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
26 - Make authentication failures single threaded.
27 - Make base64.c (bin_to_base64) take a buffer length
28 argument to avoid overruns.
29 - Fix catreq.c digestbuf at line 411 in src/dird/catreq.c
30 and verify that other buffers cannot overrun.
31 - Implement VolumeState as discussed with Arno.
33 - Document techniques for restoring large numbers of files.
34 - Document setting my.cnf to big file usage.
35 - Add example of proper index output to doc.
37 - Fix re-read of last block to check if job has actually written
38 a block, and check if block was written by a different job
39 (i.e. multiple simultaneous jobs writing).
40 - JobStatus and Termination codes.
41 - Some users claim that they must do two prune commands to get a
42 Volume marked as purged.
43 - Print warning message if LANG environment variable does not specify
45 - New dot commands from Arno.
46 .update volume [enabled|disabled|*see below]
47 .show device=xxx lists information from one storage device, including
48 devices (I'm not even sure that information exists in the DIR...)
49 .move eject device=xxx mostly the same as 'unmount xxx' but perhaps with
50 better machine-readable output like "Ok" or "Error busy"
51 .move eject device=xxx toslot=yyy the same as above, but with a new
52 target slot. The catalog should be updated accordingly.
53 .move transfer device=xxx fromslot=yyy toslot=zzz
56 === Migration from David ===
60 Name = "<poolname>-migrate"
64 Migration Selection Type = LowestUtil | OldestVol | PoolOccupancy |
65 Client | PoolResidence | Volume | JobName | SQLquery
66 Migration Selection Pattern = "regexp"
67 Next Pool = <override>
70 There should be no need for a Level (migration is always Full, since you
71 don't calculate differential/incremental differences for migration),
72 Storage should be determined by the volume types in the pool, and Client
73 is really a selection issue. Migration should always occur to the
74 NextPool defined in the pool definition. If no nextpool is defined, the
75 job should end with a reason of "no place to go". If Next Pool statement
76 is present, we override the check in the pool definition and use the
79 Here's how I'd define Migration Selection Types:
82 Client -- Migrate data from selected client only. Migration Selection
83 Pattern regexp provides pattern to select client names, eg ^FS00* makes
84 all client names starting with FS00 eligible for migration.
86 Jobname -- Migration all jobs matching name. Migration Selection Pattern
87 regexp provides pattern to select jobnames existing in pool.
89 Volume -- Migrate all data on specified volumes. Migration Selection
90 Pattern regexp provides selection criteria for volumes to be migrated.
91 Volumes must exist in pool to be eligible for migration.
95 LowestUtil -- Identify the volume in the pool with the least data on it
96 and empty it. No Migration Selection Pattern required.
98 OldestVol -- Identify the LRU volume with data written, and empty it. No
99 Migration Selection Pattern required.
101 PoolOccupancy -- if pool occupancy exceeds <highmig>, migrate volumes
102 (starting with most full volumes) until pool occupancy drops below
103 <lowmig>. Pool highmig and lowmig values are in pool definition, no
104 Migration Selection Pattern required.
108 SQLQuery -- Migrate all jobuids returned by the supplied SQL query.
109 Migration Selection Pattern contains SQL query to execute; should return
110 a list of 1 or more jobuids to migrate.
112 PoolResidence -- Migrate data sitting in pool for longer than
113 PoolResidence value in pool definition. Migration Selection Pattern
114 optional; if specified, override value in pool definition (value in
118 [ possibly a Python event -- kes ]
120 - Mount on an Autochanger with no tape in the drive causes:
121 Automatically selected Storage: LTO-changer
122 Enter autochanger drive[0]: 0
123 3301 Issuing autochanger "loaded drive 0" command.
124 3302 Autochanger "loaded drive 0", result: nothing loaded.
125 3301 Issuing autochanger "loaded drive 0" command.
126 3302 Autochanger "loaded drive 0", result: nothing loaded.
127 3902 Cannot mount Volume on Storage Device "LTO-Drive1" (/dev/nst0) because:
128 Couldn't rewind device "LTO-Drive1" (/dev/nst0): ERR=dev.c:678 Rewind error on "LTO-Drive1" (/dev/nst0). ERR=No medium found.
129 3905 Device "LTO-Drive1" (/dev/nst0) open but no Bacula volume is mounted.
130 If this is not a blank tape, try unmounting and remounting the Volume.
131 - If Drive 0 is blocked, and drive 1 is set "Autoselect=no", drive 1 will
133 - Autochanger did not change volumes.
134 select * from Storage;
135 +-----------+-------------+-------------+
136 | StorageId | Name | AutoChanger |
137 +-----------+-------------+-------------+
138 | 1 | LTO-changer | 0 |
139 +-----------+-------------+-------------+
140 05-May 03:50 roxie-sd: 3302 Autochanger "loaded drive 0", result is Slot 11.
141 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Warning: Director wanted Volume "LT
142 Current Volume "LT0-002" not acceptable because:
143 1997 Volume "LT0-002" not in catalog.
144 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Error: Autochanger Volume "LT0-002"
145 Setting InChanger to zero in catalog.
146 05-May 03:50 roxie-dir: Tibs.2006-05-05_03.05.02 Error: Unable to get Media record
148 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: Error getting Volume i
149 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: Job 530 canceled.
150 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: spool.c:249 Fatal appe
151 05-May 03:49 Tibs: Tibs.2006-05-05_03.05.02 Fatal error: c:\cygwin\home\kern\bacula
160 FirstWritten: 2006-05-05 03:11:54
161 LastWritten: 2006-05-05 03:50:23
162 LabelDate: 2005-12-26 16:52:40
173 VolRetention: 31,536,000
185 Note VolStatus is blank!!!!!
192 FirstWritten: 0000-00-00 00:00:00
193 LastWritten: 0000-00-00 00:00:00
194 LabelDate: 2005-12-26 16:52:40
205 VolRetention: 31,536,000
218 Automatically selected Storage: LTO-changer
219 Enter autochanger drive[0]: 0
220 3301 Issuing autochanger "loaded drive 0" command.
221 3302 Autochanger "loaded drive 0", result: nothing loaded.
222 3301 Issuing autochanger "loaded drive 0" command.
223 3302 Autochanger "loaded drive 0", result: nothing loaded.
224 3902 Cannot mount Volume on Storage Device "LTO-Drive1" (/dev/nst0) because:
225 Couldn't rewind device "LTO-Drive1" (/dev/nst0): ERR=dev.c:678 Rewind error on "LTO-Drive1" (/dev/nst0). ERR=No medium found.
227 3905 Device "LTO-Drive1" (/dev/nst0) open but no Bacula volume is mounted.
228 If this is not a blank tape, try unmounting and remounting the Volume.
230 - Add VolumeState (enable, disable, archive)
231 - Add VolumeLock to prevent all but lock holder (SD) from updating
232 the Volume data (with the exception of VolumeState).
233 - The btape fill command does not seem to use the Autochanger
234 - Make Windows installer default to system disk drive.
235 - Look at using ioctl(FIOBMAP, ...) on Linux, and
236 DeviceIoControl(..., FSCTL_QUERY_ALLOCATED_RANGES, ...) on
237 Win32 for sparse files.
238 http://www.flexhex.com/docs/articles/sparse-files.phtml
239 http://www.informatik.uni-frankfurt.de/~loizides/reiserfs/fibmap.html
240 - Directive: at <event> "command"
241 - Command: pycmd "command" generates "command" event. How to
242 attach to a specific job?
243 - Integrate Christopher's St. Bernard code.
244 - run_cmd() returns int should return JobId_t
245 - get_next_jobid_from_list() returns int should return JobId_t
246 - Document export LDFLAGS=-L/usr/lib64
247 - Don't attempt to restore from "Disabled" Volumes.
