query.sql.in ===
102 :List next 20 volumes to expire
104 Pool.Name AS PoolName,
109 [ FROM_UNIXTIME(UNIX_TIMESTAMP(Media.LastWritten) Media.VolRetention) AS Expire, ])dnl
111 [ media.lastwritten + interval '1 second' * media.volretention as expire, ])dnl
115 ON Media.PoolId=Pool.PoolId
116 WHERE Media.LastWritten>0
120 Command: m4 -DmySQL changequote.m4 query.sql.in >query.sql
122 The problem is that it requires m4, which is not present on all machines
124 - Get Perl replacement for bregex.c
125 - Given all the problems with FIFOs, I think the solution is to do something a
126 little different, though I will look at the code and see if there is not some
127 simple solution (i.e. some bug that was introduced). What might be a better
128 solution would be to use a FIFO as a sort of "key" to tell Bacula to read and
129 write data to a program rather than the FIFO. For example, suppose you
134 Then, I could imagine if you backup and restore this file with a direct
135 reference as is currently done for fifos, instead, during backup Bacula will
138 /home/kern/my-fifo.backup
140 and read the data that my-fifo.backup writes to stdout. For restore, Bacula
143 /home/kern/my-fifo.restore
145 and send the data backed up to stdout. These programs can either be an
146 executable or a shell script and they need only read/write to stdin/stdout.
148 I think this would give a lot of flexibility to the user without making any
149 significant changes to Bacula.
154 select FilenameId from Filename where Name='';
155 # Get list of all directories referenced in a Backup.
156 select Path.Path from Path,File where File.JobId=nnn and
157 File.FilenameId=(FilenameId-from-above) and File.PathId=Path.PathId
158 order by Path.Path ASC;
160 - Look into using Dart for testing
161 http://public.kitware.com/Dart/HTML/Index.shtml
163 - Look into replacing autotools with cmake
164 http://www.cmake.org/HTML/Index.html
166 === Migration from David ===
167 What I'd like to see:
170 Name = "<poolname>-migrate"
174 Migration Selection Type = LowestUtil | OldestVol | PoolOccupancy |
175 Client | PoolResidence | Volume | JobName | SQLquery
176 Migration Selection Pattern = "regexp"
177 Next Pool = <override>
180 There should be no need for a Level (migration is always Full, since you
181 don't calculate differential/incremental differences for migration),
182 Storage should be determined by the volume types in the pool, and Client
183 is really a selection issue. Migration should always occur to the
184 NextPool defined in the pool definition. If no nextpool is defined, the
185 job should end with a reason of "no place to go". If Next Pool statement
186 is present, we override the check in the pool definition and use the
189 Here's how I'd define Migration Selection Types:
192 Client -- Migrate data from selected client only. Migration Selection
193 Pattern regexp provides pattern to select client names, eg ^FS00* makes
194 all client names starting with FS00 eligible for migration.
196 Jobname -- Migration all jobs matching name. Migration Selection Pattern
197 regexp provides pattern to select jobnames existing in pool.
199 Volume -- Migrate all data on specified volumes. Migration Selection
200 Pattern regexp provides selection criteria for volumes to be migrated.
201 Volumes must exist in pool to be eligible for migration.
205 LowestUtil -- Identify the volume in the pool with the least data on it
206 and empty it. No Migration Selection Pattern required.
208 OldestVol -- Identify the LRU volume with data written, and empty it. No
209 Migration Selection Pattern required.
211 PoolOccupancy -- if pool occupancy exceeds <highmig>, migrate volumes
212 (starting with most full volumes) until pool occupancy drops below
213 <lowmig>. Pool highmig and lowmig values are in pool definition, no
214 Migration Selection Pattern required.
218 SQLQuery -- Migrate all jobuids returned by the supplied SQL query.
219 Migration Selection Pattern contains SQL query to execute; should return
220 a list of 1 or more jobuids to migrate.
222 PoolResidence -- Migrate data sitting in pool for longer than
223 PoolResidence value in pool definition. Migration Selection Pattern
224 optional; if specified, override value in pool definition (value in
228 [ possibly a Python event -- kes ]
230 - Mount on an Autochanger with no tape in the drive causes:
231 Automatically selected Storage: LTO-changer
232 Enter autochanger drive[0]: 0
233 3301 Issuing autochanger "loaded drive 0" command.
234 3302 Autochanger "loaded drive 0", result: nothing loaded.
235 3301 Issuing autochanger "loaded drive 0" command.
236 3302 Autochanger "loaded drive 0", result: nothing loaded.
237 3902 Cannot mount Volume on Storage Device "LTO-Drive1" (/dev/nst0) because:
238 Couldn't rewind device "LTO-Drive1" (/dev/nst0): ERR=dev.c:678 Rewind error on "LTO-Drive1" (/dev/nst0). ERR=No medium found.
239 3905 Device "LTO-Drive1" (/dev/nst0) open but no Bacula volume is mounted.
240 If this is not a blank tape, try unmounting and remounting the Volume.
241 - If Drive 0 is blocked, and drive 1 is set "Autoselect=no", drive 1 will
243 - Autochanger did not change volumes.
244 select * from Storage;
245 +-----------+-------------+-------------+
246 | StorageId | Name | AutoChanger |
247 +-----------+-------------+-------------+
248 | 1 | LTO-changer | 0 |
249 +-----------+-------------+-------------+
250 05-May 03:50 roxie-sd: 3302 Autochanger "loaded drive 0", result is Slot 11.
251 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Warning: Director wanted Volume "LT
252 Current Volume "LT0-002" not acceptable because:
253 1997 Volume "LT0-002" not in catalog.
254 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Error: Autochanger Volume "LT0-002"
255 Setting InChanger to zero in catalog.
256 05-May 03:50 roxie-dir: Tibs.2006-05-05_03.05.02 Error: Unable to get Media record
258 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: Error getting Volume i
259 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: Job 530 canceled.
260 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: spool.c:249 Fatal appe
261 05-May 03:49 Tibs: Tibs.2006-05-05_03.05.02 Fatal error: c:\cygwin\home\kern\bacula
270 FirstWritten: 2006-05-05 03:11:54
271 LastWritten: 2006-05-05 03:50:23
272 LabelDate: 2005-12-26 16:52:40
283 VolRetention: 31,536,000
295 Note VolStatus is blank!!!!!
302 FirstWritten: 0000-00-00 00:00:00
303 LastWritten: 0000-00-00 00:00:00
304 LabelDate: 2005-12-26 16:52:40
315 VolRetention: 31,536,000
328 Automatically selected Storage: LTO-changer
329 Enter autochanger drive[0]: 0
330 3301 Issuing autochanger "loaded drive 0" command.
331 3302 Autochanger "loaded drive 0", result: nothing loaded.
332 3301 Issuing autochanger "loaded drive 0" command.
333 3302 Autochanger "loaded drive 0", result: nothing loaded.
334 3902 Cannot mount Volume on Storage Device "LTO-Drive1" (/dev/nst0) because:
335 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.
340 - Add VolumeState (enable, disable, archive)
341 - Add VolumeLock to prevent all but lock holder (SD) from updating
342 the Volume data (with the exception of VolumeState).
343 - The btape fill command does not seem to use the Autochanger
344 - Make Windows installer default to system disk drive.
345 - Look at using ioctl(FIOBMAP, ...) on Linux, and
346 DeviceIoControl(..., FSCTL_QUERY_ALLOCATED_RANGES, ...) on
347 Win32 for sparse files.
348 http://www.flexhex.com/docs/articles/sparse-files.phtml
349 http://www.informatik.uni-frankfurt.de/~loizides/reiserfs/fibmap.html
350 - Directive: at <event> "command"
351 - Command: pycmd "command" generates "command" event. How to
352 attach to a specific job?
353 - Integrate Christopher's St. Bernard code.
354 - run_cmd() returns int should return JobId_t
355 - get_next_jobid_from_list() returns int should return JobId_t
356 - Document export LDFLAGS=-L/usr/lib64
357 - Don't attempt to restore from "Disabled" Volumes.
358 - Network error on Win32 should set Win32 error code.
