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.
26 - Correct the Include syntax in the m4.xxx files in examples/conf
27 - Document JobStatus and Termination codes.
28 - Fix the error with the "DVI file can't be opened" while
29 building the French PDF.
30 - Document more DVD stuff -- particularly that recycling doesn't work,
31 and all the other things too.
36 - When reading through parts on the DVD, the DVD is mounted and
37 unmounted for each part.
38 - Make sure that the restore options don't permit "seeing" other
39 Client's job data. -- mostly fixed.
40 - Implement Python event for backing up/restoring a file.
41 - Change dbcheck to tell users to use native tools for fixing
42 broken databases, and to ensure they have the proper indexes.
43 - add udev rules for Bacula devices.
44 - If a job terminates, the DIR connection can close before the
45 Volume info is updated, leaving the File count wrong.
46 - Look at why SIGPIPE during connection can cause seg fault in
47 writing the daemon message, when Dir dropped to bacula:bacula
48 - Look at zlib 32 => 64 problems.
49 - Try turning on disk seek code.
50 - Possibly turn on St. Bernard code.
51 - Fix bextract to restore ACLs, or better yet, use common
53 - Do we migrate appendable Volumes?
54 - Remove queue.c code.
55 - Some users claim that they must do two prune commands to get a
56 Volume marked as purged.
57 - Print warning message if LANG environment variable does not specify
59 - New dot commands from Arno.
60 .show device=xxx lists information from one storage device, including
61 devices (I'm not even sure that information exists in the DIR...)
62 .move eject device=xxx mostly the same as 'unmount xxx' but perhaps with
63 better machine-readable output like "Ok" or "Error busy"
64 .move eject device=xxx toslot=yyy the same as above, but with a new
65 target slot. The catalog should be updated accordingly.
66 .move transfer device=xxx fromslot=yyy toslot=zzz
69 - Check to see if jcr->stime is lost during rescheduling of
71 - Fix re-read of last block to check if job has actually written
72 a block, and check if block was written by a different job
73 (i.e. multiple simultaneous jobs writing).
74 - Figure out how to configure query.sql. Suggestion to use m4:
76 changequote(`[',`]')dnl
78 :List next 20 volumes to expire
80 Pool.Name AS PoolName,
85 [ FROM_UNIXTIME(UNIX_TIMESTAMP(Media.LastWritten) Media.VolRetention) AS Expire, ])dnl
87 [ media.lastwritten + interval '1 second' * media.volretention as expire, ])dnl
91 ON Media.PoolId=Pool.PoolId
92 WHERE Media.LastWritten>0
96 Command: m4 -DmySQL changequote.m4 query.sql.in >query.sql
98 The problem is that it requires m4, which is not present on all machines
100 - Get Perl replacement for bregex.c
101 - Given all the problems with FIFOs, I think the solution is to do something a
102 little different, though I will look at the code and see if there is not some
103 simple solution (i.e. some bug that was introduced). What might be a better
104 solution would be to use a FIFO as a sort of "key" to tell Bacula to read and
105 write data to a program rather than the FIFO. For example, suppose you
110 Then, I could imagine if you backup and restore this file with a direct
111 reference as is currently done for fifos, instead, during backup Bacula will
114 /home/kern/my-fifo.backup
116 and read the data that my-fifo.backup writes to stdout. For restore, Bacula
119 /home/kern/my-fifo.restore
121 and send the data backed up to stdout. These programs can either be an
122 executable or a shell script and they need only read/write to stdin/stdout.
124 I think this would give a lot of flexibility to the user without making any
125 significant changes to Bacula.
130 select FilenameId from Filename where Name='';
131 # Get list of all directories referenced in a Backup.
132 select Path.Path from Path,File where File.JobId=nnn and
133 File.FilenameId=(FilenameId-from-above) and File.PathId=Path.PathId
134 order by Path.Path ASC;
136 - Look into using Dart for testing
137 http://public.kitware.com/Dart/HTML/Index.shtml
139 - Look into replacing autotools with cmake
140 http://www.cmake.org/HTML/Index.html
142 === Migration from David ===
143 What I'd like to see:
146 Name = "<poolname>-migrate"
150 Migration Selection Type = LowestUtil | OldestVol | PoolOccupancy |
151 Client | PoolResidence | Volume | JobName | SQLquery
152 Migration Selection Pattern = "regexp"
153 Next Pool = <override>
156 There should be no need for a Level (migration is always Full, since you
157 don't calculate differential/incremental differences for migration),
158 Storage should be determined by the volume types in the pool, and Client
159 is really a selection issue. Migration should always occur to the
160 NextPool defined in the pool definition. If no nextpool is defined, the
161 job should end with a reason of "no place to go". If Next Pool statement
162 is present, we override the check in the pool definition and use the
165 Here's how I'd define Migration Selection Types:
168 Client -- Migrate data from selected client only. Migration Selection
169 Pattern regexp provides pattern to select client names, eg ^FS00* makes
170 all client names starting with FS00 eligible for migration.
172 Jobname -- Migration all jobs matching name. Migration Selection Pattern
173 regexp provides pattern to select jobnames existing in pool.
175 Volume -- Migrate all data on specified volumes. Migration Selection
176 Pattern regexp provides selection criteria for volumes to be migrated.
177 Volumes must exist in pool to be eligible for migration.
181 LowestUtil -- Identify the volume in the pool with the least data on it
182 and empty it. No Migration Selection Pattern required.
184 OldestVol -- Identify the LRU volume with data written, and empty it. No
185 Migration Selection Pattern required.
187 PoolOccupancy -- if pool occupancy exceeds <highmig>, migrate volumes
188 (starting with most full volumes) until pool occupancy drops below
189 <lowmig>. Pool highmig and lowmig values are in pool definition, no
190 Migration Selection Pattern required.
194 SQLQuery -- Migrate all jobuids returned by the supplied SQL query.
195 Migration Selection Pattern contains SQL query to execute; should return
196 a list of 1 or more jobuids to migrate.
198 PoolResidence -- Migrate data sitting in pool for longer than
199 PoolResidence value in pool definition. Migration Selection Pattern
200 optional; if specified, override value in pool definition (value in
204 [ possibly a Python event -- kes ]
206 - Mount on an Autochanger with no tape in the drive causes:
207 Automatically selected Storage: LTO-changer
208 Enter autochanger drive[0]: 0
209 3301 Issuing autochanger "loaded drive 0" command.
210 3302 Autochanger "loaded drive 0", result: nothing loaded.
211 3301 Issuing autochanger "loaded drive 0" command.
212 3302 Autochanger "loaded drive 0", result: nothing loaded.
213 3902 Cannot mount Volume on Storage Device "LTO-Drive1" (/dev/nst0) because:
214 Couldn't rewind device "LTO-Drive1" (/dev/nst0): ERR=dev.c:678 Rewind error on "LTO-Drive1" (/dev/nst0). ERR=No medium found.
215 3905 Device "LTO-Drive1" (/dev/nst0) open but no Bacula volume is mounted.
216 If this is not a blank tape, try unmounting and remounting the Volume.
217 - If Drive 0 is blocked, and drive 1 is set "Autoselect=no", drive 1 will
219 - Autochanger did not change volumes.
220 select * from Storage;
221 +-----------+-------------+-------------+
222 | StorageId | Name | AutoChanger |
223 +-----------+-------------+-------------+
224 | 1 | LTO-changer | 0 |
225 +-----------+-------------+-------------+
226 05-May 03:50 roxie-sd: 3302 Autochanger "loaded drive 0", result is Slot 11.
227 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Warning: Director wanted Volume "LT
228 Current Volume "LT0-002" not acceptable because:
229 1997 Volume "LT0-002" not in catalog.
230 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Error: Autochanger Volume "LT0-002"
231 Setting InChanger to zero in catalog.
232 05-May 03:50 roxie-dir: Tibs.2006-05-05_03.05.02 Error: Unable to get Media record
234 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: Error getting Volume i
235 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: Job 530 canceled.