248 - Network error on Win32 should set Win32 error code.
249 - What happens when you rename a Disk Volume?
250 - Job retention period in a Pool (and hence Volume). The job would
252 - Detect resource deadlock in Migrate when same job wants to read
253 and write the same device.
254 - Queue warning/error messages during restore so that they
255 are reported at the end of the report rather than being
256 hidden in the file listing ...
257 - Look at -D_FORTIFY_SOURCE=2
258 - Add Win32 FileSet definition somewhere
259 - Look at fixing restore status stats in SD.
260 - Make selection of Database used in restore correspond to
262 - Look at using ioctl(FIMAP) and FIGETBSZ for sparse files.
263 http://www.informatik.uni-frankfurt.de/~loizides/reiserfs/fibmap.html
264 - Implement a mode that says when a hard read error is
265 encountered, read many times (as it currently does), and if the
266 block cannot be read, skip to the next block, and try again. If
267 that fails, skip to the next file and try again, ...
269 create table LevelType (LevelType binary(1), LevelTypeLong tinyblob);
270 insert into LevelType (LevelType,LevelTypeLong) values
274 - Add ACL to restore only to original location.
275 - Show files/second in client status output.
276 - Add a recursive mark command (rmark) to restore.
277 - "Minimum Job Interval = nnn" sets minimum interval between Jobs
278 of the same level and does not permit multiple simultaneous
279 running of that Job (i.e. lets any previous invocation finish
280 before doing Interval testing).
281 - Look at simplifying File exclusions.
282 - New directive "Delete purged Volumes"
283 - new pool XXX with ScratchPoolId = MyScratchPool's PoolId and
284 let it fill itself, and RecyclePoolId = XXX's PoolId so I can
285 see if it become stable and I just have to supervise
287 - If I want to remove this pool, I set RecyclePoolId = MyScratchPool's
288 PoolId, and when it is empty remove it.
289 - Figure out how to recycle Scratch volumes back to the Scratch Pool.
291 - Allow Check Labels to be used with Bacula labels.
292 - "Resuming" a failed backup (lost line for example) by using the
293 failed backup as a sort of "base" job.
295 - Email to the user when the tape is about to need changing x
296 days before it needs changing.
297 - Command to show next tape that will be used for a job even
298 if the job is not scheduled.
299 - From: Arunav Mandal <amandal@trolltech.com>
300 1. When jobs are running and bacula for some reason crashes or if I do a
301 restart it remembers and jobs it was running before it crashed or restarted
302 as of now I loose all jobs if I restart it.
304 2. When spooling and in the midway if client is disconnected for instance a
305 laptop bacula completely discard the spool. It will be nice if it can write
306 that spool to tape so there will be some backups for that client if not all.
308 3. We have around 150 clients machines it will be nice to have a option to
309 upgrade all the client machines bacula version automatically.
311 4. Atleast one connection should be reserved for the bconsole so at heavy load
312 I should connect to the director via bconsole which at sometimes I can't
314 5. Another most important feature that is missing, say at 10am I manually
315 started backup of client abc and it was a full backup since client abc has
316 no backup history and at 10.30am bacula again automatically started backup of
317 client abc as that was in the schedule. So now we have 2 multiple Full
318 backups of the same client and if we again try to start a full backup of
319 client backup abc bacula won't complain. That should be fixed.
321 - Fix bpipe.c so that it does not modify results pointer.
322 ***FIXME*** calling sequence should be changed.
323 - For Windows disaster recovery see http://unattended.sf.net/
324 - regardless of the retention period, Bacula will not prune the
325 last Full, Diff, or Inc File data until a month after the
326 retention period for the last Full backup that was done.
327 - update volume=xxx --- add status=Full
328 - Remove old spool files on startup.
329 - Exclude SD spool/working directory.
330 - Refuse to prune last valid Full backup. Same goes for Catalog.
332 - Make a callback when Rerun failed levels is called.
333 - Give Python program access to Scheduled jobs.
334 - Add setting Volume State via Python.
335 - Python script to save with Python, not save, save with Bacula.
336 - Python script to do backup.
338 - Change the Priority, Client, Storage, JobStatus (error)
339 at the start of a job.
340 - Why is SpoolDirectory = /home/bacula/spool; not reported
341 as an error when writing a DVD?
342 - Make bootstrap file handle multiple MediaTypes (SD)
343 - Remove all old Device resource code in Dir and code to pass it
344 back in SD -- better, rework it to pass back device statistics.
345 - Check locking of resources -- be sure to lock devices where previously
346 resources were locked.
347 - The last part is left in the spool dir.
350 - In restore don't compare byte count on a raw device -- directory
351 entry does not contain bytes.
352 - To mark files as deleted, run essentially a Verify to disk, and
353 when a file is found missing (MarkId != JobId), then create
354 a new File record with FileIndex == -1. This could be done
355 by the FD at the same time as the backup.
359 MA = (last_MA * 3 + rate) / 4
360 rate = (bytes - last_bytes) / (runtime - last_runtime)
361 - Max Vols limit in Pool off by one?
362 - Implement Files/Bytes,... stats for restore job.
363 - Implement Total Bytes Written, ... for restore job.
364 - Despool attributes simultaneously with data in a separate
365 thread, rejoined at end of data spooling.
366 - 7. Implement new Console commands to allow offlining/reserving drives,
367 and possibly manipulating the autochanger (much asked for).
368 - Add start/end date editing in messages (%t %T, %e?) ...
369 - Add ClientDefs similar to JobDefs.
370 - Print more info when bextract -p accepts a bad block.
371 - Fix FD JobType to be set before RunBeforeJob in FD.
372 - Look at adding full Volume and Pool information to a Volume
373 label so that bscan can get *all* the info.
374 - If the user puts "Purge Oldest Volume = yes" or "Recycle Oldest Volume = yes"
375 and there is only one volume in the pool, refuse to do it -- otherwise
376 he fills the Volume, then immediately starts reusing it.
377 - Implement copies and stripes.
378 - Add history file to console.
379 - Each file on tape creates a JobMedia record. Peter has 4 million
380 files spread over 10000 tape files and four tapes. A restore takes
381 16 hours to build the restore list.
382 - Add and option to check if the file size changed during backup.
383 - Make sure SD deletes spool files on error exit.
384 - Delete old spool files when SD starts.
385 - When labeling tapes, if you enter 000026, Bacula uses
386 the tape index rather than the Volume name 000026.
387 - Add offline tape command to Bacula console.
389 Enter MediaId or Volume name: 32
390 Enter new Volume name: DLT-20Dec04
391 Automatically selected Pool: Default
392 Connecting to Storage daemon DLTDrive at 192.168.68.104:9103 ...
393 Sending relabel command from "DLT-28Jun03" to "DLT-20Dec04" ...
394 block.c:552 Write error at 0:0 on device /dev/nst0. ERR=Bad file descriptor.
395 Error writing final EOF to tape. This tape may not be readable.
396 dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
397 askdir.c:219 NULL Volume name. This shouldn't happen!!!
398 3912 Failed to label Volume: ERR=dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
399 Label command failed for Volume DLT-20Dec04.
400 Do not forget to mount the drive!!!
401 - Bug: if a job is manually scheduled to run later, it does not appear
402 in any status report and cannot be cancelled.
404 ==== Keeping track of deleted files ====
405 My "trick" for keeping track of deletions is the following.
406 Assuming the user turns on this option, after all the files
407 have been backed up, but before the job has terminated, the
408 FD will make a pass through all the files and send their
409 names to the DIR (*exactly* the same as what a Verify job
410 currently does). This will probably be done at the same
411 time the files are being sent to the SD avoiding a second
412 pass. The DIR will then compare that to what is stored in
413 the catalog. Any files in the catalog but not in what the
414 FD sent will receive a catalog File entry that indicates
415 that at that point in time the file was deleted.