359 - What happens when you rename a Disk Volume?
360 - Job retention period in a Pool (and hence Volume). The job would
362 - Detect resource deadlock in Migrate when same job wants to read
363 and write the same device.
364 - Queue warning/error messages during restore so that they
365 are reported at the end of the report rather than being
366 hidden in the file listing ...
367 - Look at -D_FORTIFY_SOURCE=2
368 - Add Win32 FileSet definition somewhere
369 - Look at fixing restore status stats in SD.
370 - Make selection of Database used in restore correspond to
372 - Look at using ioctl(FIMAP) and FIGETBSZ for sparse files.
373 http://www.informatik.uni-frankfurt.de/~loizides/reiserfs/fibmap.html
374 - Implement a mode that says when a hard read error is
375 encountered, read many times (as it currently does), and if the
376 block cannot be read, skip to the next block, and try again. If
377 that fails, skip to the next file and try again, ...
379 create table LevelType (LevelType binary(1), LevelTypeLong tinyblob);
380 insert into LevelType (LevelType,LevelTypeLong) values
384 - Add ACL to restore only to original location.
385 - Show files/second in client status output.
386 - Add a recursive mark command (rmark) to restore.
387 - "Minimum Job Interval = nnn" sets minimum interval between Jobs
388 of the same level and does not permit multiple simultaneous
389 running of that Job (i.e. lets any previous invocation finish
390 before doing Interval testing).
391 - Look at simplifying File exclusions.
392 - New directive "Delete purged Volumes"
393 - new pool XXX with ScratchPoolId = MyScratchPool's PoolId and
394 let it fill itself, and RecyclePoolId = XXX's PoolId so I can
395 see if it become stable and I just have to supervise
397 - If I want to remove this pool, I set RecyclePoolId = MyScratchPool's
398 PoolId, and when it is empty remove it.
399 - Figure out how to recycle Scratch volumes back to the Scratch Pool.
401 - Allow Check Labels to be used with Bacula labels.
402 - "Resuming" a failed backup (lost line for example) by using the
403 failed backup as a sort of "base" job.
405 - Email to the user when the tape is about to need changing x
406 days before it needs changing.
407 - Command to show next tape that will be used for a job even
408 if the job is not scheduled.
409 - From: Arunav Mandal <amandal@trolltech.com>
410 1. When jobs are running and bacula for some reason crashes or if I do a
411 restart it remembers and jobs it was running before it crashed or restarted
412 as of now I loose all jobs if I restart it.
414 2. When spooling and in the midway if client is disconnected for instance a
415 laptop bacula completely discard the spool. It will be nice if it can write
416 that spool to tape so there will be some backups for that client if not all.
418 3. We have around 150 clients machines it will be nice to have a option to
419 upgrade all the client machines bacula version automatically.
421 4. Atleast one connection should be reserved for the bconsole so at heavy load
422 I should connect to the director via bconsole which at sometimes I can't
424 5. Another most important feature that is missing, say at 10am I manually
425 started backup of client abc and it was a full backup since client abc has
426 no backup history and at 10.30am bacula again automatically started backup of
427 client abc as that was in the schedule. So now we have 2 multiple Full
428 backups of the same client and if we again try to start a full backup of
429 client backup abc bacula won't complain. That should be fixed.
431 - Fix bpipe.c so that it does not modify results pointer.
432 ***FIXME*** calling sequence should be changed.
433 - For Windows disaster recovery see http://unattended.sf.net/
434 - regardless of the retention period, Bacula will not prune the
435 last Full, Diff, or Inc File data until a month after the
436 retention period for the last Full backup that was done.
437 - update volume=xxx --- add status=Full
438 - Remove old spool files on startup.
439 - Exclude SD spool/working directory.
440 - Refuse to prune last valid Full backup. Same goes for Catalog.
442 - Make a callback when Rerun failed levels is called.
443 - Give Python program access to Scheduled jobs.
444 - Add setting Volume State via Python.
445 - Python script to save with Python, not save, save with Bacula.
446 - Python script to do backup.
448 - Change the Priority, Client, Storage, JobStatus (error)
449 at the start of a job.
450 - Why is SpoolDirectory = /home/bacula/spool; not reported
451 as an error when writing a DVD?
452 - Make bootstrap file handle multiple MediaTypes (SD)
453 - Remove all old Device resource code in Dir and code to pass it
454 back in SD -- better, rework it to pass back device statistics.
455 - Check locking of resources -- be sure to lock devices where previously
456 resources were locked.
457 - The last part is left in the spool dir.
460 - In restore don't compare byte count on a raw device -- directory
461 entry does not contain bytes.
465 MA = (last_MA * 3 + rate) / 4
466 rate = (bytes - last_bytes) / (runtime - last_runtime)
467 - Max Vols limit in Pool off by one?
468 - Implement Files/Bytes,... stats for restore job.
469 - Implement Total Bytes Written, ... for restore job.
470 - Despool attributes simultaneously with data in a separate
471 thread, rejoined at end of data spooling.
472 - 7. Implement new Console commands to allow offlining/reserving drives,
473 and possibly manipulating the autochanger (much asked for).
474 - Add start/end date editing in messages (%t %T, %e?) ...
475 - Add ClientDefs similar to JobDefs.
476 - Print more info when bextract -p accepts a bad block.
477 - Fix FD JobType to be set before RunBeforeJob in FD.
478 - Look at adding full Volume and Pool information to a Volume
479 label so that bscan can get *all* the info.
480 - If the user puts "Purge Oldest Volume = yes" or "Recycle Oldest Volume = yes"
481 and there is only one volume in the pool, refuse to do it -- otherwise
482 he fills the Volume, then immediately starts reusing it.
483 - Implement copies and stripes.
484 - Add history file to console.
485 - Each file on tape creates a JobMedia record. Peter has 4 million
486 files spread over 10000 tape files and four tapes. A restore takes
487 16 hours to build the restore list.
488 - Add and option to check if the file size changed during backup.
489 - Make sure SD deletes spool files on error exit.
490 - Delete old spool files when SD starts.
491 - When labeling tapes, if you enter 000026, Bacula uses
492 the tape index rather than the Volume name 000026.
493 - Add offline tape command to Bacula console.
495 Enter MediaId or Volume name: 32
496 Enter new Volume name: DLT-20Dec04
497 Automatically selected Pool: Default
498 Connecting to Storage daemon DLTDrive at 192.168.68.104:9103 ...
499 Sending relabel command from "DLT-28Jun03" to "DLT-20Dec04" ...
500 block.c:552 Write error at 0:0 on device /dev/nst0. ERR=Bad file descriptor.
501 Error writing final EOF to tape. This tape may not be readable.
502 dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
503 askdir.c:219 NULL Volume name. This shouldn't happen!!!
504 3912 Failed to label Volume: ERR=dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
505 Label command failed for Volume DLT-20Dec04.
506 Do not forget to mount the drive!!!
507 - Bug: if a job is manually scheduled to run later, it does not appear
508 in any status report and cannot be cancelled.
510 ==== Keeping track of deleted/new files ====
511 - To mark files as deleted, run essentially a Verify to disk, and
512 when a file is found missing (MarkId != JobId), then create
513 a new File record with FileIndex == -1. This could be done
514 by the FD at the same time as the backup.
516 My "trick" for keeping track of deletions is the following.
517 Assuming the user turns on this option, after all the files
518 have been backed up, but before the job has terminated, the
519 FD will make a pass through all the files and send their
520 names to the DIR (*exactly* the same as what a Verify job
521 currently does). This will probably be done at the same
522 time the files are being sent to the SD avoiding a second
523 pass. The DIR will then compare that to what is stored in
524 the catalog. Any files in the catalog but not in what the
525 FD sent will receive a catalog File entry that indicates
526 that at that point in time the file was deleted. This
527 either transmitted to the FD or simultaneously computed in
528 the FD, so that the FD can put a record on the tape that
529 indicates that the file has been deleted at this point.