236 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: spool.c:249 Fatal appe
237 05-May 03:49 Tibs: Tibs.2006-05-05_03.05.02 Fatal error: c:\cygwin\home\kern\bacula
246 FirstWritten: 2006-05-05 03:11:54
247 LastWritten: 2006-05-05 03:50:23
248 LabelDate: 2005-12-26 16:52:40
259 VolRetention: 31,536,000
271 Note VolStatus is blank!!!!!
278 FirstWritten: 0000-00-00 00:00:00
279 LastWritten: 0000-00-00 00:00:00
280 LabelDate: 2005-12-26 16:52:40
291 VolRetention: 31,536,000
304 Automatically selected Storage: LTO-changer
305 Enter autochanger drive[0]: 0
306 3301 Issuing autochanger "loaded drive 0" command.
307 3302 Autochanger "loaded drive 0", result: nothing loaded.
308 3301 Issuing autochanger "loaded drive 0" command.
309 3302 Autochanger "loaded drive 0", result: nothing loaded.
310 3902 Cannot mount Volume on Storage Device "LTO-Drive1" (/dev/nst0) because:
311 Couldn't rewind device "LTO-Drive1" (/dev/nst0): ERR=dev.c:678 Rewind error on "LTO-Drive1" (/dev/nst0). ERR=No medium found.
313 3905 Device "LTO-Drive1" (/dev/nst0) open but no Bacula volume is mounted.
314 If this is not a blank tape, try unmounting and remounting the Volume.
316 - Add VolumeState (enable, disable, archive)
317 - Add VolumeLock to prevent all but lock holder (SD) from updating
318 the Volume data (with the exception of VolumeState).
319 - The btape fill command does not seem to use the Autochanger
320 - Make Windows installer default to system disk drive.
321 - Look at using ioctl(FIOBMAP, ...) on Linux, and
322 DeviceIoControl(..., FSCTL_QUERY_ALLOCATED_RANGES, ...) on
323 Win32 for sparse files.
324 http://www.flexhex.com/docs/articles/sparse-files.phtml
325 http://www.informatik.uni-frankfurt.de/~loizides/reiserfs/fibmap.html
326 - Directive: at <event> "command"
327 - Command: pycmd "command" generates "command" event. How to
328 attach to a specific job?
329 - Integrate Christopher's St. Bernard code.
330 - run_cmd() returns int should return JobId_t
331 - get_next_jobid_from_list() returns int should return JobId_t
332 - Document export LDFLAGS=-L/usr/lib64
333 - Don't attempt to restore from "Disabled" Volumes.
334 - Network error on Win32 should set Win32 error code.
335 - What happens when you rename a Disk Volume?
336 - Job retention period in a Pool (and hence Volume). The job would
338 - Detect resource deadlock in Migrate when same job wants to read
339 and write the same device.
340 - Queue warning/error messages during restore so that they
341 are reported at the end of the report rather than being
342 hidden in the file listing ...
343 - Look at -D_FORTIFY_SOURCE=2
344 - Add Win32 FileSet definition somewhere
345 - Look at fixing restore status stats in SD.
346 - Make selection of Database used in restore correspond to
348 - Look at using ioctl(FIMAP) and FIGETBSZ for sparse files.
349 http://www.informatik.uni-frankfurt.de/~loizides/reiserfs/fibmap.html
350 - Implement a mode that says when a hard read error is
351 encountered, read many times (as it currently does), and if the
352 block cannot be read, skip to the next block, and try again. If
353 that fails, skip to the next file and try again, ...
355 create table LevelType (LevelType binary(1), LevelTypeLong tinyblob);
356 insert into LevelType (LevelType,LevelTypeLong) values
360 - Add ACL to restore only to original location.
361 - Show files/second in client status output.
362 - Add a recursive mark command (rmark) to restore.
363 - "Minimum Job Interval = nnn" sets minimum interval between Jobs
364 of the same level and does not permit multiple simultaneous
365 running of that Job (i.e. lets any previous invocation finish
366 before doing Interval testing).
367 - Look at simplifying File exclusions.
368 - New directive "Delete purged Volumes"
369 - new pool XXX with ScratchPoolId = MyScratchPool's PoolId and
370 let it fill itself, and RecyclePoolId = XXX's PoolId so I can
371 see if it become stable and I just have to supervise
373 - If I want to remove this pool, I set RecyclePoolId = MyScratchPool's
374 PoolId, and when it is empty remove it.
375 - Figure out how to recycle Scratch volumes back to the Scratch Pool.
377 - Allow Check Labels to be used with Bacula labels.
378 - "Resuming" a failed backup (lost line for example) by using the
379 failed backup as a sort of "base" job.
381 - Email to the user when the tape is about to need changing x
382 days before it needs changing.
383 - Command to show next tape that will be used for a job even
384 if the job is not scheduled.
385 - From: Arunav Mandal <amandal@trolltech.com>
386 1. When jobs are running and bacula for some reason crashes or if I do a
387 restart it remembers and jobs it was running before it crashed or restarted
388 as of now I loose all jobs if I restart it.
390 2. When spooling and in the midway if client is disconnected for instance a
391 laptop bacula completely discard the spool. It will be nice if it can write
392 that spool to tape so there will be some backups for that client if not all.
394 3. We have around 150 clients machines it will be nice to have a option to
395 upgrade all the client machines bacula version automatically.
397 4. Atleast one connection should be reserved for the bconsole so at heavy load
398 I should connect to the director via bconsole which at sometimes I can't
400 5. Another most important feature that is missing, say at 10am I manually
401 started backup of client abc and it was a full backup since client abc has
402 no backup history and at 10.30am bacula again automatically started backup of
403 client abc as that was in the schedule. So now we have 2 multiple Full
404 backups of the same client and if we again try to start a full backup of
405 client backup abc bacula won't complain. That should be fixed.
407 - Fix bpipe.c so that it does not modify results pointer.
408 ***FIXME*** calling sequence should be changed.
409 - For Windows disaster recovery see http://unattended.sf.net/
410 - regardless of the retention period, Bacula will not prune the
411 last Full, Diff, or Inc File data until a month after the
412 retention period for the last Full backup that was done.
413 - update volume=xxx --- add status=Full
414 - Remove old spool files on startup.
415 - Exclude SD spool/working directory.
416 - Refuse to prune last valid Full backup. Same goes for Catalog.
418 - Make a callback when Rerun failed levels is called.
419 - Give Python program access to Scheduled jobs.
420 - Add setting Volume State via Python.
421 - Python script to save with Python, not save, save with Bacula.
422 - Python script to do backup.
424 - Change the Priority, Client, Storage, JobStatus (error)
425 at the start of a job.
426 - Why is SpoolDirectory = /home/bacula/spool; not reported
427 as an error when writing a DVD?
428 - Make bootstrap file handle multiple MediaTypes (SD)
429 - Remove all old Device resource code in Dir and code to pass it
430 back in SD -- better, rework it to pass back device statistics.
431 - Check locking of resources -- be sure to lock devices where previously
432 resources were locked.
433 - The last part is left in the spool dir.
436 - In restore don't compare byte count on a raw device -- directory
437 entry does not contain bytes.
438 - To mark files as deleted, run essentially a Verify to disk, and
439 when a file is found missing (MarkId != JobId), then create
440 a new File record with FileIndex == -1. This could be done
441 by the FD at the same time as the backup.
445 MA = (last_MA * 3 + rate) / 4
446 rate = (bytes - last_bytes) / (runtime - last_runtime)
447 - Max Vols limit in Pool off by one?
448 - Implement Files/Bytes,... stats for restore job.
449 - Implement Total Bytes Written, ... for restore job.
450 - Despool attributes simultaneously with data in a separate
451 thread, rejoined at end of data spooling.
452 - 7. Implement new Console commands to allow offlining/reserving drives,
453 and possibly manipulating the autochanger (much asked for).
454 - Add start/end date editing in messages (%t %T, %e?) ...
455 - Add ClientDefs similar to JobDefs.
456 - Print more info when bextract -p accepts a bad block.
457 - Fix FD JobType to be set before RunBeforeJob in FD.
458 - Look at adding full Volume and Pool information to a Volume
459 label so that bscan can get *all* the info.
460 - If the user puts "Purge Oldest Volume = yes" or "Recycle Oldest Volume = yes"
461 and there is only one volume in the pool, refuse to do it -- otherwise
462 he fills the Volume, then immediately starts reusing it.
463 - Implement copies and stripes.
464 - Add history file to console.
465 - Each file on tape creates a JobMedia record. Peter has 4 million
466 files spread over 10000 tape files and four tapes. A restore takes
467 16 hours to build the restore list.