417 During a restore, any file initially picked up by some
418 backup (Full, ...) then subsequently having a File entry
419 marked "delete" will be removed from the tree, so will not
420 be restored. If a file with the same name is later OK it
421 will be inserted in the tree -- this already happens. All
422 will be consistent except for possible changes during the
425 Since I'm on the subject, some of you may be wondering what
426 the utility of the in memory tree is if you are going to
427 restore everything (at least it comes up from time to time
428 on the list). Well, it is still *very* useful because it
429 allows only the last item found for a particular filename
430 (full path) to be entered into the tree, and thus if a file
431 is backed up 10 times, only the last copy will be restored.
432 I recently (last Friday) restored a complete directory, and
433 the Full and all the Differential and Incremental backups
434 spanned 3 Volumes. The first Volume was not even mounted
435 because all the files had been updated and hence backed up
436 since the Full backup was made. In this case, the tree
437 saved me a *lot* of time.
439 Make sure this information is stored on the tape too so
440 that it can be restored directly from the tape.
442 Comments from Martin Simmons (I think they are all covered):
443 Ok, that should cover the basics. There are few issues though:
445 - Restore will depend on the catalog. I think it is better to include the
446 extra data in the backup as well, so it can be seen by bscan and bextract.
448 - I'm not sure if it will preserve multiple hard links to the same inode. Or
449 maybe adding or removing links will cause the data to be dumped again?
451 - I'm not sure if it will handle renamed directories. Possibly it will work
452 by dumping the whole tree under a renamed directory?
454 - It remains to be seen how the backup performance of the DIR's will be
455 affected when comparing the catalog for a large filesystem.
459 How about introducing a Type = MgmtPolicy job type? That job type would
460 be responsible for scanning the Bacula environment looking for specific
461 conditions, and submitting the appropriate jobs for implementing said
465 Name = "Migration-Policy"
467 Policy Selection Job Type = Migrate
468 Scope = "<keyword> <operator> <regexp>"
469 Threshold = "<keyword> <operator> <regexp>"
470 Job Template = <template-name>
473 Where <keyword> is any legal job keyword, <operator> is a comparison
474 operator (=,<,>,!=, logical operators AND/OR/NOT) and <regexp> is a
475 appropriate regexp. I could see an argument for Scope and Threshold
476 being SQL queries if we want to support full flexibility. The
477 Migration-Policy job would then get scheduled as frequently as a site
478 felt necessary (suggested default: every 15 minutes).
483 Name = "Migration-Policy"
485 Policy Selection Job Type = Migration
487 Threshold = "Migration Selection Type = LowestUtil"
488 Job Template = "MigrationTemplate"
491 would select all pools for examination and generate a job based on
492 MigrationTemplate to automatically select the volume with the lowest
493 usage and migrate it's contents to the nextpool defined for that pool.
495 This policy abstraction would be really handy for adjusting the behavior
496 of Bacula according to site-selectable criteria (one thing that pops
497 into mind is Amanda's ability to automatically adjust backup levels
498 depending on various criteria).
504 - Add Pool/Storage override regression test.
505 - Add delete JobId to regression.
506 - Add a regression test for dbcheck.
507 - New test to add bscan to four-concurrent-jobs regression,
508 i.e. after the four-concurrent jobs zap the
509 database as is done in the bscan-test, then use bscan to
510 restore the database, do a restore and compare with the
512 - Add restore of specific JobId to regression (item 3
513 on the restore prompt)
514 - Add IPv6 to regression
515 - Add database test to regression. Test each function like delete,
518 - AntiVir can slow down backups on Win32 systems.
519 - Win32 systems with FAT32 can be much slower than NTFS for
520 more than 1000 files per directory.
524 - A HOLD command to stop all jobs from starting.
525 - A PAUSE command to pause all running jobs ==> release the
527 - Media Type = LTO,LTO-2,LTO-3
528 Media Type Read = LTO,LTO2,LTO3
529 Media Type Write = LTO2, LTO3
531 === From Carsten Menke <bootsy52@gmx.net>
533 Following is a list of what I think in the situations where I'm faced with,
534 could be a usefull enhancement to bacula, which I'm certain other users will
535 benefit from as well.
537 1. NextJob/NextJobs Directive within a Job Resource in the form of
538 NextJobs = job1,job2.
541 I currently solved the problem with running multiple jobs each after each
542 by setting the Max Wait Time for a job to 8 hours, and give
543 the jobs different Priorities. However, there scenarios where
544 1 Job is directly depending on another job, so if the former job fails,
545 the job after it needn't to be run
546 while maybe other jobs should run despite of that
549 A Backup Job and a Verify job, if the backup job fails there is no need to run
550 the verify job, as the backup job already failed. However, one may like
551 to backup the Catalog to disk despite of that the main backup job failed.
554 I see that this is related to the Event Handlers which are on the ToDo
555 list, also it is maybe a good idea to check for the return value and
556 execute different actions based on the return value
559 3. offline capability to bconsole
562 Currently I use a script which I execute within the last Job via the
563 RunAfterJob Directive, to release and eject the tape.
564 So I have to call bconsole "release=Storage-Name" and afterwards
565 mt -f /dev/nst0 eject to get the tape out.
567 If I have multiple Storage Devices, than these may not be /dev/nst0 and
568 I have to modify the script or call it with parameters etc.
569 This would actually not be needed, as everything is already defined
570 in bacula-sd.conf and if I can invoke bconsole with the
571 storage name via $1 in the script than I'm done and information is
574 4. %s for Storage Name added to the chars being substituted in "RunAfterJob"
578 For the reason mentioned in 3. to have the ability to call a
579 script with /scripts/foobar %s and in the script use $1
580 to pass the Storage Name to bconsole
582 5. Setting Volume State within a Job Resource
585 Instead of using "Maximum Volume Jobs" in the Pool Resource,
586 I would have the possibilty to define
587 in a Job Resource that after this certain job is run, the Volume State
588 should be set to "Volume State = Used", this give more flexibility (IMHO).
590 6. Localization of Bacula Messages
593 Unfortunatley many,many people I work with don't speak english very well.
594 So if at least the Reporting messages would be localized then they
595 would understand that they have to change the tape,etc. etc.
597 I volunteer to do the german translations, and if I can convince my wife also
598 french and Morre (western african language).
600 7. OK, this is evil, probably bound to security risks and maybe not possible
601 due to the design of bacula.
603 Implementation of Backtics ( `command` ) for shell comand execution to
604 the "Label Format" Directive.
608 Currently I have defined BACULA_DAY_OF_WEEK="day1|day2..." resulting in
609 Label Format = "HolyBackup-${BACULA_DAY_OF_WEEK[${WeekDay}]}". If I could
610 use backticks than I could use "Label Format = HolyBackup-`date +%A` to have
611 the localized name for the day of the week appended to the
612 format string. Then I have the tape labeled automatically with weekday
613 name in the correct language.
615 - Yes, that is surely the case. I probably should turn those into Warning
616 errors. In addition, you just made me think that it might not be bad to
617 add an option to check the file size after backing up the file and
618 report if it changes. This would be done as an option because it would
621 Kern, good idea. If you do do that, mention in the output: file
622 shrunk, or file expanded, just to make it obvious to the user
623 (without having to the refer to file size), just how the file size
626 Would this option be for all file, or just one file? Or a fileset?
627 - Make output from status use html table tags for nicely
628 presenting in a browser.
629 - Can one write tapes faster with 8192 byte block sizes?
630 - Document security problems with the same password for everyone in
631 rpm and Win32 releases.
632 - Browse generations of files.