530 A delete file entry could potentially be one with a FileIndex
531 of 0 or perhaps -1 (need to check if FileIndex is used for
532 some other thing as many of the Bacula fields are "overloaded"
535 During a restore, any file initially picked up by some
536 backup (Full, ...) then subsequently having a File entry
537 marked "delete" will be removed from the tree, so will not
538 be restored. If a file with the same name is later OK it
539 will be inserted in the tree -- this already happens. All
540 will be consistent except for possible changes during the
543 Since I'm on the subject, some of you may be wondering what
544 the utility of the in memory tree is if you are going to
545 restore everything (at least it comes up from time to time
546 on the list). Well, it is still *very* useful because it
547 allows only the last item found for a particular filename
548 (full path) to be entered into the tree, and thus if a file
549 is backed up 10 times, only the last copy will be restored.
550 I recently (last Friday) restored a complete directory, and
551 the Full and all the Differential and Incremental backups
552 spanned 3 Volumes. The first Volume was not even mounted
553 because all the files had been updated and hence backed up
554 since the Full backup was made. In this case, the tree
555 saved me a *lot* of time.
557 Make sure this information is stored on the tape too so
558 that it can be restored directly from the tape.
560 All the code (with the exception of formally generating and
561 saving the delete file entries) already exists in the Verify
562 Catalog command. It explicitly recognizes added/deleted files since
563 the last InitCatalog. It is more or less a "simple" matter of
564 taking that code and adapting it slightly to work for backups.
566 Comments from Martin Simmons (I think they are all covered):
567 Ok, that should cover the basics. There are few issues though:
569 - Restore will depend on the catalog. I think it is better to include the
570 extra data in the backup as well, so it can be seen by bscan and bextract.
572 - I'm not sure if it will preserve multiple hard links to the same inode. Or
573 maybe adding or removing links will cause the data to be dumped again?
575 - I'm not sure if it will handle renamed directories. Possibly it will work
576 by dumping the whole tree under a renamed directory?
578 - It remains to be seen how the backup performance of the DIR's will be
579 affected when comparing the catalog for a large filesystem.
583 How about introducing a Type = MgmtPolicy job type? That job type would
584 be responsible for scanning the Bacula environment looking for specific
585 conditions, and submitting the appropriate jobs for implementing said
589 Name = "Migration-Policy"
591 Policy Selection Job Type = Migrate
592 Scope = "<keyword> <operator> <regexp>"
593 Threshold = "<keyword> <operator> <regexp>"
594 Job Template = <template-name>
597 Where <keyword> is any legal job keyword, <operator> is a comparison
598 operator (=,<,>,!=, logical operators AND/OR/NOT) and <regexp> is a
599 appropriate regexp. I could see an argument for Scope and Threshold
600 being SQL queries if we want to support full flexibility. The
601 Migration-Policy job would then get scheduled as frequently as a site
602 felt necessary (suggested default: every 15 minutes).
607 Name = "Migration-Policy"
609 Policy Selection Job Type = Migration
611 Threshold = "Migration Selection Type = LowestUtil"
612 Job Template = "MigrationTemplate"
615 would select all pools for examination and generate a job based on
616 MigrationTemplate to automatically select the volume with the lowest
617 usage and migrate it's contents to the nextpool defined for that pool.
619 This policy abstraction would be really handy for adjusting the behavior
620 of Bacula according to site-selectable criteria (one thing that pops
621 into mind is Amanda's ability to automatically adjust backup levels
622 depending on various criteria).
628 - Add Pool/Storage override regression test.
629 - Add delete JobId to regression.
630 - Add a regression test for dbcheck.
631 - New test to add bscan to four-concurrent-jobs regression,
632 i.e. after the four-concurrent jobs zap the
633 database as is done in the bscan-test, then use bscan to
634 restore the database, do a restore and compare with the
636 - Add restore of specific JobId to regression (item 3
637 on the restore prompt)
638 - Add IPv6 to regression
639 - Add database test to regression. Test each function like delete,
642 - AntiVir can slow down backups on Win32 systems.
643 - Win32 systems with FAT32 can be much slower than NTFS for
644 more than 1000 files per directory.
648 - A HOLD command to stop all jobs from starting.
649 - A PAUSE command to pause all running jobs ==> release the
651 - Media Type = LTO,LTO-2,LTO-3
652 Media Type Read = LTO,LTO2,LTO3
653 Media Type Write = LTO2, LTO3
655 === From Carsten Menke <bootsy52@gmx.net>
657 Following is a list of what I think in the situations where I'm faced with,
658 could be a usefull enhancement to bacula, which I'm certain other users will
659 benefit from as well.
661 1. NextJob/NextJobs Directive within a Job Resource in the form of
662 NextJobs = job1,job2.
665 I currently solved the problem with running multiple jobs each after each
666 by setting the Max Wait Time for a job to 8 hours, and give
667 the jobs different Priorities. However, there scenarios where
668 1 Job is directly depending on another job, so if the former job fails,
669 the job after it needn't to be run
670 while maybe other jobs should run despite of that
673 A Backup Job and a Verify job, if the backup job fails there is no need to run
674 the verify job, as the backup job already failed. However, one may like
675 to backup the Catalog to disk despite of that the main backup job failed.
678 I see that this is related to the Event Handlers which are on the ToDo
679 list, also it is maybe a good idea to check for the return value and
680 execute different actions based on the return value
683 3. offline capability to bconsole
686 Currently I use a script which I execute within the last Job via the
687 RunAfterJob Directive, to release and eject the tape.
688 So I have to call bconsole "release=Storage-Name" and afterwards
689 mt -f /dev/nst0 eject to get the tape out.
691 If I have multiple Storage Devices, than these may not be /dev/nst0 and
692 I have to modify the script or call it with parameters etc.
693 This would actually not be needed, as everything is already defined
694 in bacula-sd.conf and if I can invoke bconsole with the
695 storage name via $1 in the script than I'm done and information is
698 4. %s for Storage Name added to the chars being substituted in "RunAfterJob"
702 For the reason mentioned in 3. to have the ability to call a
703 script with /scripts/foobar %s and in the script use $1
704 to pass the Storage Name to bconsole
706 5. Setting Volume State within a Job Resource
709 Instead of using "Maximum Volume Jobs" in the Pool Resource,
710 I would have the possibilty to define
711 in a Job Resource that after this certain job is run, the Volume State
712 should be set to "Volume State = Used", this give more flexibility (IMHO).
714 6. Localization of Bacula Messages
717 Unfortunatley many,many people I work with don't speak english very well.
718 So if at least the Reporting messages would be localized then they
719 would understand that they have to change the tape,etc. etc.
721 I volunteer to do the german translations, and if I can convince my wife also
722 french and Morre (western african language).
724 7. OK, this is evil, probably bound to security risks and maybe not possible
725 due to the design of bacula.
727 Implementation of Backtics ( `command` ) for shell comand execution to
728 the "Label Format" Directive.
732 Currently I have defined BACULA_DAY_OF_WEEK="day1|day2..." resulting in
733 Label Format = "HolyBackup-${BACULA_DAY_OF_WEEK[${WeekDay}]}". If I could
734 use backticks than I could use "Label Format = HolyBackup-`date +%A` to have
735 the localized name for the day of the week appended to the
736 format string. Then I have the tape labeled automatically with weekday
737 name in the correct language.
739 - Yes, that is surely the case. I probably should turn those into Warning
740 errors. In addition, you just made me think that it might not be bad to
741 add an option to check the file size after backing up the file and
742 report if it changes. This would be done as an option because it would
745 Kern, good idea. If you do do that, mention in the output: file
746 shrunk, or file expanded, just to make it obvious to the user
747 (without having to the refer to file size), just how the file size
750 Would this option be for all file, or just one file? Or a fileset?
751 - Make output from status use html table tags for nicely
752 presenting in a browser.
753 - Can one write tapes faster with 8192 byte block sizes?
754 - Document security problems with the same password for everyone in
755 rpm and Win32 releases.
756 - Browse generations of files.
757 - I've seen an error when my catalog's File table fills up. I
758 then have to recreate the File table with a larger maximum row
759 size. Relevant information is at
760 http://dev.mysql.com/doc/mysql/en/Full_table.html ; I think the
761 "Installing and Configuring MySQL" chapter should talk a bit
762 about this potential problem, and recommend a solution.