468 - Add and option to check if the file size changed during backup.
469 - Make sure SD deletes spool files on error exit.
470 - Delete old spool files when SD starts.
471 - When labeling tapes, if you enter 000026, Bacula uses
472 the tape index rather than the Volume name 000026.
473 - Add offline tape command to Bacula console.
475 Enter MediaId or Volume name: 32
476 Enter new Volume name: DLT-20Dec04
477 Automatically selected Pool: Default
478 Connecting to Storage daemon DLTDrive at 192.168.68.104:9103 ...
479 Sending relabel command from "DLT-28Jun03" to "DLT-20Dec04" ...
480 block.c:552 Write error at 0:0 on device /dev/nst0. ERR=Bad file descriptor.
481 Error writing final EOF to tape. This tape may not be readable.
482 dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
483 askdir.c:219 NULL Volume name. This shouldn't happen!!!
484 3912 Failed to label Volume: ERR=dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
485 Label command failed for Volume DLT-20Dec04.
486 Do not forget to mount the drive!!!
487 - Bug: if a job is manually scheduled to run later, it does not appear
488 in any status report and cannot be cancelled.
490 ==== Keeping track of deleted files ====
491 My "trick" for keeping track of deletions is the following.
492 Assuming the user turns on this option, after all the files
493 have been backed up, but before the job has terminated, the
494 FD will make a pass through all the files and send their
495 names to the DIR (*exactly* the same as what a Verify job
496 currently does). This will probably be done at the same
497 time the files are being sent to the SD avoiding a second
498 pass. The DIR will then compare that to what is stored in
499 the catalog. Any files in the catalog but not in what the
500 FD sent will receive a catalog File entry that indicates
501 that at that point in time the file was deleted.
503 During a restore, any file initially picked up by some
504 backup (Full, ...) then subsequently having a File entry
505 marked "delete" will be removed from the tree, so will not
506 be restored. If a file with the same name is later OK it
507 will be inserted in the tree -- this already happens. All
508 will be consistent except for possible changes during the
511 Since I'm on the subject, some of you may be wondering what
512 the utility of the in memory tree is if you are going to
513 restore everything (at least it comes up from time to time
514 on the list). Well, it is still *very* useful because it
515 allows only the last item found for a particular filename
516 (full path) to be entered into the tree, and thus if a file
517 is backed up 10 times, only the last copy will be restored.
518 I recently (last Friday) restored a complete directory, and
519 the Full and all the Differential and Incremental backups
520 spanned 3 Volumes. The first Volume was not even mounted
521 because all the files had been updated and hence backed up
522 since the Full backup was made. In this case, the tree
523 saved me a *lot* of time.
525 Make sure this information is stored on the tape too so
526 that it can be restored directly from the tape.
528 Comments from Martin Simmons (I think they are all covered):
529 Ok, that should cover the basics. There are few issues though:
531 - Restore will depend on the catalog. I think it is better to include the
532 extra data in the backup as well, so it can be seen by bscan and bextract.
534 - I'm not sure if it will preserve multiple hard links to the same inode. Or
535 maybe adding or removing links will cause the data to be dumped again?
537 - I'm not sure if it will handle renamed directories. Possibly it will work
538 by dumping the whole tree under a renamed directory?
540 - It remains to be seen how the backup performance of the DIR's will be
541 affected when comparing the catalog for a large filesystem.
545 How about introducing a Type = MgmtPolicy job type? That job type would
546 be responsible for scanning the Bacula environment looking for specific
547 conditions, and submitting the appropriate jobs for implementing said
551 Name = "Migration-Policy"
553 Policy Selection Job Type = Migrate
554 Scope = "<keyword> <operator> <regexp>"
555 Threshold = "<keyword> <operator> <regexp>"
556 Job Template = <template-name>
559 Where <keyword> is any legal job keyword, <operator> is a comparison
560 operator (=,<,>,!=, logical operators AND/OR/NOT) and <regexp> is a
561 appropriate regexp. I could see an argument for Scope and Threshold
562 being SQL queries if we want to support full flexibility. The
563 Migration-Policy job would then get scheduled as frequently as a site
564 felt necessary (suggested default: every 15 minutes).
569 Name = "Migration-Policy"
571 Policy Selection Job Type = Migration
573 Threshold = "Migration Selection Type = LowestUtil"
574 Job Template = "MigrationTemplate"
577 would select all pools for examination and generate a job based on
578 MigrationTemplate to automatically select the volume with the lowest
579 usage and migrate it's contents to the nextpool defined for that pool.
581 This policy abstraction would be really handy for adjusting the behavior
582 of Bacula according to site-selectable criteria (one thing that pops
583 into mind is Amanda's ability to automatically adjust backup levels
584 depending on various criteria).
590 - Add Pool/Storage override regression test.
591 - Add delete JobId to regression.
592 - Add a regression test for dbcheck.
593 - New test to add bscan to four-concurrent-jobs regression,
594 i.e. after the four-concurrent jobs zap the
595 database as is done in the bscan-test, then use bscan to
596 restore the database, do a restore and compare with the
598 - Add restore of specific JobId to regression (item 3
599 on the restore prompt)
600 - Add IPv6 to regression
601 - Add database test to regression. Test each function like delete,
604 - AntiVir can slow down backups on Win32 systems.
605 - Win32 systems with FAT32 can be much slower than NTFS for
606 more than 1000 files per directory.
610 - A HOLD command to stop all jobs from starting.
611 - A PAUSE command to pause all running jobs ==> release the
613 - Media Type = LTO,LTO-2,LTO-3
614 Media Type Read = LTO,LTO2,LTO3
615 Media Type Write = LTO2, LTO3
617 === From Carsten Menke <bootsy52@gmx.net>
619 Following is a list of what I think in the situations where I'm faced with,
620 could be a usefull enhancement to bacula, which I'm certain other users will
621 benefit from as well.
623 1. NextJob/NextJobs Directive within a Job Resource in the form of
624 NextJobs = job1,job2.
627 I currently solved the problem with running multiple jobs each after each
628 by setting the Max Wait Time for a job to 8 hours, and give
629 the jobs different Priorities. However, there scenarios where
630 1 Job is directly depending on another job, so if the former job fails,
631 the job after it needn't to be run
632 while maybe other jobs should run despite of that
635 A Backup Job and a Verify job, if the backup job fails there is no need to run
636 the verify job, as the backup job already failed. However, one may like
637 to backup the Catalog to disk despite of that the main backup job failed.
640 I see that this is related to the Event Handlers which are on the ToDo
641 list, also it is maybe a good idea to check for the return value and
642 execute different actions based on the return value
645 3. offline capability to bconsole
648 Currently I use a script which I execute within the last Job via the
649 RunAfterJob Directive, to release and eject the tape.
650 So I have to call bconsole "release=Storage-Name" and afterwards
651 mt -f /dev/nst0 eject to get the tape out.
653 If I have multiple Storage Devices, than these may not be /dev/nst0 and
654 I have to modify the script or call it with parameters etc.
655 This would actually not be needed, as everything is already defined
656 in bacula-sd.conf and if I can invoke bconsole with the
657 storage name via $1 in the script than I'm done and information is
660 4. %s for Storage Name added to the chars being substituted in "RunAfterJob"
664 For the reason mentioned in 3. to have the ability to call a
665 script with /scripts/foobar %s and in the script use $1
666 to pass the Storage Name to bconsole
668 5. Setting Volume State within a Job Resource
671 Instead of using "Maximum Volume Jobs" in the Pool Resource,
672 I would have the possibilty to define
673 in a Job Resource that after this certain job is run, the Volume State
674 should be set to "Volume State = Used", this give more flexibility (IMHO).
676 6. Localization of Bacula Messages
679 Unfortunatley many,many people I work with don't speak english very well.
680 So if at least the Reporting messages would be localized then they
681 would understand that they have to change the tape,etc. etc.
683 I volunteer to do the german translations, and if I can convince my wife also
684 french and Morre (western african language).
686 7. OK, this is evil, probably bound to security risks and maybe not possible
687 due to the design of bacula.
689 Implementation of Backtics ( `command` ) for shell comand execution to
690 the "Label Format" Directive.