633 - I've seen an error when my catalog's File table fills up. I
634 then have to recreate the File table with a larger maximum row
635 size. Relevant information is at
636 http://dev.mysql.com/doc/mysql/en/Full_table.html ; I think the
637 "Installing and Configuring MySQL" chapter should talk a bit
638 about this potential problem, and recommend a solution.
639 - For Solaris must use POSIX awk.
640 - Want speed of writing to tape while despooling.
641 - Supported autochanger:
649 Wangtek 6525ES (SCSI-1 QIC drive, 525MB), under Linux 2.4.something,
650 bacula 1.36.0/1 works with blocksize 16k INSIDE bacula-sd.conf.
651 - Add regex from http://www.pcre.org to Bacula for Win32.
652 - Use only shell tools no make in CDROM package.
653 - Include within include does it work?
654 - Implement a Pool of type Cleaning?
655 - Implement VolReadTime and VolWriteTime in SD
656 - Modify Backing up Your Database to include a bootstrap file.
657 - Think about making certain database errors fatal.
658 - Look at correcting the time jump in the scheduler for daylight
659 savings time changes.
660 - Add a "real" timer to network connections.
661 - Promote to Full = Time period
662 - Check dates entered by user for correctness (month/day/... ranges)
663 - Compress restore Volume listing by date and first file.
664 - Look at patches/bacula_db.b2z postgresql that loops during restore.
666 - Perhaps add read/write programs and/or plugins to FileSets.
667 - How to handle backing up portables ...
668 - Add some sort of guaranteed Interval for upgrading jobs.
669 - Can we write the state file after every job terminates? On Win32
670 the system crashes and the state file is not updated.
673 Documentation to do: (any release a little bit at a time)
674 - Doc to do unmount before removing magazine.
675 - Alternative to static linking "ldd prog" save all binaries listed,
676 restore them and point LD_LIBRARY_PATH to them.
677 - Document add "</dev/null >/dev/null 2>&1" to the bacula-fd command line
678 - Document query file format.
679 - Add more documentation for bsr files.
680 - Document problems with Verify and pruning.
681 - Document how to use multiple databases.
682 - VXA drives have a "cleaning required"
683 indicator, but Exabyte recommends preventive cleaning after every 75
686 In this context, it should be noted that Exabyte has a command-line
687 vxatool utility available for free download. (The current version is
688 vxatool-3.72.) It can get diagnostic info, read, write and erase tapes,
689 test the drive, unload tapes, change drive settings, flash new firmware,
691 Of particular interest in this context is that vxatool <device> -i will
692 report, among other details, the time since last cleaning in tape motion
693 minutes. This information can be retrieved (and settings changed, for
694 that matter) through the generic-SCSI device even when Bacula has the
695 regular tape device locked. (Needless to say, I don't recommend
696 changing tape settings while a job is running.)
697 - Lookup HP cleaning recommendations.
698 - Lookup HP tape replacement recommendations (see trouble shooting autochanger)
699 - Document doing table repair
702 ===================================
703 - Add macro expansions in JobDefs.
704 Run Before Job = "SomeFile %{Level} %{Client}"
705 Write Bootstrap="/some/dir/%{JobName}_%{Client}.bsr"
706 - Use non-blocking network I/O but if no data is available, use
708 - Use gather write() for network I/O.
709 - Autorestart on crash.
710 - Add bandwidth limiting.
711 - Add acks every once and a while from the SD to keep
712 the line from timing out.
713 - When an error in input occurs and conio beeps, you can back
714 up through the prompt.
715 - Detect fixed tape block mode during positioning by looking at
716 block numbers in btape "test". Possibly adjust in Bacula.
717 - Fix list volumes to output volume retention in some other
718 units, perhaps via a directive.
719 - If opening a tape in read/write mode fails attempt to open
720 it in read-only mode, and mark the tape for read only.
721 - Allow Simultaneous Priorities = yes => run up to Max concurrent jobs even
722 with multiple priorities.
723 - If you use restore replace=never, the directory attributes for
724 non-existent directories will not be restored properly.
726 - see lzma401.zip in others directory for new compression
728 - Minimal autochanger handling in Bacula and in btape.
729 - Look into how tar does not save sockets and the possiblity of
730 not saving them in Bacula (Martin Simmons reported this).
731 - Fix restore jobs so that multiple jobs can run if they
732 are not using the same tape(s).
733 - Allow the user to select JobType for manual pruning/purging.
734 - bscan does not put first of two volumes back with all info in
736 - Implement the FreeBSD nodump flag in chflags.
737 - Figure out how to make named console messages go only to that
738 console and to the non-restricted console (new console class?).
739 - Make restricted console prompt for password if *ask* is set or
740 perhaps if password is undefined.
741 - Implement "from ISO-date/time every x hours/days/weeks/months" in
744 ==== from Marc Schoechlin
745 - the help-command should be more verbose
746 (it should explain the paramters of the different
748 -> its time-comsuming to consult the manual anytime
749 you need a special parameter
750 -> maybe its more easy to maintain this, if the
751 descriptions of that commands are outsourced to
753 - the cd-command should allow complete paths
754 i.e. cd /foo/bar/foo/bar
755 -> if a customer mails me the path to a certain file,
756 its faster to enter the specified directory
757 - if the password is not configured in bconsole.conf
758 you should be asked for it.
759 -> sometimes you like to do restore on a customer-machine
760 which shouldnt know the password for bacula.
761 -> adding the password to the file favours admins
762 to forget to remove the password after usage
764 the protection of that file is less important
765 - long-listed-output of commands should be scrollable
766 like the unix more/less-command does
767 -> if someone runs 200 and more machines, the lists could
768 be a little long and complex
769 - command-output should be shown column by column
770 to reduce scrolling and to increase clarity
772 - lsmark should list the selected files with full
774 - wildcards for selecting and file and directories would be nice
775 - any actions should be interuptable with STRG+C
776 - command-expansion would be pretty cool
778 - When the replace Never option is set, new directory permissions
779 are not restored. See bug 213. To fix this requires creating a
780 list of newly restored directories so that those directory
781 permissions *can* be restored.
782 - Compaction of Disk space by "migrating" Volumes that have pruned
783 Jobs (what criteria? size, #jobs, time).
784 - Add prune all command
785 - Document fact that purge can destroy a part of a restore by purging
786 one volume while others remain valid -- perhaps mark Jobs.
787 - Add multiple-media-types.txt
788 - look at mxt-changer.html
789 - Make ? do a help command (no return needed).
790 - Implement restore directory.
791 - Document streams and how to implement them.
792 - Try not to re-backup a file if a new hard link is added.
793 - Add feature to backup hard links only, but not the data.
794 - Fix stream handling to be simpler.
795 - Add Priority and Bootstrap to Run a Job.
796 - Eliminate Restore "Run Restore Job" prompt by allowing new "run command
798 - Remove View FileSet button from Run a Job dialog.
799 - Handle prompt for restore job at end of Restore command.
800 - Add display of total selected files to Restore window.
801 - Add tree pane to left of window.
802 - Add progress meter.
803 - Max wait time or max run time causes seg fault -- see runtime-bug.txt
804 - Document writing to a CD/DVD with Bacula.
805 - Add a "base" package to the window installer for pthreadsVCE.dll
806 which is needed by all packages.
807 - Add message to user to check for fixed block size when the forward
808 space test fails in btape.
809 - When unmarking a directory check if all files below are unmarked and
810 then remove the + flag -- in the restore tree.
811 - Possibly implement: Action = Unmount Device="TapeDrive1" in Admin jobs.
812 - Setup lrrd graphs: (http://www.linpro.no/projects/lrrd/) Mike Acar.
813 - Revisit the question of multiple Volumes (disk) on a single device.
814 - Add a block copy option to bcopy.