763 - For Solaris must use POSIX awk.
764 - Want speed of writing to tape while despooling.
765 - Supported autochanger:
773 Wangtek 6525ES (SCSI-1 QIC drive, 525MB), under Linux 2.4.something,
774 bacula 1.36.0/1 works with blocksize 16k INSIDE bacula-sd.conf.
775 - Add regex from http://www.pcre.org to Bacula for Win32.
776 - Use only shell tools no make in CDROM package.
777 - Include within include does it work?
778 - Implement a Pool of type Cleaning?
779 - Implement VolReadTime and VolWriteTime in SD
780 - Modify Backing up Your Database to include a bootstrap file.
781 - Think about making certain database errors fatal.
782 - Look at correcting the time jump in the scheduler for daylight
783 savings time changes.
784 - Add a "real" timer to network connections.
785 - Promote to Full = Time period
786 - Check dates entered by user for correctness (month/day/... ranges)
787 - Compress restore Volume listing by date and first file.
788 - Look at patches/bacula_db.b2z postgresql that loops during restore.
790 - Perhaps add read/write programs and/or plugins to FileSets.
791 - How to handle backing up portables ...
792 - Add some sort of guaranteed Interval for upgrading jobs.
793 - Can we write the state file after every job terminates? On Win32
794 the system crashes and the state file is not updated.
797 Documentation to do: (any release a little bit at a time)
798 - Doc to do unmount before removing magazine.
799 - Alternative to static linking "ldd prog" save all binaries listed,
800 restore them and point LD_LIBRARY_PATH to them.
801 - Document add "</dev/null >/dev/null 2>&1" to the bacula-fd command line
802 - Document query file format.
803 - Add more documentation for bsr files.
804 - Document problems with Verify and pruning.
805 - Document how to use multiple databases.
806 - VXA drives have a "cleaning required"
807 indicator, but Exabyte recommends preventive cleaning after every 75
810 In this context, it should be noted that Exabyte has a command-line
811 vxatool utility available for free download. (The current version is
812 vxatool-3.72.) It can get diagnostic info, read, write and erase tapes,
813 test the drive, unload tapes, change drive settings, flash new firmware,
815 Of particular interest in this context is that vxatool <device> -i will
816 report, among other details, the time since last cleaning in tape motion
817 minutes. This information can be retrieved (and settings changed, for
818 that matter) through the generic-SCSI device even when Bacula has the
819 regular tape device locked. (Needless to say, I don't recommend
820 changing tape settings while a job is running.)
821 - Lookup HP cleaning recommendations.
822 - Lookup HP tape replacement recommendations (see trouble shooting autochanger)
823 - Document doing table repair
826 ===================================
827 - Add macro expansions in JobDefs.
828 Run Before Job = "SomeFile %{Level} %{Client}"
829 Write Bootstrap="/some/dir/%{JobName}_%{Client}.bsr"
830 - Use non-blocking network I/O but if no data is available, use
832 - Use gather write() for network I/O.
833 - Autorestart on crash.
834 - Add bandwidth limiting.
835 - Add acks every once and a while from the SD to keep
836 the line from timing out.
837 - When an error in input occurs and conio beeps, you can back
838 up through the prompt.
839 - Detect fixed tape block mode during positioning by looking at
840 block numbers in btape "test". Possibly adjust in Bacula.
841 - Fix list volumes to output volume retention in some other
842 units, perhaps via a directive.
843 - If opening a tape in read/write mode fails attempt to open
844 it in read-only mode, and mark the tape for read only.
845 - Allow Simultaneous Priorities = yes => run up to Max concurrent jobs even
846 with multiple priorities.
847 - If you use restore replace=never, the directory attributes for
848 non-existent directories will not be restored properly.
850 - see lzma401.zip in others directory for new compression
852 - Minimal autochanger handling in Bacula and in btape.
853 - Look into how tar does not save sockets and the possiblity of
854 not saving them in Bacula (Martin Simmons reported this).
855 - Fix restore jobs so that multiple jobs can run if they
856 are not using the same tape(s).
857 - Allow the user to select JobType for manual pruning/purging.
858 - bscan does not put first of two volumes back with all info in
860 - Implement the FreeBSD nodump flag in chflags.
861 - Figure out how to make named console messages go only to that
862 console and to the non-restricted console (new console class?).
863 - Make restricted console prompt for password if *ask* is set or
864 perhaps if password is undefined.
865 - Implement "from ISO-date/time every x hours/days/weeks/months" in
868 ==== from Marc Schoechlin
869 - the help-command should be more verbose
870 (it should explain the paramters of the different
872 -> its time-comsuming to consult the manual anytime
873 you need a special parameter
874 -> maybe its more easy to maintain this, if the
875 descriptions of that commands are outsourced to
877 - the cd-command should allow complete paths
878 i.e. cd /foo/bar/foo/bar
879 -> if a customer mails me the path to a certain file,
880 its faster to enter the specified directory
881 - if the password is not configured in bconsole.conf
882 you should be asked for it.
883 -> sometimes you like to do restore on a customer-machine
884 which shouldnt know the password for bacula.
885 -> adding the password to the file favours admins
886 to forget to remove the password after usage
888 the protection of that file is less important
889 - long-listed-output of commands should be scrollable
890 like the unix more/less-command does
891 -> if someone runs 200 and more machines, the lists could
892 be a little long and complex
893 - command-output should be shown column by column
894 to reduce scrolling and to increase clarity
896 - lsmark should list the selected files with full
898 - wildcards for selecting and file and directories would be nice
899 - any actions should be interuptable with STRG+C
900 - command-expansion would be pretty cool
902 - When the replace Never option is set, new directory permissions
903 are not restored. See bug 213. To fix this requires creating a
904 list of newly restored directories so that those directory
905 permissions *can* be restored.
906 - Compaction of Disk space by "migrating" Volumes that have pruned
907 Jobs (what criteria? size, #jobs, time).
908 - Add prune all command
909 - Document fact that purge can destroy a part of a restore by purging
910 one volume while others remain valid -- perhaps mark Jobs.
911 - Add multiple-media-types.txt
912 - look at mxt-changer.html
913 - Make ? do a help command (no return needed).
914 - Implement restore directory.
915 - Document streams and how to implement them.
916 - Try not to re-backup a file if a new hard link is added.
917 - Add feature to backup hard links only, but not the data.
918 - Fix stream handling to be simpler.
919 - Add Priority and Bootstrap to Run a Job.
920 - Eliminate Restore "Run Restore Job" prompt by allowing new "run command
922 - Remove View FileSet button from Run a Job dialog.
923 - Handle prompt for restore job at end of Restore command.
924 - Add display of total selected files to Restore window.
925 - Add tree pane to left of window.
926 - Add progress meter.
927 - Max wait time or max run time causes seg fault -- see runtime-bug.txt
928 - Document writing to a CD/DVD with Bacula.
929 - Add a "base" package to the window installer for pthreadsVCE.dll
930 which is needed by all packages.
931 - Add message to user to check for fixed block size when the forward
932 space test fails in btape.
933 - When unmarking a directory check if all files below are unmarked and
934 then remove the + flag -- in the restore tree.
935 - Possibly implement: Action = Unmount Device="TapeDrive1" in Admin jobs.
936 - Setup lrrd graphs: (http://www.linpro.no/projects/lrrd/) Mike Acar.
937 - Revisit the question of multiple Volumes (disk) on a single device.
938 - Add a block copy option to bcopy.
939 - Investigate adding Mac Resource Forks.
940 - Finish work on Gnome restore GUI.
941 - Fix "llist jobid=xx" where no fileset or client exists.
942 - For each job type (Admin, Restore, ...) require only the really necessary
943 fields.- Pass Director resource name as an option to the Console.
944 - Add a "batch" mode to the Console (no unsolicited queries, ...).
945 - Add a .list all files in the restore tree (probably also a list all files)
946 Do both a long and short form.
947 - Allow browsing the catalog to see all versions of a file (with
948 stat data on each file).
949 - Restore attributes of directory if replace=never set but directory
951 - Use SHA1 on authentication if possible.