694 Currently I have defined BACULA_DAY_OF_WEEK="day1|day2..." resulting in
695 Label Format = "HolyBackup-${BACULA_DAY_OF_WEEK[${WeekDay}]}". If I could
696 use backticks than I could use "Label Format = HolyBackup-`date +%A` to have
697 the localized name for the day of the week appended to the
698 format string. Then I have the tape labeled automatically with weekday
699 name in the correct language.
701 - Yes, that is surely the case. I probably should turn those into Warning
702 errors. In addition, you just made me think that it might not be bad to
703 add an option to check the file size after backing up the file and
704 report if it changes. This would be done as an option because it would
707 Kern, good idea. If you do do that, mention in the output: file
708 shrunk, or file expanded, just to make it obvious to the user
709 (without having to the refer to file size), just how the file size
712 Would this option be for all file, or just one file? Or a fileset?
713 - Make output from status use html table tags for nicely
714 presenting in a browser.
715 - Can one write tapes faster with 8192 byte block sizes?
716 - Document security problems with the same password for everyone in
717 rpm and Win32 releases.
718 - Browse generations of files.
719 - I've seen an error when my catalog's File table fills up. I
720 then have to recreate the File table with a larger maximum row
721 size. Relevant information is at
722 http://dev.mysql.com/doc/mysql/en/Full_table.html ; I think the
723 "Installing and Configuring MySQL" chapter should talk a bit
724 about this potential problem, and recommend a solution.
725 - For Solaris must use POSIX awk.
726 - Want speed of writing to tape while despooling.
727 - Supported autochanger:
735 Wangtek 6525ES (SCSI-1 QIC drive, 525MB), under Linux 2.4.something,
736 bacula 1.36.0/1 works with blocksize 16k INSIDE bacula-sd.conf.
737 - Add regex from http://www.pcre.org to Bacula for Win32.
738 - Use only shell tools no make in CDROM package.
739 - Include within include does it work?
740 - Implement a Pool of type Cleaning?
741 - Implement VolReadTime and VolWriteTime in SD
742 - Modify Backing up Your Database to include a bootstrap file.
743 - Think about making certain database errors fatal.
744 - Look at correcting the time jump in the scheduler for daylight
745 savings time changes.
746 - Add a "real" timer to network connections.
747 - Promote to Full = Time period
748 - Check dates entered by user for correctness (month/day/... ranges)
749 - Compress restore Volume listing by date and first file.
750 - Look at patches/bacula_db.b2z postgresql that loops during restore.
752 - Perhaps add read/write programs and/or plugins to FileSets.
753 - How to handle backing up portables ...
754 - Add some sort of guaranteed Interval for upgrading jobs.
755 - Can we write the state file after every job terminates? On Win32
756 the system crashes and the state file is not updated.
759 Documentation to do: (any release a little bit at a time)
760 - Doc to do unmount before removing magazine.
761 - Alternative to static linking "ldd prog" save all binaries listed,
762 restore them and point LD_LIBRARY_PATH to them.
763 - Document add "</dev/null >/dev/null 2>&1" to the bacula-fd command line
764 - Document query file format.
765 - Add more documentation for bsr files.
766 - Document problems with Verify and pruning.
767 - Document how to use multiple databases.
768 - VXA drives have a "cleaning required"
769 indicator, but Exabyte recommends preventive cleaning after every 75
772 In this context, it should be noted that Exabyte has a command-line
773 vxatool utility available for free download. (The current version is
774 vxatool-3.72.) It can get diagnostic info, read, write and erase tapes,
775 test the drive, unload tapes, change drive settings, flash new firmware,
777 Of particular interest in this context is that vxatool <device> -i will
778 report, among other details, the time since last cleaning in tape motion
779 minutes. This information can be retrieved (and settings changed, for
780 that matter) through the generic-SCSI device even when Bacula has the
781 regular tape device locked. (Needless to say, I don't recommend
782 changing tape settings while a job is running.)
783 - Lookup HP cleaning recommendations.
784 - Lookup HP tape replacement recommendations (see trouble shooting autochanger)
785 - Document doing table repair
788 ===================================
789 - Add macro expansions in JobDefs.
790 Run Before Job = "SomeFile %{Level} %{Client}"
791 Write Bootstrap="/some/dir/%{JobName}_%{Client}.bsr"
792 - Use non-blocking network I/O but if no data is available, use
794 - Use gather write() for network I/O.
795 - Autorestart on crash.
796 - Add bandwidth limiting.
797 - Add acks every once and a while from the SD to keep
798 the line from timing out.
799 - When an error in input occurs and conio beeps, you can back
800 up through the prompt.
801 - Detect fixed tape block mode during positioning by looking at
802 block numbers in btape "test". Possibly adjust in Bacula.
803 - Fix list volumes to output volume retention in some other
804 units, perhaps via a directive.
805 - If opening a tape in read/write mode fails attempt to open
806 it in read-only mode, and mark the tape for read only.
807 - Allow Simultaneous Priorities = yes => run up to Max concurrent jobs even
808 with multiple priorities.
809 - If you use restore replace=never, the directory attributes for
810 non-existent directories will not be restored properly.
812 - see lzma401.zip in others directory for new compression
814 - Minimal autochanger handling in Bacula and in btape.
815 - Look into how tar does not save sockets and the possiblity of
816 not saving them in Bacula (Martin Simmons reported this).
817 - Fix restore jobs so that multiple jobs can run if they
818 are not using the same tape(s).
819 - Allow the user to select JobType for manual pruning/purging.
820 - bscan does not put first of two volumes back with all info in
822 - Implement the FreeBSD nodump flag in chflags.
823 - Figure out how to make named console messages go only to that
824 console and to the non-restricted console (new console class?).
825 - Make restricted console prompt for password if *ask* is set or
826 perhaps if password is undefined.
827 - Implement "from ISO-date/time every x hours/days/weeks/months" in
830 ==== from Marc Schoechlin
831 - the help-command should be more verbose
832 (it should explain the paramters of the different
834 -> its time-comsuming to consult the manual anytime
835 you need a special parameter
836 -> maybe its more easy to maintain this, if the
837 descriptions of that commands are outsourced to
839 - the cd-command should allow complete paths
840 i.e. cd /foo/bar/foo/bar
841 -> if a customer mails me the path to a certain file,
842 its faster to enter the specified directory
843 - if the password is not configured in bconsole.conf
844 you should be asked for it.
845 -> sometimes you like to do restore on a customer-machine
846 which shouldnt know the password for bacula.
847 -> adding the password to the file favours admins
848 to forget to remove the password after usage
850 the protection of that file is less important
851 - long-listed-output of commands should be scrollable
852 like the unix more/less-command does
853 -> if someone runs 200 and more machines, the lists could
854 be a little long and complex
855 - command-output should be shown column by column
856 to reduce scrolling and to increase clarity
858 - lsmark should list the selected files with full
860 - wildcards for selecting and file and directories would be nice
861 - any actions should be interuptable with STRG+C
862 - command-expansion would be pretty cool
864 - When the replace Never option is set, new directory permissions
865 are not restored. See bug 213. To fix this requires creating a
866 list of newly restored directories so that those directory
867 permissions *can* be restored.
868 - Compaction of Disk space by "migrating" Volumes that have pruned
869 Jobs (what criteria? size, #jobs, time).
870 - Add prune all command
871 - Document fact that purge can destroy a part of a restore by purging
872 one volume while others remain valid -- perhaps mark Jobs.
873 - Add multiple-media-types.txt
874 - look at mxt-changer.html
875 - Make ? do a help command (no return needed).
876 - Implement restore directory.
877 - Document streams and how to implement them.
878 - Try not to re-backup a file if a new hard link is added.
879 - Add feature to backup hard links only, but not the data.
880 - Fix stream handling to be simpler.
881 - Add Priority and Bootstrap to Run a Job.
882 - Eliminate Restore "Run Restore Job" prompt by allowing new "run command
884 - Remove View FileSet button from Run a Job dialog.
885 - Handle prompt for restore job at end of Restore command.
886 - Add display of total selected files to Restore window.
887 - Add tree pane to left of window.
888 - Add progress meter.
889 - Max wait time or max run time causes seg fault -- see runtime-bug.txt
890 - Document writing to a CD/DVD with Bacula.
891 - Add a "base" package to the window installer for pthreadsVCE.dll
892 which is needed by all packages.