815 - Investigate adding Mac Resource Forks.
816 - Finish work on Gnome restore GUI.
817 - Fix "llist jobid=xx" where no fileset or client exists.
818 - For each job type (Admin, Restore, ...) require only the really necessary
819 fields.- Pass Director resource name as an option to the Console.
820 - Add a "batch" mode to the Console (no unsolicited queries, ...).
821 - Add a .list all files in the restore tree (probably also a list all files)
822 Do both a long and short form.
823 - Allow browsing the catalog to see all versions of a file (with
824 stat data on each file).
825 - Restore attributes of directory if replace=never set but directory
827 - Use SHA1 on authentication if possible.
828 - See comtest-xxx.zip for Windows code to talk to USB.
829 - Add John's appended files:
830 Appended = { /files/server/logs/http/*log }
831 and such files would be treated as follows.On a FULL backup, they would
832 be backed up like any other file.On an INCREMENTAL backup, where a
833 previous INCREMENTAL or FULL was already in thecatalogue and the length
834 of the file wasgreater than the length of the last backup, only thedata
835 added since the last backup will be dumped.On an INCREMENTAL backup, if
836 the length of the file is less than thelength of the file with the same
837 name last backed up, the completefile is dumped.On Windows systems, with
838 creation date of files, we can be evensmarter about this and not count
839 entirely upon the length.On a restore, the full and all incrementals
840 since it will beapplied in sequence to restore the file.
841 - Check new HAVE_WIN32 open bits.
842 - Check if the tape has moved before writing.
843 - Handling removable disks -- see below:
844 - Keep track of tape use time, and report when cleaning is necessary.
845 - Add FromClient and ToClient keywords on restore command (or
846 BackupClient RestoreClient).
847 - Implement a JobSet, which groups any number of jobs. If the
848 JobSet is started, all the jobs are started together.
849 Allow Pool, Level, and Schedule overrides.
850 - Enhance cancel to timeout BSOCK packets after a specific delay.
851 - Do scheduling by UTC using gmtime_r() in run_conf, scheduler, and
852 ua_status.!!! Thanks to Alan Brown for this tip.
853 - Look at updating Volume Jobs so that Max Volume Jobs = 1 will work
854 correctly for multiple simultaneous jobs.
855 - Correct code so that FileSet MD5 is calculated for < and | filename
857 - Implement the Media record flag that indicates that the Volume does disk
859 - Implement VolAddr, which is used when Volume is addressed like a disk,
860 and form it from VolFile and VolBlock.
861 - Make multiple restore jobs for multiple media types specifying
862 the proper storage type.
863 - Fix fast block rejection (stored/read_record.c:118). It passes a null
864 pointer (rec) to try_repositioning().
865 - Look at extracting Win data from BackupRead.
866 - Implement RestoreJobRetention? Maybe better "JobRetention" in a Job,
867 which would take precidence over the Catalog "JobRetention".
868 - Implement Label Format in Add and Label console commands.
869 - Possibly up network buffers to 65K. Put on variable.
870 - Put email tape request delays on one or more variables. User wants
871 to cancel the job after a certain time interval. Maximum Mount Wait?
872 - Job, Client, Device, Pool, or Volume?
873 Is it possible to make this a directive which is *optional* in multiple
874 resources, like Level? If so, I think I'd make it an optional directive
875 in Job, Client, and Pool, with precedence such that Job overrides Client
876 which in turn overrides Pool.
878 - New Storage specifications:
879 - Want to write to multiple storage devices simultaneously
880 - Want to write to multiple storage devices sequentially (in one job)
881 - Want to read/write simultaneously
882 - Key is MediaType -- it must match
884 Passed to SD as a sort of BSR record called Storage Specification
888 MediaType -> Next MediaType
890 Device -> Next Device
892 Allow multiple Storage specifications
900 Allow Multiple Pool specifications (note, Pool currently
902 Allow Multiple MediaType specifications in Dir conf
903 Allow Multiple Device specifications in Dir conf
904 Perhaps keep this in a single SSR
905 Tie a Volume to a specific device by using a MediaType that
906 is contained in only one device.
907 In SD allow Device to have Multiple MediaTypes
909 - Ideas from Jerry Scharf:
910 First let's point out some big pluses that bacula has for this
912 more importantly it's active. Thank you so much for that
913 even more important, it's not flaky
914 it has an open access catalog, opening many possibilities
915 it's pushing toward heterogeneous systems capability
917 Macintosh file client
918 macs are an interesting niche, but I fear a server is a rathole
919 working bare iron recovery for windows
920 the option for inc/diff backups not reset on fileset revision
921 a) use both change and inode update time against base time
922 b) do the full catalog check (expensive but accurate)
923 sizing guide (how much system is needed to back up N systems/files)
924 consultants on using bacula in building a disaster recovery system
926 or how to get at fancy things that one could do with bacula
927 logwatch code for bacula logs (or similar)
928 linux distro inclusion of bacula (brings good and bad, but necessary)
929 win2k/XP server capability (icky but you asked)
930 support for Oracle database ??
932 - Look at adding SQL server and Exchange support for Windows.
933 - Make dev->file and dev->block_num signed integers so that -1 can
934 be an invalid value which happens with BSR.
935 - Create VolAddr for disk files in place of VolFile and VolBlock. This
936 is needed to properly specify ranges.
937 - Add progress of files/bytes to SD and FD.
938 - Print warning message if FileId > 4 billion
939 - do a "messages" before the first prompt in Console
940 - Client does not show busy during Estimate command.
941 - Implement Console mtx commands.
942 - Implement a Mount Command and an Unmount Command where
943 the users could specify a system command to be performed
944 to do the mount, after which Bacula could attempt to
945 read the device. This is for Removeable media such as a CDROM.
946 - Most likely, this mount command would be invoked explicitly
947 by the user using the current Console "mount" and "unmount"
948 commands -- the Storage Daemon would do the right thing
949 depending on the exact nature of the device.
950 - As with tape drives, when Bacula wanted a new removable
951 disk mounted, it would unmount the old one, and send a message
952 to the user, who would then use "mount" as described above
953 once he had actually inserted the disk.
954 - Implement dump/print label to UA
955 - Spool to disk only when the tape is full, then when a tape is hung move
957 - bextract is sending everything to the log file ****FIXME****
958 - Allow multiple Storage specifications (or multiple names on
959 a single Storage specification) in the Job record. Thus a job
960 can be backed up to a number of storage devices.
961 - Implement some way for the File daemon to contact the Director
962 to start a job or pass its DHCP obtained IP number.
963 - Implement a query tape prompt/replace feature for a console
964 - Copy console @ code to gnome2-console
965 - Make AES the only encryption algorithm see
966 http://csrc.nist.gov/CryptoToolkit/aes/). It's
967 an officially adopted standard, has survived peer
968 review, and provides keys up to 256 bits.
969 - Take a careful look at SetACL http://setacl.sourceforge.net
970 - Make tree walk routines like cd, ls, ... more user friendly
971 by handling spaces better.
972 - Make sure that Bacula rechecks the tape after the 20 min wait.
973 - Set IO_NOWAIT on Bacula TCP/IP packets.
974 - Try doing a raw partition backup and restore by mounting a
977 Yes, it would allow to highly automatic the request for new tapes. If a
978 tape is empty, bacula reads the barcodes (native or simulated), and if
979 an unused tape is found, it runs the label command with all the
980 necessary parameters.
982 By the way can bacula automatically "move" an empty/purged volume say
983 in the "short" pool to the "long" pool if this pool runs out of volume
985 - What to do about "list files job=xxx".
986 - Get and test MySQL 4.0
987 - Look at how fuser works and /proc/PID/fd that is how Nic found the
988 file descriptor leak in Bacula.