952 - See comtest-xxx.zip for Windows code to talk to USB.
953 - Add John's appended files:
954 Appended = { /files/server/logs/http/*log }
955 and such files would be treated as follows.On a FULL backup, they would
956 be backed up like any other file.On an INCREMENTAL backup, where a
957 previous INCREMENTAL or FULL was already in thecatalogue and the length
958 of the file wasgreater than the length of the last backup, only thedata
959 added since the last backup will be dumped.On an INCREMENTAL backup, if
960 the length of the file is less than thelength of the file with the same
961 name last backed up, the completefile is dumped.On Windows systems, with
962 creation date of files, we can be evensmarter about this and not count
963 entirely upon the length.On a restore, the full and all incrementals
964 since it will beapplied in sequence to restore the file.
965 - Check new HAVE_WIN32 open bits.
966 - Check if the tape has moved before writing.
967 - Handling removable disks -- see below:
968 - Keep track of tape use time, and report when cleaning is necessary.
969 - Add FromClient and ToClient keywords on restore command (or
970 BackupClient RestoreClient).
971 - Implement a JobSet, which groups any number of jobs. If the
972 JobSet is started, all the jobs are started together.
973 Allow Pool, Level, and Schedule overrides.
974 - Enhance cancel to timeout BSOCK packets after a specific delay.
975 - Do scheduling by UTC using gmtime_r() in run_conf, scheduler, and
976 ua_status.!!! Thanks to Alan Brown for this tip.
977 - Look at updating Volume Jobs so that Max Volume Jobs = 1 will work
978 correctly for multiple simultaneous jobs.
979 - Correct code so that FileSet MD5 is calculated for < and | filename
981 - Implement the Media record flag that indicates that the Volume does disk
983 - Implement VolAddr, which is used when Volume is addressed like a disk,
984 and form it from VolFile and VolBlock.
985 - Make multiple restore jobs for multiple media types specifying
986 the proper storage type.
987 - Fix fast block rejection (stored/read_record.c:118). It passes a null
988 pointer (rec) to try_repositioning().
989 - Look at extracting Win data from BackupRead.
990 - Implement RestoreJobRetention? Maybe better "JobRetention" in a Job,
991 which would take precidence over the Catalog "JobRetention".
992 - Implement Label Format in Add and Label console commands.
993 - Possibly up network buffers to 65K. Put on variable.
994 - Put email tape request delays on one or more variables. User wants
995 to cancel the job after a certain time interval. Maximum Mount Wait?
996 - Job, Client, Device, Pool, or Volume?
997 Is it possible to make this a directive which is *optional* in multiple
998 resources, like Level? If so, I think I'd make it an optional directive
999 in Job, Client, and Pool, with precedence such that Job overrides Client
1000 which in turn overrides Pool.
1002 - New Storage specifications:
1003 - Want to write to multiple storage devices simultaneously
1004 - Want to write to multiple storage devices sequentially (in one job)
1005 - Want to read/write simultaneously
1006 - Key is MediaType -- it must match
1008 Passed to SD as a sort of BSR record called Storage Specification
1012 MediaType -> Next MediaType
1014 Device -> Next Device
1016 Allow multiple Storage specifications
1024 Allow Multiple Pool specifications (note, Pool currently
1026 Allow Multiple MediaType specifications in Dir conf
1027 Allow Multiple Device specifications in Dir conf
1028 Perhaps keep this in a single SSR
1029 Tie a Volume to a specific device by using a MediaType that
1030 is contained in only one device.
1031 In SD allow Device to have Multiple MediaTypes
1033 - Ideas from Jerry Scharf:
1034 First let's point out some big pluses that bacula has for this
1036 more importantly it's active. Thank you so much for that
1037 even more important, it's not flaky
1038 it has an open access catalog, opening many possibilities
1039 it's pushing toward heterogeneous systems capability
1041 Macintosh file client
1042 macs are an interesting niche, but I fear a server is a rathole
1043 working bare iron recovery for windows
1044 the option for inc/diff backups not reset on fileset revision
1045 a) use both change and inode update time against base time
1046 b) do the full catalog check (expensive but accurate)
1047 sizing guide (how much system is needed to back up N systems/files)
1048 consultants on using bacula in building a disaster recovery system
1049 an integration guide
1050 or how to get at fancy things that one could do with bacula
1051 logwatch code for bacula logs (or similar)
1052 linux distro inclusion of bacula (brings good and bad, but necessary)
1053 win2k/XP server capability (icky but you asked)
1054 support for Oracle database ??
1056 - Look at adding SQL server and Exchange support for Windows.
1057 - Make dev->file and dev->block_num signed integers so that -1 can
1058 be an invalid value which happens with BSR.
1059 - Create VolAddr for disk files in place of VolFile and VolBlock. This
1060 is needed to properly specify ranges.
1061 - Add progress of files/bytes to SD and FD.
1062 - Print warning message if FileId > 4 billion
1063 - do a "messages" before the first prompt in Console
1064 - Client does not show busy during Estimate command.
1065 - Implement Console mtx commands.
1066 - Implement a Mount Command and an Unmount Command where
1067 the users could specify a system command to be performed
1068 to do the mount, after which Bacula could attempt to
1069 read the device. This is for Removeable media such as a CDROM.
1070 - Most likely, this mount command would be invoked explicitly
1071 by the user using the current Console "mount" and "unmount"
1072 commands -- the Storage Daemon would do the right thing
1073 depending on the exact nature of the device.
1074 - As with tape drives, when Bacula wanted a new removable
1075 disk mounted, it would unmount the old one, and send a message
1076 to the user, who would then use "mount" as described above
1077 once he had actually inserted the disk.
1078 - Implement dump/print label to UA
1079 - Spool to disk only when the tape is full, then when a tape is hung move
1081 - bextract is sending everything to the log file ****FIXME****
1082 - Allow multiple Storage specifications (or multiple names on
1083 a single Storage specification) in the Job record. Thus a job
1084 can be backed up to a number of storage devices.
1085 - Implement some way for the File daemon to contact the Director
1086 to start a job or pass its DHCP obtained IP number.
1087 - Implement a query tape prompt/replace feature for a console
1088 - Copy console @ code to gnome2-console
1089 - Make AES the only encryption algorithm see
1090 http://csrc.nist.gov/CryptoToolkit/aes/). It's
1091 an officially adopted standard, has survived peer
1092 review, and provides keys up to 256 bits.
1093 - Take a careful look at SetACL http://setacl.sourceforge.net
1094 - Make tree walk routines like cd, ls, ... more user friendly
1095 by handling spaces better.
1096 - Make sure that Bacula rechecks the tape after the 20 min wait.
1097 - Set IO_NOWAIT on Bacula TCP/IP packets.
1098 - Try doing a raw partition backup and restore by mounting a
1100 - From Lars Kellers:
1101 Yes, it would allow to highly automatic the request for new tapes. If a
1102 tape is empty, bacula reads the barcodes (native or simulated), and if
1103 an unused tape is found, it runs the label command with all the
1104 necessary parameters.
1106 By the way can bacula automatically "move" an empty/purged volume say
1107 in the "short" pool to the "long" pool if this pool runs out of volume
1109 - What to do about "list files job=xxx".
1110 - Get and test MySQL 4.0
1111 - Look at how fuser works and /proc/PID/fd that is how Nic found the
1112 file descriptor leak in Bacula.
1113 - Implement WrapCounters in Counters.
1114 - Add heartbeat from FD to SD if hb interval expires.
1115 - Can we dynamically change FileSets?
1116 - If pool specified to label command and Label Format is specified,
1117 automatically generate the Volume name.
1118 - Why can't SQL do the filename sort for restore?
1119 - Add ExhautiveRestoreSearch
1120 - Look at the possibility of loading only the necessary
1121 data into the restore tree (i.e. do it one directory at a
1122 time as the user walks through the tree).
1123 - Possibly use the hash code if the user selects all for a restore command.
1124 - Fix "restore all" to bypass building the tree.
1125 - Prohibit backing up archive device (findlib/find_one.c:128)
1126 - Implement Release Device in the Job resource to unmount a drive.