893 - Add message to user to check for fixed block size when the forward
894 space test fails in btape.
895 - When unmarking a directory check if all files below are unmarked and
896 then remove the + flag -- in the restore tree.
897 - Possibly implement: Action = Unmount Device="TapeDrive1" in Admin jobs.
898 - Setup lrrd graphs: (http://www.linpro.no/projects/lrrd/) Mike Acar.
899 - Revisit the question of multiple Volumes (disk) on a single device.
900 - Add a block copy option to bcopy.
901 - Investigate adding Mac Resource Forks.
902 - Finish work on Gnome restore GUI.
903 - Fix "llist jobid=xx" where no fileset or client exists.
904 - For each job type (Admin, Restore, ...) require only the really necessary
905 fields.- Pass Director resource name as an option to the Console.
906 - Add a "batch" mode to the Console (no unsolicited queries, ...).
907 - Add a .list all files in the restore tree (probably also a list all files)
908 Do both a long and short form.
909 - Allow browsing the catalog to see all versions of a file (with
910 stat data on each file).
911 - Restore attributes of directory if replace=never set but directory
913 - Use SHA1 on authentication if possible.
914 - See comtest-xxx.zip for Windows code to talk to USB.
915 - Add John's appended files:
916 Appended = { /files/server/logs/http/*log }
917 and such files would be treated as follows.On a FULL backup, they would
918 be backed up like any other file.On an INCREMENTAL backup, where a
919 previous INCREMENTAL or FULL was already in thecatalogue and the length
920 of the file wasgreater than the length of the last backup, only thedata
921 added since the last backup will be dumped.On an INCREMENTAL backup, if
922 the length of the file is less than thelength of the file with the same
923 name last backed up, the completefile is dumped.On Windows systems, with
924 creation date of files, we can be evensmarter about this and not count
925 entirely upon the length.On a restore, the full and all incrementals
926 since it will beapplied in sequence to restore the file.
927 - Check new HAVE_WIN32 open bits.
928 - Check if the tape has moved before writing.
929 - Handling removable disks -- see below:
930 - Keep track of tape use time, and report when cleaning is necessary.
931 - Add FromClient and ToClient keywords on restore command (or
932 BackupClient RestoreClient).
933 - Implement a JobSet, which groups any number of jobs. If the
934 JobSet is started, all the jobs are started together.
935 Allow Pool, Level, and Schedule overrides.
936 - Enhance cancel to timeout BSOCK packets after a specific delay.
937 - Do scheduling by UTC using gmtime_r() in run_conf, scheduler, and
938 ua_status.!!! Thanks to Alan Brown for this tip.
939 - Look at updating Volume Jobs so that Max Volume Jobs = 1 will work
940 correctly for multiple simultaneous jobs.
941 - Correct code so that FileSet MD5 is calculated for < and | filename
943 - Implement the Media record flag that indicates that the Volume does disk
945 - Implement VolAddr, which is used when Volume is addressed like a disk,
946 and form it from VolFile and VolBlock.
947 - Make multiple restore jobs for multiple media types specifying
948 the proper storage type.
949 - Fix fast block rejection (stored/read_record.c:118). It passes a null
950 pointer (rec) to try_repositioning().
951 - Look at extracting Win data from BackupRead.
952 - Implement RestoreJobRetention? Maybe better "JobRetention" in a Job,
953 which would take precidence over the Catalog "JobRetention".
954 - Implement Label Format in Add and Label console commands.
955 - Possibly up network buffers to 65K. Put on variable.
956 - Put email tape request delays on one or more variables. User wants
957 to cancel the job after a certain time interval. Maximum Mount Wait?
958 - Job, Client, Device, Pool, or Volume?
959 Is it possible to make this a directive which is *optional* in multiple
960 resources, like Level? If so, I think I'd make it an optional directive
961 in Job, Client, and Pool, with precedence such that Job overrides Client
962 which in turn overrides Pool.
964 - New Storage specifications:
965 - Want to write to multiple storage devices simultaneously
966 - Want to write to multiple storage devices sequentially (in one job)
967 - Want to read/write simultaneously
968 - Key is MediaType -- it must match
970 Passed to SD as a sort of BSR record called Storage Specification
974 MediaType -> Next MediaType
976 Device -> Next Device
978 Allow multiple Storage specifications
986 Allow Multiple Pool specifications (note, Pool currently
988 Allow Multiple MediaType specifications in Dir conf
989 Allow Multiple Device specifications in Dir conf
990 Perhaps keep this in a single SSR
991 Tie a Volume to a specific device by using a MediaType that
992 is contained in only one device.
993 In SD allow Device to have Multiple MediaTypes
995 - Ideas from Jerry Scharf:
996 First let's point out some big pluses that bacula has for this
998 more importantly it's active. Thank you so much for that
999 even more important, it's not flaky
1000 it has an open access catalog, opening many possibilities
1001 it's pushing toward heterogeneous systems capability
1003 Macintosh file client
1004 macs are an interesting niche, but I fear a server is a rathole
1005 working bare iron recovery for windows
1006 the option for inc/diff backups not reset on fileset revision
1007 a) use both change and inode update time against base time
1008 b) do the full catalog check (expensive but accurate)
1009 sizing guide (how much system is needed to back up N systems/files)
1010 consultants on using bacula in building a disaster recovery system
1011 an integration guide
1012 or how to get at fancy things that one could do with bacula
1013 logwatch code for bacula logs (or similar)
1014 linux distro inclusion of bacula (brings good and bad, but necessary)
1015 win2k/XP server capability (icky but you asked)
1016 support for Oracle database ??
1018 - Look at adding SQL server and Exchange support for Windows.
1019 - Make dev->file and dev->block_num signed integers so that -1 can
1020 be an invalid value which happens with BSR.
1021 - Create VolAddr for disk files in place of VolFile and VolBlock. This
1022 is needed to properly specify ranges.
1023 - Add progress of files/bytes to SD and FD.
1024 - Print warning message if FileId > 4 billion
1025 - do a "messages" before the first prompt in Console
1026 - Client does not show busy during Estimate command.
1027 - Implement Console mtx commands.
1028 - Implement a Mount Command and an Unmount Command where
1029 the users could specify a system command to be performed
1030 to do the mount, after which Bacula could attempt to
1031 read the device. This is for Removeable media such as a CDROM.
1032 - Most likely, this mount command would be invoked explicitly
1033 by the user using the current Console "mount" and "unmount"
1034 commands -- the Storage Daemon would do the right thing
1035 depending on the exact nature of the device.
1036 - As with tape drives, when Bacula wanted a new removable
1037 disk mounted, it would unmount the old one, and send a message
1038 to the user, who would then use "mount" as described above
1039 once he had actually inserted the disk.
1040 - Implement dump/print label to UA
1041 - Spool to disk only when the tape is full, then when a tape is hung move
1043 - bextract is sending everything to the log file ****FIXME****
1044 - Allow multiple Storage specifications (or multiple names on
1045 a single Storage specification) in the Job record. Thus a job
1046 can be backed up to a number of storage devices.
1047 - Implement some way for the File daemon to contact the Director
1048 to start a job or pass its DHCP obtained IP number.
1049 - Implement a query tape prompt/replace feature for a console
1050 - Copy console @ code to gnome2-console
1051 - Make AES the only encryption algorithm see
1052 http://csrc.nist.gov/CryptoToolkit/aes/). It's
1053 an officially adopted standard, has survived peer
1054 review, and provides keys up to 256 bits.
1055 - Take a careful look at SetACL http://setacl.sourceforge.net
1056 - Make tree walk routines like cd, ls, ... more user friendly
1057 by handling spaces better.
1058 - Make sure that Bacula rechecks the tape after the 20 min wait.
1059 - Set IO_NOWAIT on Bacula TCP/IP packets.
1060 - Try doing a raw partition backup and restore by mounting a
1062 - From Lars Kellers:
1063 Yes, it would allow to highly automatic the request for new tapes. If a
1064 tape is empty, bacula reads the barcodes (native or simulated), and if
1065 an unused tape is found, it runs the label command with all the
1066 necessary parameters.
1068 By the way can bacula automatically "move" an empty/purged volume say
1069 in the "short" pool to the "long" pool if this pool runs out of volume
1071 - What to do about "list files job=xxx".