989 - Implement WrapCounters in Counters.
990 - Add heartbeat from FD to SD if hb interval expires.
991 - Can we dynamically change FileSets?
992 - If pool specified to label command and Label Format is specified,
993 automatically generate the Volume name.
994 - Why can't SQL do the filename sort for restore?
995 - Add ExhautiveRestoreSearch
996 - Look at the possibility of loading only the necessary
997 data into the restore tree (i.e. do it one directory at a
998 time as the user walks through the tree).
999 - Possibly use the hash code if the user selects all for a restore command.
1000 - Fix "restore all" to bypass building the tree.
1001 - Prohibit backing up archive device (findlib/find_one.c:128)
1002 - Implement Release Device in the Job resource to unmount a drive.
1003 - Implement Acquire Device in the Job resource to mount a drive,
1004 be sure this works with admin jobs so that the user can get
1005 prompted to insert the correct tape. Possibly some way to say to
1006 run the job but don't save the files.
1007 - Make things like list where a file is saved case independent for
1010 - Use autochanger to handle multiple devices.
1011 - On Windows with very long path names, it may be impossible to create
1012 a file (and thus restore it) because the total length is too long.
1013 We must cd into the directory then create the file without the
1015 - Implement a Recycle command
1016 - Test a second language e.g. french.
1017 - Start working on Base jobs.
1018 - Implement UnsavedFiles DB record.
1019 - From Phil Stracchino:
1020 It would probably be a per-client option, and would be called
1021 something like, say, "Automatically purge obsoleted jobs". What it
1022 would do is, when you successfully complete a Differential backup of a
1023 client, it would automatically purge all Incremental backups for that
1024 client that are rendered redundant by that Differential. Likewise,
1025 when a Full backup on a client completed, it would automatically purge
1026 all Differential and Incremental jobs obsoleted by that Full backup.
1027 This would let people minimize the number of tapes they're keeping on
1028 hand without having to master the art of retention times.
1029 - When doing a Backup send all attributes back to the Director, who
1030 would then figure out what files have been deleted.
1031 - Currently in mount.c:236 the SD simply creates a Volume. It should have
1032 explicit permission to do so. It should also mark the tape in error
1033 if there is an error.
1034 - Cancel waiting for Client connect in SD if FD goes away.
1036 - Implement timeout in response() when it should come quickly.
1037 - Implement a Slot priority (loaded/not loaded).
1038 - Implement "vacation" Incremental only saves.
1039 - Implement create "FileSet"?
1040 - Add prefixlinks to where or not where absolute links to FD.
1041 - Issue message to mount a new tape before the rewind.
1042 - Simplified client job initiation for portables.
1043 - If SD cannot open a drive, make it periodically retry.
1044 - Add more of the config info to the tape label.
1046 - If tape is marked read-only, then try opening it read-only rather than
1047 failing, and remember that it cannot be written.
1048 - Refine SD waiting output:
1049 Device is being positioned
1050 > Device is being positioned for append
1051 > Device is being positioned to file x
1053 - Figure out some way to estimate output size and to avoid splitting
1054 a backup across two Volumes -- this could be useful for writing CDROMs
1055 where you really prefer not to have it split -- not serious.
1056 - Have SD compute MD5 or SHA1 and compare to what FD computes.
1057 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
1058 actual data on the Volume and compare it.
1059 - Implement Bacula plugins -- design API
1060 - Make bcopy read through bad tape records.
1061 - Program files (i.e. execute a program to read/write files).
1062 Pass read date of last backup, size of file last time.
1063 - Add Signature type to File DB record.
1064 - CD into subdirectory when open()ing files for backup to
1065 speed up things. Test with testfind().
1066 - Priority job to go to top of list.
1067 - Why are save/restore of device different sizes (sparse?) Yup! Fix it.
1068 - Implement some way for the Console to dynamically create a job.
1069 - Solaris -I on tar for include list
1070 - Need a verbose mode in restore, perhaps to bsr.
1071 - bscan without -v is too quiet -- perhaps show jobs.
1072 - Add code to reject whole blocks if not wanted on restore.
1073 - Check if we can increase Bacula FD priorty in Win2000
1074 - Make sure the MaxVolFiles is fully implemented in SD
1075 - Check if both CatalogFiles and UseCatalog are set to SD.
1076 - Possibly add email to Watchdog if drive is unmounted too
1077 long and a job is waiting on the drive.
1078 - After unmount, if restore job started, ask to mount.
1079 - Add UA rc and history files.
1080 - put termcap (used by console) in ./configure and
1081 allow -with-termcap-dir.
1082 - Fix Autoprune for Volumes to respect need for full save.
1083 - Compare tape to Client files (attributes, or attributes and data)
1084 - Make all database Ids 64 bit.
1085 - Allow console commands to detach or run in background.
1086 - Fix status delay on storage daemon during rewind.
1087 - Add SD message variables to control operator wait time
1088 - Maximum Operator Wait
1089 - Minimum Message Interval
1090 - Maximum Message Interval
1091 - Send Operator message when cannot read tape label.
1092 - Verify level=Volume (scan only), level=Data (compare of data to file).
1093 Verify level=Catalog, level=InitCatalog
1095 - Add keyword search to show command in Console.
1096 - Events : tape has more than xxx bytes.
1097 - Complete code in Bacula Resources -- this will permit
1098 reading a new config file at any time.
1099 - Handle ctl-c in Console
1100 - Implement script driven addition of File daemon to config files.
1101 - Think about how to make Bacula work better with File (non-tape) archives.
1102 - Write Unix emulator for Windows.
1103 - Put memory utilization in Status output of each daemon
1104 if full status requested or if some level of debug on.
1105 - Make database type selectable by .conf files i.e. at runtime
1106 - Set flag for uname -a. Add to Volume label.
1107 - Restore files modified after date
1108 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
1109 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
1110 - Timout a job or terminate if link goes down, or reopen link and query.
1111 - Concept of precious tapes (cannot be reused).
1112 - Make bcopy copy with a single tape drive.
1113 - Permit changing ownership during restore.
1116 > My suggestion: Add a feature on the systray menu-icon menu to request
1117 > an immediate backup now. This would be useful for laptop users who may
1118 > not be on the network when the regular scheduled backup is run.
1120 > My wife's suggestion: Add a setting to the win32 client to allow it to
1121 > shut down the machine after backup is complete (after, of course,
1122 > displaying a "System will shut down in one minute, click here to cancel"
1123 > warning dialog). This would be useful for sites that want user
1124 > woorkstations to be shut down overnight to save power.
1127 - Autolabel should be specified by DIR instead of SD.
1129 - Add media capacity
1130 - AutoScan (check checksum of tape)
1131 - Format command = "format /dev/nst0"
1135 - Seek resolution (usually corresponds to buffer size)
1136 - EODErrorCode=ENOSPC or code
1137 - Partial Read error code
1138 - Partial write error code
1139 - Nonformatted read error
1140 - Nonformatted write error
1141 - WriteProtected error
1145 - IgnoreCloseErrors=yes
1155 - FD sends unsaved file list to Director at end of job (see
1157 - File daemon should build list of files skipped, and then
1158 at end of save retry and report any errors.
1159 - Write a Storage daemon that uses pipes and
1160 standard Unix programs to write to the tape.
1162 - Need something that monitors the JCR queue and
1163 times out jobs by asking the deamons where they are.
1164 - Enhance Jmsg code to permit buffering and saving to disk.
1165 - device driver = "xxxx" for drives.
1166 - Verify from Volume
1167 - Ensure that /dev/null works
1168 - Need report class for messages. Perhaps
1169 report resource where report=group of messages
1170 - enhance scan_attrib and rename scan_jobtype, and
1171 fill in code for "since" option
1172 - Director needs a time after which the report status is sent
1173 anyway -- or better yet, a retry time for the job.