1127 - Implement Acquire Device in the Job resource to mount a drive,
1128 be sure this works with admin jobs so that the user can get
1129 prompted to insert the correct tape. Possibly some way to say to
1130 run the job but don't save the files.
1131 - Make things like list where a file is saved case independent for
1134 - Use autochanger to handle multiple devices.
1135 - On Windows with very long path names, it may be impossible to create
1136 a file (and thus restore it) because the total length is too long.
1137 We must cd into the directory then create the file without the
1139 - Implement a Recycle command
1140 - Test a second language e.g. french.
1141 - Start working on Base jobs.
1142 - Implement UnsavedFiles DB record.
1143 - From Phil Stracchino:
1144 It would probably be a per-client option, and would be called
1145 something like, say, "Automatically purge obsoleted jobs". What it
1146 would do is, when you successfully complete a Differential backup of a
1147 client, it would automatically purge all Incremental backups for that
1148 client that are rendered redundant by that Differential. Likewise,
1149 when a Full backup on a client completed, it would automatically purge
1150 all Differential and Incremental jobs obsoleted by that Full backup.
1151 This would let people minimize the number of tapes they're keeping on
1152 hand without having to master the art of retention times.
1153 - When doing a Backup send all attributes back to the Director, who
1154 would then figure out what files have been deleted.
1155 - Currently in mount.c:236 the SD simply creates a Volume. It should have
1156 explicit permission to do so. It should also mark the tape in error
1157 if there is an error.
1158 - Cancel waiting for Client connect in SD if FD goes away.
1160 - Implement timeout in response() when it should come quickly.
1161 - Implement a Slot priority (loaded/not loaded).
1162 - Implement "vacation" Incremental only saves.
1163 - Implement create "FileSet"?
1164 - Add prefixlinks to where or not where absolute links to FD.
1165 - Issue message to mount a new tape before the rewind.
1166 - Simplified client job initiation for portables.
1167 - If SD cannot open a drive, make it periodically retry.
1168 - Add more of the config info to the tape label.
1170 - If tape is marked read-only, then try opening it read-only rather than
1171 failing, and remember that it cannot be written.
1172 - Refine SD waiting output:
1173 Device is being positioned
1174 > Device is being positioned for append
1175 > Device is being positioned to file x
1177 - Figure out some way to estimate output size and to avoid splitting
1178 a backup across two Volumes -- this could be useful for writing CDROMs
1179 where you really prefer not to have it split -- not serious.
1180 - Have SD compute MD5 or SHA1 and compare to what FD computes.
1181 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
1182 actual data on the Volume and compare it.
1183 - Implement Bacula plugins -- design API
1184 - Make bcopy read through bad tape records.
1185 - Program files (i.e. execute a program to read/write files).
1186 Pass read date of last backup, size of file last time.
1187 - Add Signature type to File DB record.
1188 - CD into subdirectory when open()ing files for backup to
1189 speed up things. Test with testfind().
1190 - Priority job to go to top of list.
1191 - Why are save/restore of device different sizes (sparse?) Yup! Fix it.
1192 - Implement some way for the Console to dynamically create a job.
1193 - Solaris -I on tar for include list
1194 - Need a verbose mode in restore, perhaps to bsr.
1195 - bscan without -v is too quiet -- perhaps show jobs.
1196 - Add code to reject whole blocks if not wanted on restore.
1197 - Check if we can increase Bacula FD priorty in Win2000
1198 - Make sure the MaxVolFiles is fully implemented in SD
1199 - Check if both CatalogFiles and UseCatalog are set to SD.
1200 - Possibly add email to Watchdog if drive is unmounted too
1201 long and a job is waiting on the drive.
1202 - After unmount, if restore job started, ask to mount.
1203 - Add UA rc and history files.
1204 - put termcap (used by console) in ./configure and
1205 allow -with-termcap-dir.
1206 - Fix Autoprune for Volumes to respect need for full save.
1207 - Compare tape to Client files (attributes, or attributes and data)
1208 - Make all database Ids 64 bit.
1209 - Allow console commands to detach or run in background.
1210 - Fix status delay on storage daemon during rewind.
1211 - Add SD message variables to control operator wait time
1212 - Maximum Operator Wait
1213 - Minimum Message Interval
1214 - Maximum Message Interval
1215 - Send Operator message when cannot read tape label.
1216 - Verify level=Volume (scan only), level=Data (compare of data to file).
1217 Verify level=Catalog, level=InitCatalog
1219 - Add keyword search to show command in Console.
1220 - Events : tape has more than xxx bytes.
1221 - Complete code in Bacula Resources -- this will permit
1222 reading a new config file at any time.
1223 - Handle ctl-c in Console
1224 - Implement script driven addition of File daemon to config files.
1225 - Think about how to make Bacula work better with File (non-tape) archives.
1226 - Write Unix emulator for Windows.
1227 - Put memory utilization in Status output of each daemon
1228 if full status requested or if some level of debug on.
1229 - Make database type selectable by .conf files i.e. at runtime
1230 - Set flag for uname -a. Add to Volume label.
1231 - Restore files modified after date
1232 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
1233 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
1234 - Timout a job or terminate if link goes down, or reopen link and query.
1235 - Concept of precious tapes (cannot be reused).
1236 - Make bcopy copy with a single tape drive.
1237 - Permit changing ownership during restore.
1240 > My suggestion: Add a feature on the systray menu-icon menu to request
1241 > an immediate backup now. This would be useful for laptop users who may
1242 > not be on the network when the regular scheduled backup is run.
1244 > My wife's suggestion: Add a setting to the win32 client to allow it to
1245 > shut down the machine after backup is complete (after, of course,
1246 > displaying a "System will shut down in one minute, click here to cancel"
1247 > warning dialog). This would be useful for sites that want user
1248 > woorkstations to be shut down overnight to save power.
1251 - Autolabel should be specified by DIR instead of SD.
1253 - Add media capacity
1254 - AutoScan (check checksum of tape)
1255 - Format command = "format /dev/nst0"
1259 - Seek resolution (usually corresponds to buffer size)
1260 - EODErrorCode=ENOSPC or code
1261 - Partial Read error code
1262 - Partial write error code
1263 - Nonformatted read error
1264 - Nonformatted write error
1265 - WriteProtected error
1269 - IgnoreCloseErrors=yes
1279 - FD sends unsaved file list to Director at end of job (see
1281 - File daemon should build list of files skipped, and then
1282 at end of save retry and report any errors.
1283 - Write a Storage daemon that uses pipes and
1284 standard Unix programs to write to the tape.
1286 - Need something that monitors the JCR queue and
1287 times out jobs by asking the deamons where they are.
1288 - Enhance Jmsg code to permit buffering and saving to disk.
1289 - device driver = "xxxx" for drives.
1290 - Verify from Volume
1291 - Ensure that /dev/null works
1292 - Need report class for messages. Perhaps
1293 report resource where report=group of messages
1294 - enhance scan_attrib and rename scan_jobtype, and
1295 fill in code for "since" option
1296 - Director needs a time after which the report status is sent
1297 anyway -- or better yet, a retry time for the job.
1298 - Don't reschedule a job if previous incarnation is still running.
1299 - Some way to automatically backup everything is needed????
1300 - Need a structure for pending actions:
1302 - termination status (part of buffered msgs?)
1304 Read, Write, Clean, Delete
1305 - Login to Bacula; Bacula users with different permissions:
1306 owner, group, user, quotas
1307 - Store info on each file system type (probably in the job header on tape.
1308 This could be the output of df; or perhaps some sort of /etc/mtab record.
1310 ========= ideas ===============
1311 From: "Jerry K. Schieffer" <jerry@skylinetechnology.com>
1312 To: <kern@sibbald.com>
1313 Subject: RE: [Bacula-users] future large programming jobs
1314 Date: Thu, 26 Feb 2004 11:34:54 -0600
1316 I noticed the subject thread and thought I would offer the following
1317 merely as sources of ideas, i.e. something to think about, not even as
1318 strong as a request. In my former life (before retiring) I often
1319 dealt with backups and storage management issues/products as a
1320 developer and as a consultant. I am currently migrating my personal
1321 network from amanda to bacula specifically because of the ability to
1322 cross media boundaries during storing backups.