1072 - Get and test MySQL 4.0
1073 - Look at how fuser works and /proc/PID/fd that is how Nic found the
1074 file descriptor leak in Bacula.
1075 - Implement WrapCounters in Counters.
1076 - Add heartbeat from FD to SD if hb interval expires.
1077 - Can we dynamically change FileSets?
1078 - If pool specified to label command and Label Format is specified,
1079 automatically generate the Volume name.
1080 - Why can't SQL do the filename sort for restore?
1081 - Add ExhautiveRestoreSearch
1082 - Look at the possibility of loading only the necessary
1083 data into the restore tree (i.e. do it one directory at a
1084 time as the user walks through the tree).
1085 - Possibly use the hash code if the user selects all for a restore command.
1086 - Fix "restore all" to bypass building the tree.
1087 - Prohibit backing up archive device (findlib/find_one.c:128)
1088 - Implement Release Device in the Job resource to unmount a drive.
1089 - Implement Acquire Device in the Job resource to mount a drive,
1090 be sure this works with admin jobs so that the user can get
1091 prompted to insert the correct tape. Possibly some way to say to
1092 run the job but don't save the files.
1093 - Make things like list where a file is saved case independent for
1096 - Use autochanger to handle multiple devices.
1097 - On Windows with very long path names, it may be impossible to create
1098 a file (and thus restore it) because the total length is too long.
1099 We must cd into the directory then create the file without the
1101 - Implement a Recycle command
1102 - Test a second language e.g. french.
1103 - Start working on Base jobs.
1104 - Implement UnsavedFiles DB record.
1105 - From Phil Stracchino:
1106 It would probably be a per-client option, and would be called
1107 something like, say, "Automatically purge obsoleted jobs". What it
1108 would do is, when you successfully complete a Differential backup of a
1109 client, it would automatically purge all Incremental backups for that
1110 client that are rendered redundant by that Differential. Likewise,
1111 when a Full backup on a client completed, it would automatically purge
1112 all Differential and Incremental jobs obsoleted by that Full backup.
1113 This would let people minimize the number of tapes they're keeping on
1114 hand without having to master the art of retention times.
1115 - When doing a Backup send all attributes back to the Director, who
1116 would then figure out what files have been deleted.
1117 - Currently in mount.c:236 the SD simply creates a Volume. It should have
1118 explicit permission to do so. It should also mark the tape in error
1119 if there is an error.
1120 - Cancel waiting for Client connect in SD if FD goes away.
1122 - Implement timeout in response() when it should come quickly.
1123 - Implement a Slot priority (loaded/not loaded).
1124 - Implement "vacation" Incremental only saves.
1125 - Implement create "FileSet"?
1126 - Add prefixlinks to where or not where absolute links to FD.
1127 - Issue message to mount a new tape before the rewind.
1128 - Simplified client job initiation for portables.
1129 - If SD cannot open a drive, make it periodically retry.
1130 - Add more of the config info to the tape label.
1132 - If tape is marked read-only, then try opening it read-only rather than
1133 failing, and remember that it cannot be written.
1134 - Refine SD waiting output:
1135 Device is being positioned
1136 > Device is being positioned for append
1137 > Device is being positioned to file x
1139 - Figure out some way to estimate output size and to avoid splitting
1140 a backup across two Volumes -- this could be useful for writing CDROMs
1141 where you really prefer not to have it split -- not serious.
1142 - Have SD compute MD5 or SHA1 and compare to what FD computes.
1143 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
1144 actual data on the Volume and compare it.
1145 - Implement Bacula plugins -- design API
1146 - Make bcopy read through bad tape records.
1147 - Program files (i.e. execute a program to read/write files).
1148 Pass read date of last backup, size of file last time.
1149 - Add Signature type to File DB record.
1150 - CD into subdirectory when open()ing files for backup to
1151 speed up things. Test with testfind().
1152 - Priority job to go to top of list.
1153 - Why are save/restore of device different sizes (sparse?) Yup! Fix it.
1154 - Implement some way for the Console to dynamically create a job.
1155 - Solaris -I on tar for include list
1156 - Need a verbose mode in restore, perhaps to bsr.
1157 - bscan without -v is too quiet -- perhaps show jobs.
1158 - Add code to reject whole blocks if not wanted on restore.
1159 - Check if we can increase Bacula FD priorty in Win2000
1160 - Make sure the MaxVolFiles is fully implemented in SD
1161 - Check if both CatalogFiles and UseCatalog are set to SD.
1162 - Possibly add email to Watchdog if drive is unmounted too
1163 long and a job is waiting on the drive.
1164 - After unmount, if restore job started, ask to mount.
1165 - Add UA rc and history files.
1166 - put termcap (used by console) in ./configure and
1167 allow -with-termcap-dir.
1168 - Fix Autoprune for Volumes to respect need for full save.
1169 - Compare tape to Client files (attributes, or attributes and data)
1170 - Make all database Ids 64 bit.
1171 - Allow console commands to detach or run in background.
1172 - Fix status delay on storage daemon during rewind.
1173 - Add SD message variables to control operator wait time
1174 - Maximum Operator Wait
1175 - Minimum Message Interval
1176 - Maximum Message Interval
1177 - Send Operator message when cannot read tape label.
1178 - Verify level=Volume (scan only), level=Data (compare of data to file).
1179 Verify level=Catalog, level=InitCatalog
1181 - Add keyword search to show command in Console.
1182 - Events : tape has more than xxx bytes.
1183 - Complete code in Bacula Resources -- this will permit
1184 reading a new config file at any time.
1185 - Handle ctl-c in Console
1186 - Implement script driven addition of File daemon to config files.
1187 - Think about how to make Bacula work better with File (non-tape) archives.
1188 - Write Unix emulator for Windows.
1189 - Put memory utilization in Status output of each daemon
1190 if full status requested or if some level of debug on.
1191 - Make database type selectable by .conf files i.e. at runtime
1192 - Set flag for uname -a. Add to Volume label.
1193 - Restore files modified after date
1194 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
1195 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
1196 - Timout a job or terminate if link goes down, or reopen link and query.
1197 - Concept of precious tapes (cannot be reused).
1198 - Make bcopy copy with a single tape drive.
1199 - Permit changing ownership during restore.
1202 > My suggestion: Add a feature on the systray menu-icon menu to request
1203 > an immediate backup now. This would be useful for laptop users who may
1204 > not be on the network when the regular scheduled backup is run.
1206 > My wife's suggestion: Add a setting to the win32 client to allow it to
1207 > shut down the machine after backup is complete (after, of course,
1208 > displaying a "System will shut down in one minute, click here to cancel"
1209 > warning dialog). This would be useful for sites that want user
1210 > woorkstations to be shut down overnight to save power.
1213 - Autolabel should be specified by DIR instead of SD.
1215 - Add media capacity
1216 - AutoScan (check checksum of tape)
1217 - Format command = "format /dev/nst0"
1221 - Seek resolution (usually corresponds to buffer size)
1222 - EODErrorCode=ENOSPC or code
1223 - Partial Read error code
1224 - Partial write error code
1225 - Nonformatted read error
1226 - Nonformatted write error
1227 - WriteProtected error
1231 - IgnoreCloseErrors=yes
1241 - FD sends unsaved file list to Director at end of job (see
1243 - File daemon should build list of files skipped, and then
1244 at end of save retry and report any errors.
1245 - Write a Storage daemon that uses pipes and
1246 standard Unix programs to write to the tape.
1248 - Need something that monitors the JCR queue and
1249 times out jobs by asking the deamons where they are.
1250 - Enhance Jmsg code to permit buffering and saving to disk.
1251 - device driver = "xxxx" for drives.
1252 - Verify from Volume
1253 - Ensure that /dev/null works
1254 - Need report class for messages. Perhaps
1255 report resource where report=group of messages
1256 - enhance scan_attrib and rename scan_jobtype, and
1257 fill in code for "since" option
1258 - Director needs a time after which the report status is sent
1259 anyway -- or better yet, a retry time for the job.
1260 - Don't reschedule a job if previous incarnation is still running.
1261 - Some way to automatically backup everything is needed????
1262 - Need a structure for pending actions:
1264 - termination status (part of buffered msgs?)
1266 Read, Write, Clean, Delete
1267 - Login to Bacula; Bacula users with different permissions:
1268 owner, group, user, quotas
1269 - Store info on each file system type (probably in the job header on tape.