1174 - Don't reschedule a job if previous incarnation is still running.
1175 - Some way to automatically backup everything is needed????
1176 - Need a structure for pending actions:
1178 - termination status (part of buffered msgs?)
1180 Read, Write, Clean, Delete
1181 - Login to Bacula; Bacula users with different permissions:
1182 owner, group, user, quotas
1183 - Store info on each file system type (probably in the job header on tape.
1184 This could be the output of df; or perhaps some sort of /etc/mtab record.
1186 ========= ideas ===============
1187 From: "Jerry K. Schieffer" <jerry@skylinetechnology.com>
1188 To: <kern@sibbald.com>
1189 Subject: RE: [Bacula-users] future large programming jobs
1190 Date: Thu, 26 Feb 2004 11:34:54 -0600
1192 I noticed the subject thread and thought I would offer the following
1193 merely as sources of ideas, i.e. something to think about, not even as
1194 strong as a request. In my former life (before retiring) I often
1195 dealt with backups and storage management issues/products as a
1196 developer and as a consultant. I am currently migrating my personal
1197 network from amanda to bacula specifically because of the ability to
1198 cross media boundaries during storing backups.
1199 Are you familiar with the commercial product called ADSM (I think IBM
1200 now sells it under the Tivoli label)? It has a couple of interesting
1201 ideas that may apply to the following topics.
1203 1. Migration: Consider that when you need to restore a system, there
1204 may be pressure to hurry. If all the information for a single client
1205 can eventually end up on the same media (and in chronological order),
1206 the restore is facillitated by not having to search past information
1207 from other clients. ADSM has the concept of "client affinity" that
1208 may be associated with it's storage pools. It seems to me that this
1209 concept (as an optional feature) might fit in your architecture for
1212 ADSM also has the concept of defining one or more storage pools as
1213 "copy pools" (almost mirrors, but only in the sense of contents).
1214 These pools provide the ability to have duplicte data stored both
1215 onsite and offsite. The copy process can be scheduled to be handled
1216 by their storage manager during periods when there is no backup
1217 activity. Again, the migration process might be a place to consider
1218 implementing something like this.
1221 > It strikes me that it would be very nice to be able to do things
1223 > have the Job(s) backing up the machines run, and once they have all
1224 > completed, start a migration job to copy the data from disks Volumes
1226 > a tape library and then to offsite storage. Maybe this can already
1228 > done with some careful scheduling and Job prioritzation; the events
1229 > mechanism described below would probably make it very easy.
1231 This is the goal. In the first step (before events), you simply
1233 the Migration to tape later.
1235 2. Base jobs: In ADSM, each copy of each stored file is tracked in
1236 the database. Once a file (unique by path and metadata such as dates,
1237 size, ownership, etc.) is in a copy pool, no more copies are made. In
1238 other words, when you start ADSM, it begins like your concept of a
1239 base job. After that it is in the "incremental" mode. You can
1240 configure the number of "generations" of files to be retained, plus a
1241 retention date after which even old generations are purged. The
1242 database tracks the contents of media and projects the percentage of
1243 each volume that is valid. When the valid content of a volume drops
1244 below a configured percentage, the valid data are migrated to another
1245 volume and the old volume is marked as empty. Note, this requires
1246 ADSM to have an idea of the contents of a client, i.e. marking the
1247 database when an existing file was deleted, but this would solve your
1248 issue of restoring a client without restoring deleted files.
1250 This is pretty far from what bacula now does, but if you are going to
1251 rip things up for Base jobs,.....
1252 Also, the benefits of this are huge for very large shops, especially
1253 with media robots, but are a pain for shops with manual media
1257 > Base jobs sound pretty useful, but I'm not dying for them.
1259 Nobody is dying for them, but when you see what it does, you will die
1262 3. Restoring deleted files: Since I think my comments in (2) above
1263 have low probability of implementation, I'll also suggest that you
1264 could approach the issue of deleted files by a mechanism of having the
1265 fd report to the dir, a list of all files on the client for every
1266 backup job. The dir could note in the database entry for each file
1267 the date that the file was seen. Then if a restore as of date X takes
1268 place, only files that exist from before X until after X would be
1269 restored. Probably the major cost here is the extra date container in
1270 each row of the files table.
1272 Thanks for "listening". I hope some of this helps. If you want to
1273 contact me, please send me an email - I read some but not all of the
1274 mailing list traffic and might miss a reply there.
1276 Please accept my compliments for bacula. It is doing a great job for
1277 me!! I sympathize with you in the need to wrestle with excelence in
1278 execution vs. excelence in feature inclusion.
1283 ==============================
1286 - Design at hierarchial storage for Bacula. Migration and Clone.
1287 - Implement FSM (File System Modules).
1288 - Audit M_ error codes to ensure they are correct and consistent.
1289 - Add variable break characters to lex analyzer.
1290 Either a bit mask or a string of chars so that
1291 the caller can change the break characters.
1292 - Make a single T_BREAK to replace T_COMMA, etc.
1293 - Ensure that File daemon and Storage daemon can
1294 continue a save if the Director goes down (this
1295 is NOT currently the case). Must detect socket error,
1296 buffer messages for later.
1297 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
1298 - Add ability to backup to two Storage devices (two SD sessions) at
1299 the same time -- e.g. onsite, offsite.
1300 - Add the ability to consolidate old backup sets (basically do a restore
1301 to tape and appropriately update the catalog). Compress Volume sets.
1302 Might need to spool via file is only one drive is available.
1303 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
1304 someway to do so with every volume that has less than x% valid
1308 Migration: Move a backup from one Volume to another
1309 Clone: Copy a backup -- two Volumes
1311 Bacula Migration is based on Jobs (apparently Networker is file by file).
1313 Migration triggered by:
1317 Highwater mark (keep total size)
1322 ======================================================
1324 It is somewhat like a Full save becomes an incremental since
1325 the Base job (or jobs) plus other non-base files.
1327 - A Base backup is same as Full backup, just different type.
1328 - New BaseFiles table that contains:
1330 BaseJobId - Base JobId referenced for this FileId (needed ???)
1331 JobId - JobId currently running
1332 FileId - File not backed up, exists in Base Job
1333 FileIndex - FileIndex from Base Job.
1334 i.e. for each base file that exists but is not saved because
1335 it has not changed, the File daemon sends the JobId, BaseId,
1336 FileId, FileIndex back to the Director who creates the DB entry.
1337 - To initiate a Base save, the Director sends the FD
1338 the FileId, and full filename for each file in the Base.
1339 - When the FD finds a Base file, he requests the Director to
1340 send him the full File entry (stat packet plus MD5), or
1341 conversely, the FD sends it to the Director and the Director
1342 says yes or no. This can be quite rapid if the FileId is kept
1343 by the FD for each Base Filename.
1344 - It is probably better to have the comparison done by the FD
1345 despite the fact that the File entry must be sent across the
1347 - An alternative would be to send the FD the whole File entry
1348 from the start. The disadvantage is that it requires a lot of
1349 space. The advantage is that it requires less communications
1351 - The Job record must be updated to indicate that one or more
1353 - At end of Job, FD returns:
1354 1. Count of base files/bytes not written to tape (i.e. matches)
1355 2. Count of base file that were saved i.e. had changed.
1356 - No tape record would be written for a Base file that matches, in the
1357 same way that no tape record is written for Incremental jobs where
1358 the file is not saved because it is unchanged.