1323 Are you familiar with the commercial product called ADSM (I think IBM
1324 now sells it under the Tivoli label)? It has a couple of interesting
1325 ideas that may apply to the following topics.
1327 1. Migration: Consider that when you need to restore a system, there
1328 may be pressure to hurry. If all the information for a single client
1329 can eventually end up on the same media (and in chronological order),
1330 the restore is facillitated by not having to search past information
1331 from other clients. ADSM has the concept of "client affinity" that
1332 may be associated with it's storage pools. It seems to me that this
1333 concept (as an optional feature) might fit in your architecture for
1336 ADSM also has the concept of defining one or more storage pools as
1337 "copy pools" (almost mirrors, but only in the sense of contents).
1338 These pools provide the ability to have duplicte data stored both
1339 onsite and offsite. The copy process can be scheduled to be handled
1340 by their storage manager during periods when there is no backup
1341 activity. Again, the migration process might be a place to consider
1342 implementing something like this.
1345 > It strikes me that it would be very nice to be able to do things
1347 > have the Job(s) backing up the machines run, and once they have all
1348 > completed, start a migration job to copy the data from disks Volumes
1350 > a tape library and then to offsite storage. Maybe this can already
1352 > done with some careful scheduling and Job prioritzation; the events
1353 > mechanism described below would probably make it very easy.
1355 This is the goal. In the first step (before events), you simply
1357 the Migration to tape later.
1359 2. Base jobs: In ADSM, each copy of each stored file is tracked in
1360 the database. Once a file (unique by path and metadata such as dates,
1361 size, ownership, etc.) is in a copy pool, no more copies are made. In
1362 other words, when you start ADSM, it begins like your concept of a
1363 base job. After that it is in the "incremental" mode. You can
1364 configure the number of "generations" of files to be retained, plus a
1365 retention date after which even old generations are purged. The
1366 database tracks the contents of media and projects the percentage of
1367 each volume that is valid. When the valid content of a volume drops
1368 below a configured percentage, the valid data are migrated to another
1369 volume and the old volume is marked as empty. Note, this requires
1370 ADSM to have an idea of the contents of a client, i.e. marking the
1371 database when an existing file was deleted, but this would solve your
1372 issue of restoring a client without restoring deleted files.
1374 This is pretty far from what bacula now does, but if you are going to
1375 rip things up for Base jobs,.....
1376 Also, the benefits of this are huge for very large shops, especially
1377 with media robots, but are a pain for shops with manual media
1381 > Base jobs sound pretty useful, but I'm not dying for them.
1383 Nobody is dying for them, but when you see what it does, you will die
1386 3. Restoring deleted files: Since I think my comments in (2) above
1387 have low probability of implementation, I'll also suggest that you
1388 could approach the issue of deleted files by a mechanism of having the
1389 fd report to the dir, a list of all files on the client for every
1390 backup job. The dir could note in the database entry for each file
1391 the date that the file was seen. Then if a restore as of date X takes
1392 place, only files that exist from before X until after X would be
1393 restored. Probably the major cost here is the extra date container in
1394 each row of the files table.
1396 Thanks for "listening". I hope some of this helps. If you want to
1397 contact me, please send me an email - I read some but not all of the
1398 mailing list traffic and might miss a reply there.
1400 Please accept my compliments for bacula. It is doing a great job for
1401 me!! I sympathize with you in the need to wrestle with excelence in
1402 execution vs. excelence in feature inclusion.
1407 ==============================
1410 - Design at hierarchial storage for Bacula. Migration and Clone.
1411 - Implement FSM (File System Modules).
1412 - Audit M_ error codes to ensure they are correct and consistent.
1413 - Add variable break characters to lex analyzer.
1414 Either a bit mask or a string of chars so that
1415 the caller can change the break characters.
1416 - Make a single T_BREAK to replace T_COMMA, etc.
1417 - Ensure that File daemon and Storage daemon can
1418 continue a save if the Director goes down (this
1419 is NOT currently the case). Must detect socket error,
1420 buffer messages for later.
1421 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
1422 - Add ability to backup to two Storage devices (two SD sessions) at
1423 the same time -- e.g. onsite, offsite.
1424 - Add the ability to consolidate old backup sets (basically do a restore
1425 to tape and appropriately update the catalog). Compress Volume sets.
1426 Might need to spool via file is only one drive is available.
1427 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
1428 someway to do so with every volume that has less than x% valid
1432 Migration: Move a backup from one Volume to another
1433 Clone: Copy a backup -- two Volumes
1435 Bacula Migration is based on Jobs (apparently Networker is file by file).
1437 Migration triggered by:
1441 Highwater mark (keep total size)
1446 ======================================================
1448 It is somewhat like a Full save becomes an incremental since
1449 the Base job (or jobs) plus other non-base files.
1451 - A Base backup is same as Full backup, just different type.
1452 - New BaseFiles table that contains:
1454 BaseJobId - Base JobId referenced for this FileId (needed ???)
1455 JobId - JobId currently running
1456 FileId - File not backed up, exists in Base Job
1457 FileIndex - FileIndex from Base Job.
1458 i.e. for each base file that exists but is not saved because
1459 it has not changed, the File daemon sends the JobId, BaseId,
1460 FileId, FileIndex back to the Director who creates the DB entry.
1461 - To initiate a Base save, the Director sends the FD
1462 the FileId, and full filename for each file in the Base.
1463 - When the FD finds a Base file, he requests the Director to
1464 send him the full File entry (stat packet plus MD5), or
1465 conversely, the FD sends it to the Director and the Director
1466 says yes or no. This can be quite rapid if the FileId is kept
1467 by the FD for each Base Filename.
1468 - It is probably better to have the comparison done by the FD
1469 despite the fact that the File entry must be sent across the
1471 - An alternative would be to send the FD the whole File entry
1472 from the start. The disadvantage is that it requires a lot of
1473 space. The advantage is that it requires less communications
1475 - The Job record must be updated to indicate that one or more
1477 - At end of Job, FD returns:
1478 1. Count of base files/bytes not written to tape (i.e. matches)
1479 2. Count of base file that were saved i.e. had changed.
1480 - No tape record would be written for a Base file that matches, in the
1481 same way that no tape record is written for Incremental jobs where
1482 the file is not saved because it is unchanged.
1483 - On a restore, all the Base file records must explicitly be
1484 found from the BaseFile tape. I.e. for each Full save that is marked
1485 to have one or more Base Jobs, search the BaseFile for all occurrences
1487 - An optimization might be to make the BaseFile have:
1493 This would avoid the need to explicitly fetch each File record for
1494 the Base job. The Base Job record will be fetched to get the
1495 VolSessionId and VolSessionTime.
1496 =========================================================
1499 ==========================================================
1501 For each Incremental job that is run, there may be files that
1502 were found but not saved because they were locked (this applies
1503 only to Windows). Such a system could send back to the Director
1504 a list of Unsaved files.
1506 - New UnSavedFiles table that contains:
1510 - Then in the next Incremental job, the list of Unsaved Files will be
1511 feed to the FD, who will ensure that they are explicitly chosen even
1512 if standard date/time check would not have selected them.
1513 =============================================================
1517 Multiple drive autochanger data: see Alan Brown
1518 mtx -f xxx unloadStorage Element 1 is Already Full(drive 0 was empty)
1519 Unloading Data Transfer Element into Storage Element 1...source Element
1520 Address 480 is Empty
1522 (drive 0 was empty and so was slot 1)
1523 > mtx -f xxx load 15 0
1524 no response, just returns to the command prompt when complete.