1270 This could be the output of df; or perhaps some sort of /etc/mtab record.
1272 ========= ideas ===============
1273 From: "Jerry K. Schieffer" <jerry@skylinetechnology.com>
1274 To: <kern@sibbald.com>
1275 Subject: RE: [Bacula-users] future large programming jobs
1276 Date: Thu, 26 Feb 2004 11:34:54 -0600
1278 I noticed the subject thread and thought I would offer the following
1279 merely as sources of ideas, i.e. something to think about, not even as
1280 strong as a request. In my former life (before retiring) I often
1281 dealt with backups and storage management issues/products as a
1282 developer and as a consultant. I am currently migrating my personal
1283 network from amanda to bacula specifically because of the ability to
1284 cross media boundaries during storing backups.
1285 Are you familiar with the commercial product called ADSM (I think IBM
1286 now sells it under the Tivoli label)? It has a couple of interesting
1287 ideas that may apply to the following topics.
1289 1. Migration: Consider that when you need to restore a system, there
1290 may be pressure to hurry. If all the information for a single client
1291 can eventually end up on the same media (and in chronological order),
1292 the restore is facillitated by not having to search past information
1293 from other clients. ADSM has the concept of "client affinity" that
1294 may be associated with it's storage pools. It seems to me that this
1295 concept (as an optional feature) might fit in your architecture for
1298 ADSM also has the concept of defining one or more storage pools as
1299 "copy pools" (almost mirrors, but only in the sense of contents).
1300 These pools provide the ability to have duplicte data stored both
1301 onsite and offsite. The copy process can be scheduled to be handled
1302 by their storage manager during periods when there is no backup
1303 activity. Again, the migration process might be a place to consider
1304 implementing something like this.
1307 > It strikes me that it would be very nice to be able to do things
1309 > have the Job(s) backing up the machines run, and once they have all
1310 > completed, start a migration job to copy the data from disks Volumes
1312 > a tape library and then to offsite storage. Maybe this can already
1314 > done with some careful scheduling and Job prioritzation; the events
1315 > mechanism described below would probably make it very easy.
1317 This is the goal. In the first step (before events), you simply
1319 the Migration to tape later.
1321 2. Base jobs: In ADSM, each copy of each stored file is tracked in
1322 the database. Once a file (unique by path and metadata such as dates,
1323 size, ownership, etc.) is in a copy pool, no more copies are made. In
1324 other words, when you start ADSM, it begins like your concept of a
1325 base job. After that it is in the "incremental" mode. You can
1326 configure the number of "generations" of files to be retained, plus a
1327 retention date after which even old generations are purged. The
1328 database tracks the contents of media and projects the percentage of
1329 each volume that is valid. When the valid content of a volume drops
1330 below a configured percentage, the valid data are migrated to another
1331 volume and the old volume is marked as empty. Note, this requires
1332 ADSM to have an idea of the contents of a client, i.e. marking the
1333 database when an existing file was deleted, but this would solve your
1334 issue of restoring a client without restoring deleted files.
1336 This is pretty far from what bacula now does, but if you are going to
1337 rip things up for Base jobs,.....
1338 Also, the benefits of this are huge for very large shops, especially
1339 with media robots, but are a pain for shops with manual media
1343 > Base jobs sound pretty useful, but I'm not dying for them.
1345 Nobody is dying for them, but when you see what it does, you will die
1348 3. Restoring deleted files: Since I think my comments in (2) above
1349 have low probability of implementation, I'll also suggest that you
1350 could approach the issue of deleted files by a mechanism of having the
1351 fd report to the dir, a list of all files on the client for every
1352 backup job. The dir could note in the database entry for each file
1353 the date that the file was seen. Then if a restore as of date X takes
1354 place, only files that exist from before X until after X would be
1355 restored. Probably the major cost here is the extra date container in
1356 each row of the files table.
1358 Thanks for "listening". I hope some of this helps. If you want to
1359 contact me, please send me an email - I read some but not all of the
1360 mailing list traffic and might miss a reply there.
1362 Please accept my compliments for bacula. It is doing a great job for
1363 me!! I sympathize with you in the need to wrestle with excelence in
1364 execution vs. excelence in feature inclusion.
1369 ==============================
1372 - Design at hierarchial storage for Bacula. Migration and Clone.
1373 - Implement FSM (File System Modules).
1374 - Audit M_ error codes to ensure they are correct and consistent.
1375 - Add variable break characters to lex analyzer.
1376 Either a bit mask or a string of chars so that
1377 the caller can change the break characters.
1378 - Make a single T_BREAK to replace T_COMMA, etc.
1379 - Ensure that File daemon and Storage daemon can
1380 continue a save if the Director goes down (this
1381 is NOT currently the case). Must detect socket error,
1382 buffer messages for later.
1383 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
1384 - Add ability to backup to two Storage devices (two SD sessions) at
1385 the same time -- e.g. onsite, offsite.
1386 - Add the ability to consolidate old backup sets (basically do a restore
1387 to tape and appropriately update the catalog). Compress Volume sets.
1388 Might need to spool via file is only one drive is available.
1389 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
1390 someway to do so with every volume that has less than x% valid
1394 Migration: Move a backup from one Volume to another
1395 Clone: Copy a backup -- two Volumes
1397 Bacula Migration is based on Jobs (apparently Networker is file by file).
1399 Migration triggered by:
1403 Highwater mark (keep total size)
1408 ======================================================
1410 It is somewhat like a Full save becomes an incremental since
1411 the Base job (or jobs) plus other non-base files.
1413 - A Base backup is same as Full backup, just different type.
1414 - New BaseFiles table that contains:
1416 BaseJobId - Base JobId referenced for this FileId (needed ???)
1417 JobId - JobId currently running
1418 FileId - File not backed up, exists in Base Job
1419 FileIndex - FileIndex from Base Job.
1420 i.e. for each base file that exists but is not saved because
1421 it has not changed, the File daemon sends the JobId, BaseId,
1422 FileId, FileIndex back to the Director who creates the DB entry.
1423 - To initiate a Base save, the Director sends the FD
1424 the FileId, and full filename for each file in the Base.
1425 - When the FD finds a Base file, he requests the Director to
1426 send him the full File entry (stat packet plus MD5), or
1427 conversely, the FD sends it to the Director and the Director
1428 says yes or no. This can be quite rapid if the FileId is kept
1429 by the FD for each Base Filename.
1430 - It is probably better to have the comparison done by the FD
1431 despite the fact that the File entry must be sent across the
1433 - An alternative would be to send the FD the whole File entry
1434 from the start. The disadvantage is that it requires a lot of
1435 space. The advantage is that it requires less communications
1437 - The Job record must be updated to indicate that one or more
1439 - At end of Job, FD returns:
1440 1. Count of base files/bytes not written to tape (i.e. matches)
1441 2. Count of base file that were saved i.e. had changed.
1442 - No tape record would be written for a Base file that matches, in the
1443 same way that no tape record is written for Incremental jobs where
1444 the file is not saved because it is unchanged.
1445 - On a restore, all the Base file records must explicitly be
1446 found from the BaseFile tape. I.e. for each Full save that is marked
1447 to have one or more Base Jobs, search the BaseFile for all occurrences
1449 - An optimization might be to make the BaseFile have:
1455 This would avoid the need to explicitly fetch each File record for
1456 the Base job. The Base Job record will be fetched to get the
1457 VolSessionId and VolSessionTime.
1458 =========================================================
1461 ==========================================================
1463 For each Incremental job that is run, there may be files that
1464 were found but not saved because they were locked (this applies
1465 only to Windows). Such a system could send back to the Director
1466 a list of Unsaved files.
1468 - New UnSavedFiles table that contains:
1472 - Then in the next Incremental job, the list of Unsaved Files will be
1473 feed to the FD, who will ensure that they are explicitly chosen even
1474 if standard date/time check would not have selected them.
1475 =============================================================
1479 Multiple drive autochanger data: see Alan Brown
1480 mtx -f xxx unloadStorage Element 1 is Already Full(drive 0 was empty)
1481 Unloading Data Transfer Element into Storage Element 1...source Element
1482 Address 480 is Empty
1484 (drive 0 was empty and so was slot 1)
1485 > mtx -f xxx load 15 0
1486 no response, just returns to the command prompt when complete.