1359 - On a restore, all the Base file records must explicitly be
1360 found from the BaseFile tape. I.e. for each Full save that is marked
1361 to have one or more Base Jobs, search the BaseFile for all occurrences
1363 - An optimization might be to make the BaseFile have:
1369 This would avoid the need to explicitly fetch each File record for
1370 the Base job. The Base Job record will be fetched to get the
1371 VolSessionId and VolSessionTime.
1372 =========================================================
1375 ==========================================================
1377 For each Incremental job that is run, there may be files that
1378 were found but not saved because they were locked (this applies
1379 only to Windows). Such a system could send back to the Director
1380 a list of Unsaved files.
1382 - New UnSavedFiles table that contains:
1386 - Then in the next Incremental job, the list of Unsaved Files will be
1387 feed to the FD, who will ensure that they are explicitly chosen even
1388 if standard date/time check would not have selected them.
1389 =============================================================
1393 Multiple drive autochanger data: see Alan Brown
1394 mtx -f xxx unloadStorage Element 1 is Already Full(drive 0 was empty)
1395 Unloading Data Transfer Element into Storage Element 1...source Element
1396 Address 480 is Empty
1398 (drive 0 was empty and so was slot 1)
1399 > mtx -f xxx load 15 0
1400 no response, just returns to the command prompt when complete.
1401 > mtx -f xxx status Storage Changer /dev/changer:2 Drives, 60 Slots ( 2 Import/Export )
1402 Data Transfer Element 0:Full (Storage Element 15 Loaded):VolumeTag = HX001
1403 Data Transfer Element 1:Empty
1404 Storage Element 1:Empty
1405 Storage Element 2:Full :VolumeTag=HX002
1406 Storage Element 3:Full :VolumeTag=HX003
1407 Storage Element 4:Full :VolumeTag=HX004
1408 Storage Element 5:Full :VolumeTag=HX005
1409 Storage Element 6:Full :VolumeTag=HX006
1410 Storage Element 7:Full :VolumeTag=HX007
1411 Storage Element 8:Full :VolumeTag=HX008
1412 Storage Element 9:Full :VolumeTag=HX009
1413 Storage Element 10:Full :VolumeTag=HX010
1414 Storage Element 11:Empty
1415 Storage Element 12:Empty
1416 Storage Element 13:Empty
1417 Storage Element 14:Empty
1418 Storage Element 15:Empty
1419 Storage Element 16:Empty....
1420 Storage Element 28:Empty
1421 Storage Element 29:Full :VolumeTag=CLNU01L1
1422 Storage Element 30:Empty....
1423 Storage Element 57:Empty
1424 Storage Element 58:Full :VolumeTag=NEX261L2
1425 Storage Element 59 IMPORT/EXPORT:Empty
1426 Storage Element 60 IMPORT/EXPORT:Empty
1428 Unloading Data Transfer Element into Storage Element 15...done
1430 (just to verify it remembers where it came from, however it can be
1431 overrriden with mtx unload {slotnumber} to go to any storage slot.)
1433 There needs to be a table of drive # to devices somewhere - If there are
1434 multiple changers or drives there may not be a 1:1 correspondance between
1435 changer drive number and system device name - and depending on the way the
1436 drives are hooked up to scsi busses, they may not be linearly numbered
1437 from an offset point either.something like
1439 Autochanger drives = 2
1440 Autochanger drive 0 = /dev/nst1
1441 Autochanger drive 1 = /dev/nst2
1442 IMHO, it would be _safest_ to use explicit mtx unload commands at all
1443 times, not just for multidrive changers. For a 1 drive changer, that's
1449 MTX's manpage (1.2.15):
1450 unload [<slotnum>] [ <drivenum> ]
1451 Unloads media from drive <drivenum> into slot
1452 <slotnum>. If <drivenum> is omitted, defaults to
1453 drive 0 (as do all commands). If <slotnum> is
1454 omitted, defaults to the slot that the drive was
1455 loaded from. Note that there's currently no way
1456 to say 'unload drive 1's media to the slot it
1457 came from', other than to explicitly use that
1458 slot number as the destination.AB
1464 undef# camcontrol devlist
1465 <WANGTEK 51000 SCSI M74H 12B3> at scbus0 target 2 lun 0 (pass0,sa0)
1466 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 0 (pass1,sa1)
1467 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 1 (pass2)
1469 tapeinfo -f /dev/sg0 with a bad tape in drive 1:
1470 [kern@rufus mtx-1.2.17kes]$ ./tapeinfo -f /dev/sg0
1471 Product Type: Tape Drive
1473 Product ID: 'C5713A '
1475 Attached Changer: No
1476 TapeAlert[3]: Hard Error: Uncorrectable read/write error.
1477 TapeAlert[20]: Clean Now: The tape drive neads cleaning NOW.
1484 Medium Type: Not Loaded
1487 DataCompEnabled: yes
1488 DataCompCapable: yes
1489 DataDeCompEnabled: yes
1496 Handling removable disks
1498 From: Karl Cunningham <karlc@keckec.com>
1500 My backups are only to hard disk these days, in removable bays. This is my
1501 idea of how a backup to hard disk would work more smoothly. Some of these
1502 things Bacula does already, but I mention them for completeness. If others
1503 have better ways to do this, I'd like to hear about it.
1505 1. Accommodate several disks, rotated similar to how tapes are. Identified
1506 by partition volume ID or perhaps by the name of a subdirectory.
1507 2. Abort & notify the admin if the wrong disk is in the bay.
1508 3. Write backups to different subdirectories for each machine to be backed
1510 4. Volumes (files) get created as needed in the proper subdirectory, one
1512 5. When a disk is recycled, remove or zero all old backup files. This is
1513 important as the disk being recycled may be close to full. This may be
1514 better done manually since the backup files for many machines may be
1515 scattered in many subdirectories.
1520 - Make sure that all do_prompt() calls in Dir check for
1521 -1 (error) and -2 (cancel) returns.
1522 - Fix foreach_jcr() to have free_jcr() inside next().
1523 jcr=jcr_walk_start();
1524 for ( ; jcr; (jcr=jcr_walk_next(jcr)) )
1527 - A Volume taken from Scratch should take on the retention period
1529 - Correct doc for Maximum Changer Wait (and others) accepting only
1531 - Implement status that shows why a job is being held in reserve, or
1532 rather why none of the drives are suitable.
1533 - Implement a way to disable a drive (so you can use the second
1534 drive of an autochanger, and the first one will not be used or
1536 - Make sure Maximum Volumes is respected in Pools when adding
1537 Volumes (e.g. when pulling a Scratch volume).
1538 - Keep same dcr when switching device ...
1539 - Implement code that makes the Dir aware that a drive is an
1540 autochanger (so the user doesn't need to use the Autochanger = yes
1542 - Make catalog respect ACL.
1543 - Add recycle count to Media record.
1544 - Add initial write date to Media record.
1545 - Fix store_yesno to be store_bitmask.
1546 --- create_file.c.orig Fri Jul 8 12:13:05 2005
1547 +++ create_file.c Fri Jul 8 12:13:07 2005
1549 attr->ofname, be.strerror());
1552 + } else if(S_ISSOCK(attr->statp.st_mode)) {
1553 + Dmsg1(200, "Skipping socket: %s\n", attr->ofname);
1555 Dmsg1(200, "Restore node: %s\n", attr->ofname);
1556 if (mknod(attr->ofname, attr->statp.st_mode, attr->statp.st_rdev) != 0 && errno != EEXIST) {
1557 - Add true/false to conf same as yes/no
1558 - Reserve blocks other restore jobs when first cannot connect to SD.
1559 - Fix Maximum Changer Wait, Maximum Open Wait, Maximum Rewind Wait to
1560 accept time qualifiers.
1561 - Does ClientRunAfterJob fail the job on a bad return code?
1562 - Make hardlink code at line 240 of find_one.c use binary search.
1563 - Add ACL error messages in src/filed/acl.c.
projects / bacula / bacula / blob