1525 > mtx -f xxx status Storage Changer /dev/changer:2 Drives, 60 Slots ( 2 Import/Export )
1526 Data Transfer Element 0:Full (Storage Element 15 Loaded):VolumeTag = HX001
1527 Data Transfer Element 1:Empty
1528 Storage Element 1:Empty
1529 Storage Element 2:Full :VolumeTag=HX002
1530 Storage Element 3:Full :VolumeTag=HX003
1531 Storage Element 4:Full :VolumeTag=HX004
1532 Storage Element 5:Full :VolumeTag=HX005
1533 Storage Element 6:Full :VolumeTag=HX006
1534 Storage Element 7:Full :VolumeTag=HX007
1535 Storage Element 8:Full :VolumeTag=HX008
1536 Storage Element 9:Full :VolumeTag=HX009
1537 Storage Element 10:Full :VolumeTag=HX010
1538 Storage Element 11:Empty
1539 Storage Element 12:Empty
1540 Storage Element 13:Empty
1541 Storage Element 14:Empty
1542 Storage Element 15:Empty
1543 Storage Element 16:Empty....
1544 Storage Element 28:Empty
1545 Storage Element 29:Full :VolumeTag=CLNU01L1
1546 Storage Element 30:Empty....
1547 Storage Element 57:Empty
1548 Storage Element 58:Full :VolumeTag=NEX261L2
1549 Storage Element 59 IMPORT/EXPORT:Empty
1550 Storage Element 60 IMPORT/EXPORT:Empty
1552 Unloading Data Transfer Element into Storage Element 15...done
1554 (just to verify it remembers where it came from, however it can be
1555 overrriden with mtx unload {slotnumber} to go to any storage slot.)
1557 There needs to be a table of drive # to devices somewhere - If there are
1558 multiple changers or drives there may not be a 1:1 correspondance between
1559 changer drive number and system device name - and depending on the way the
1560 drives are hooked up to scsi busses, they may not be linearly numbered
1561 from an offset point either.something like
1563 Autochanger drives = 2
1564 Autochanger drive 0 = /dev/nst1
1565 Autochanger drive 1 = /dev/nst2
1566 IMHO, it would be _safest_ to use explicit mtx unload commands at all
1567 times, not just for multidrive changers. For a 1 drive changer, that's
1573 MTX's manpage (1.2.15):
1574 unload [<slotnum>] [ <drivenum> ]
1575 Unloads media from drive <drivenum> into slot
1576 <slotnum>. If <drivenum> is omitted, defaults to
1577 drive 0 (as do all commands). If <slotnum> is
1578 omitted, defaults to the slot that the drive was
1579 loaded from. Note that there's currently no way
1580 to say 'unload drive 1's media to the slot it
1581 came from', other than to explicitly use that
1582 slot number as the destination.AB
1588 undef# camcontrol devlist
1589 <WANGTEK 51000 SCSI M74H 12B3> at scbus0 target 2 lun 0 (pass0,sa0)
1590 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 0 (pass1,sa1)
1591 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 1 (pass2)
1593 tapeinfo -f /dev/sg0 with a bad tape in drive 1:
1594 [kern@rufus mtx-1.2.17kes]$ ./tapeinfo -f /dev/sg0
1595 Product Type: Tape Drive
1597 Product ID: 'C5713A '
1599 Attached Changer: No
1600 TapeAlert[3]: Hard Error: Uncorrectable read/write error.
1601 TapeAlert[20]: Clean Now: The tape drive neads cleaning NOW.
1608 Medium Type: Not Loaded
1611 DataCompEnabled: yes
1612 DataCompCapable: yes
1613 DataDeCompEnabled: yes
1620 Handling removable disks
1622 From: Karl Cunningham <karlc@keckec.com>
1624 My backups are only to hard disk these days, in removable bays. This is my
1625 idea of how a backup to hard disk would work more smoothly. Some of these
1626 things Bacula does already, but I mention them for completeness. If others
1627 have better ways to do this, I'd like to hear about it.
1629 1. Accommodate several disks, rotated similar to how tapes are. Identified
1630 by partition volume ID or perhaps by the name of a subdirectory.
1631 2. Abort & notify the admin if the wrong disk is in the bay.
1632 3. Write backups to different subdirectories for each machine to be backed
1634 4. Volumes (files) get created as needed in the proper subdirectory, one
1636 5. When a disk is recycled, remove or zero all old backup files. This is
1637 important as the disk being recycled may be close to full. This may be
1638 better done manually since the backup files for many machines may be
1639 scattered in many subdirectories.
1644 - Make sure that all do_prompt() calls in Dir check for
1645 -1 (error) and -2 (cancel) returns.
1646 - Fix foreach_jcr() to have free_jcr() inside next().
1647 jcr=jcr_walk_start();
1648 for ( ; jcr; (jcr=jcr_walk_next(jcr)) )
1651 - A Volume taken from Scratch should take on the retention period
1653 - Correct doc for Maximum Changer Wait (and others) accepting only
1655 - Implement status that shows why a job is being held in reserve, or
1656 rather why none of the drives are suitable.
1657 - Implement a way to disable a drive (so you can use the second
1658 drive of an autochanger, and the first one will not be used or
1660 - Make sure Maximum Volumes is respected in Pools when adding
1661 Volumes (e.g. when pulling a Scratch volume).
1662 - Keep same dcr when switching device ...
1663 - Implement code that makes the Dir aware that a drive is an
1664 autochanger (so the user doesn't need to use the Autochanger = yes
1666 - Make catalog respect ACL.
1667 - Add recycle count to Media record.
1668 - Add initial write date to Media record.
1669 - Fix store_yesno to be store_bitmask.
1670 --- create_file.c.orig Fri Jul 8 12:13:05 2005
1671 +++ create_file.c Fri Jul 8 12:13:07 2005
1673 attr->ofname, be.strerror());
1676 + } else if(S_ISSOCK(attr->statp.st_mode)) {
1677 + Dmsg1(200, "Skipping socket: %s\n", attr->ofname);
1679 Dmsg1(200, "Restore node: %s\n", attr->ofname);
1680 if (mknod(attr->ofname, attr->statp.st_mode, attr->statp.st_rdev) != 0 && errno != EEXIST) {
1681 - Add true/false to conf same as yes/no
1682 - Reserve blocks other restore jobs when first cannot connect to SD.
1683 - Fix Maximum Changer Wait, Maximum Open Wait, Maximum Rewind Wait to
1684 accept time qualifiers.
1685 - Does ClientRunAfterJob fail the job on a bad return code?
1686 - Make hardlink code at line 240 of find_one.c use binary search.
1687 - Add ACL error messages in src/filed/acl.c.
1688 - Make authentication failures single threaded.
1689 - Make Dir and SD authentication errors single threaded.
1690 - Fix catreq.c digestbuf at line 411 in src/dird/catreq.c
1691 - Make base64.c (bin_to_base64) take a buffer length
1692 argument to avoid overruns.
1693 and verify that other buffers cannot overrun.
1694 - Implement VolumeState as discussed with Arno.
1695 - Add LocationId to update volume
1703 - Add Comment to Media record
1704 - Fix auth compatibility with 1.38
1705 - Update dbcheck to include Log table
1706 - Update llist to include new fields.
1707 - Make unmount unload autochanger. Make mount load slot.
1708 - Fix bscan to report the JobType when restoring a job.
1709 - Fix wx-console scanning problem with commas in names.
1710 - Add manpages to the list of directories for make install. Notify
1712 - Add bconsole option to use stdin/out instead of conio.
1713 - Fix ClientRunBefore/AfterJob compatibility.
1714 - Ensure that connection to daemon failure always indicates what
1715 daemon it was trying to connect to.
1716 - Freespace on DVD requested over and over even with no intervening
1718 - .update volume [enabled|disabled|*see below]
1719 > However, I could easily imagine an option to "update slots" that says
1720 > "enable=yes|no" that would automatically enable or disable all the Volumes
1721 > found in the autochanger. This will permit the user to optionally mark all
1722 > the Volumes in the magazine disabled prior to taking them offsite, and mark
1723 > them all enabled when bringing them back on site. Coupled with the options
1724 > to the slots keyword, you can apply the enable/disable to any or all volumes.
1725 - Restricted consoles start in the Default catalog even if it
1727 - When reading through parts on the DVD, the DVD is mounted and
1728 unmounted for each part.
1729 - Make sure that the restore options don't permit "seeing" other
1731 - Restore of a raw drive should not try to check the volume size.
1732 - Lock tape drive door when open()
1733 - Make release unload any autochanger.
1734 - Arno's reservation deadlock.