1487 > mtx -f xxx status Storage Changer /dev/changer:2 Drives, 60 Slots ( 2 Import/Export )
1488 Data Transfer Element 0:Full (Storage Element 15 Loaded):VolumeTag = HX001
1489 Data Transfer Element 1:Empty
1490 Storage Element 1:Empty
1491 Storage Element 2:Full :VolumeTag=HX002
1492 Storage Element 3:Full :VolumeTag=HX003
1493 Storage Element 4:Full :VolumeTag=HX004
1494 Storage Element 5:Full :VolumeTag=HX005
1495 Storage Element 6:Full :VolumeTag=HX006
1496 Storage Element 7:Full :VolumeTag=HX007
1497 Storage Element 8:Full :VolumeTag=HX008
1498 Storage Element 9:Full :VolumeTag=HX009
1499 Storage Element 10:Full :VolumeTag=HX010
1500 Storage Element 11:Empty
1501 Storage Element 12:Empty
1502 Storage Element 13:Empty
1503 Storage Element 14:Empty
1504 Storage Element 15:Empty
1505 Storage Element 16:Empty....
1506 Storage Element 28:Empty
1507 Storage Element 29:Full :VolumeTag=CLNU01L1
1508 Storage Element 30:Empty....
1509 Storage Element 57:Empty
1510 Storage Element 58:Full :VolumeTag=NEX261L2
1511 Storage Element 59 IMPORT/EXPORT:Empty
1512 Storage Element 60 IMPORT/EXPORT:Empty
1514 Unloading Data Transfer Element into Storage Element 15...done
1516 (just to verify it remembers where it came from, however it can be
1517 overrriden with mtx unload {slotnumber} to go to any storage slot.)
1519 There needs to be a table of drive # to devices somewhere - If there are
1520 multiple changers or drives there may not be a 1:1 correspondance between
1521 changer drive number and system device name - and depending on the way the
1522 drives are hooked up to scsi busses, they may not be linearly numbered
1523 from an offset point either.something like
1525 Autochanger drives = 2
1526 Autochanger drive 0 = /dev/nst1
1527 Autochanger drive 1 = /dev/nst2
1528 IMHO, it would be _safest_ to use explicit mtx unload commands at all
1529 times, not just for multidrive changers. For a 1 drive changer, that's
1535 MTX's manpage (1.2.15):
1536 unload [<slotnum>] [ <drivenum> ]
1537 Unloads media from drive <drivenum> into slot
1538 <slotnum>. If <drivenum> is omitted, defaults to
1539 drive 0 (as do all commands). If <slotnum> is
1540 omitted, defaults to the slot that the drive was
1541 loaded from. Note that there's currently no way
1542 to say 'unload drive 1's media to the slot it
1543 came from', other than to explicitly use that
1544 slot number as the destination.AB
1550 undef# camcontrol devlist
1551 <WANGTEK 51000 SCSI M74H 12B3> at scbus0 target 2 lun 0 (pass0,sa0)
1552 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 0 (pass1,sa1)
1553 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 1 (pass2)
1555 tapeinfo -f /dev/sg0 with a bad tape in drive 1:
1556 [kern@rufus mtx-1.2.17kes]$ ./tapeinfo -f /dev/sg0
1557 Product Type: Tape Drive
1559 Product ID: 'C5713A '
1561 Attached Changer: No
1562 TapeAlert[3]: Hard Error: Uncorrectable read/write error.
1563 TapeAlert[20]: Clean Now: The tape drive neads cleaning NOW.
1570 Medium Type: Not Loaded
1573 DataCompEnabled: yes
1574 DataCompCapable: yes
1575 DataDeCompEnabled: yes
1582 Handling removable disks
1584 From: Karl Cunningham <karlc@keckec.com>
1586 My backups are only to hard disk these days, in removable bays. This is my
1587 idea of how a backup to hard disk would work more smoothly. Some of these
1588 things Bacula does already, but I mention them for completeness. If others
1589 have better ways to do this, I'd like to hear about it.
1591 1. Accommodate several disks, rotated similar to how tapes are. Identified
1592 by partition volume ID or perhaps by the name of a subdirectory.
1593 2. Abort & notify the admin if the wrong disk is in the bay.
1594 3. Write backups to different subdirectories for each machine to be backed
1596 4. Volumes (files) get created as needed in the proper subdirectory, one
1598 5. When a disk is recycled, remove or zero all old backup files. This is
1599 important as the disk being recycled may be close to full. This may be
1600 better done manually since the backup files for many machines may be
1601 scattered in many subdirectories.
1606 - Make sure that all do_prompt() calls in Dir check for
1607 -1 (error) and -2 (cancel) returns.
1608 - Fix foreach_jcr() to have free_jcr() inside next().
1609 jcr=jcr_walk_start();
1610 for ( ; jcr; (jcr=jcr_walk_next(jcr)) )
1613 - A Volume taken from Scratch should take on the retention period
1615 - Correct doc for Maximum Changer Wait (and others) accepting only
1617 - Implement status that shows why a job is being held in reserve, or
1618 rather why none of the drives are suitable.
1619 - Implement a way to disable a drive (so you can use the second
1620 drive of an autochanger, and the first one will not be used or
1622 - Make sure Maximum Volumes is respected in Pools when adding
1623 Volumes (e.g. when pulling a Scratch volume).
1624 - Keep same dcr when switching device ...
1625 - Implement code that makes the Dir aware that a drive is an
1626 autochanger (so the user doesn't need to use the Autochanger = yes
1628 - Make catalog respect ACL.
1629 - Add recycle count to Media record.
1630 - Add initial write date to Media record.
1631 - Fix store_yesno to be store_bitmask.
1632 --- create_file.c.orig Fri Jul 8 12:13:05 2005
1633 +++ create_file.c Fri Jul 8 12:13:07 2005
1635 attr->ofname, be.strerror());
1638 + } else if(S_ISSOCK(attr->statp.st_mode)) {
1639 + Dmsg1(200, "Skipping socket: %s\n", attr->ofname);
1641 Dmsg1(200, "Restore node: %s\n", attr->ofname);
1642 if (mknod(attr->ofname, attr->statp.st_mode, attr->statp.st_rdev) != 0 && errno != EEXIST) {
1643 - Add true/false to conf same as yes/no
1644 - Reserve blocks other restore jobs when first cannot connect to SD.
1645 - Fix Maximum Changer Wait, Maximum Open Wait, Maximum Rewind Wait to
1646 accept time qualifiers.
1647 - Does ClientRunAfterJob fail the job on a bad return code?
1648 - Make hardlink code at line 240 of find_one.c use binary search.
1649 - Add ACL error messages in src/filed/acl.c.
1650 - Make authentication failures single threaded.
1651 - Make Dir and SD authentication errors single threaded.
1652 - Fix catreq.c digestbuf at line 411 in src/dird/catreq.c
1653 - Make base64.c (bin_to_base64) take a buffer length
1654 argument to avoid overruns.
1655 and verify that other buffers cannot overrun.
1656 - Implement VolumeState as discussed with Arno.
1657 - Add LocationId to update volume
1665 - Add Comment to Media record
1666 - Fix auth compatibility with 1.38
1667 - Update dbcheck to include Log table
1668 - Update llist to include new fields.
1669 - Make unmount unload autochanger. Make mount load slot.
1670 - Fix bscan to report the JobType when restoring a job.
1671 - Fix wx-console scanning problem with commas in names.
1672 - Add manpages to the list of directories for make install. Notify
1674 - Add bconsole option to use stdin/out instead of conio.
1675 - Fix ClientRunBefore/AfterJob compatibility.
1676 - Ensure that connection to daemon failure always indicates what
1677 daemon it was trying to connect to.
1678 - Freespace on DVD requested over and over even with no intervening
1680 - .update volume [enabled|disabled|*see below]
1681 > However, I could easily imagine an option to "update slots" that says
1682 > "enable=yes|no" that would automatically enable or disable all the Volumes
1683 > found in the autochanger. This will permit the user to optionally mark all
1684 > the Volumes in the magazine disabled prior to taking them offsite, and mark
1685 > them all enabled when bringing them back on site. Coupled with the options
1686 > to the slots keyword, you can apply the enable/disable to any or all volumes.
1687 - Restricted consoles start in the Default catalog even if it