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 - Remove queue.c code.
27 - Correct the Include syntax in the m4.xxx files in examples/conf
28 - Get Perl replacement for bregex.c
29 - Make base64.c (bin_to_base64) take a buffer length
30 argument to avoid overruns.
31 and verify that other buffers cannot overrun.
32 - 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
55 - Given all the problems with FIFOs, I think the solution is to do something a
56 little different, though I will look at the code and see if there is not some
57 simple solution (i.e. some bug that was introduced). What might be a better
58 solution would be to use a FIFO as a sort of "key" to tell Bacula to read and
59 write data to a program rather than the FIFO. For example, suppose you
64 Then, I could imagine if you backup and restore this file with a direct
65 reference as is currently done for fifos, instead, during backup Bacula will
68 /home/kern/my-fifo.backup
70 and read the data that my-fifo.backup writes to stdout. For restore, Bacula
73 /home/kern/my-fifo.restore
75 and send the data backed up to stdout. These programs can either be an
76 executable or a shell script and they need only read/write to stdin/stdout.
78 I think this would give a lot of flexibility to the user without making any
79 significant changes to Bacula.
84 select FilenameId from Filename where Name='';
85 # Get list of all directories referenced in a Backup.
86 select Path.Path from Path,File where File.JobId=nnn and
87 File.FilenameId=(FilenameId-from-above) and File.PathId=Path.PathId
88 order by Path.Path ASC;
90 - Look into using Dart for testing
91 http://public.kitware.com/Dart/HTML/Index.shtml
93 - Look into replacing autotools with cmake
94 http://www.cmake.org/HTML/Index.html
96 === Migration from David ===
100 Name = "<poolname>-migrate"
104 Migration Selection Type = LowestUtil | OldestVol | PoolOccupancy |
105 Client | PoolResidence | Volume | JobName | SQLquery
106 Migration Selection Pattern = "regexp"
107 Next Pool = <override>
110 There should be no need for a Level (migration is always Full, since you
111 don't calculate differential/incremental differences for migration),
112 Storage should be determined by the volume types in the pool, and Client
113 is really a selection issue. Migration should always occur to the
114 NextPool defined in the pool definition. If no nextpool is defined, the
115 job should end with a reason of "no place to go". If Next Pool statement
116 is present, we override the check in the pool definition and use the
119 Here's how I'd define Migration Selection Types:
122 Client -- Migrate data from selected client only. Migration Selection
123 Pattern regexp provides pattern to select client names, eg ^FS00* makes
124 all client names starting with FS00 eligible for migration.
126 Jobname -- Migration all jobs matching name. Migration Selection Pattern
127 regexp provides pattern to select jobnames existing in pool.
129 Volume -- Migrate all data on specified volumes. Migration Selection
130 Pattern regexp provides selection criteria for volumes to be migrated.
131 Volumes must exist in pool to be eligible for migration.
135 LowestUtil -- Identify the volume in the pool with the least data on it
136 and empty it. No Migration Selection Pattern required.
138 OldestVol -- Identify the LRU volume with data written, and empty it. No
139 Migration Selection Pattern required.
141 PoolOccupancy -- if pool occupancy exceeds <highmig>, migrate volumes
142 (starting with most full volumes) until pool occupancy drops below
143 <lowmig>. Pool highmig and lowmig values are in pool definition, no
144 Migration Selection Pattern required.
148 SQLQuery -- Migrate all jobuids returned by the supplied SQL query.
149 Migration Selection Pattern contains SQL query to execute; should return
150 a list of 1 or more jobuids to migrate.
152 PoolResidence -- Migrate data sitting in pool for longer than
153 PoolResidence value in pool definition. Migration Selection Pattern
154 optional; if specified, override value in pool definition (value in
158 [ possibly a Python event -- kes ]
160 - Mount on an Autochanger with no tape in the drive causes:
161 Automatically selected Storage: LTO-changer
162 Enter autochanger drive[0]: 0
163 3301 Issuing autochanger "loaded drive 0" command.
164 3302 Autochanger "loaded drive 0", result: nothing loaded.
165 3301 Issuing autochanger "loaded drive 0" command.
166 3302 Autochanger "loaded drive 0", result: nothing loaded.
167 3902 Cannot mount Volume on Storage Device "LTO-Drive1" (/dev/nst0) because:
168 Couldn't rewind device "LTO-Drive1" (/dev/nst0): ERR=dev.c:678 Rewind error on "LTO-Drive1" (/dev/nst0). ERR=No medium found.
169 3905 Device "LTO-Drive1" (/dev/nst0) open but no Bacula volume is mounted.
170 If this is not a blank tape, try unmounting and remounting the Volume.
171 - If Drive 0 is blocked, and drive 1 is set "Autoselect=no", drive 1 will
173 - Autochanger did not change volumes.
174 select * from Storage;
175 +-----------+-------------+-------------+
176 | StorageId | Name | AutoChanger |
177 +-----------+-------------+-------------+
178 | 1 | LTO-changer | 0 |
179 +-----------+-------------+-------------+
180 05-May 03:50 roxie-sd: 3302 Autochanger "loaded drive 0", result is Slot 11.
181 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Warning: Director wanted Volume "LT
182 Current Volume "LT0-002" not acceptable because:
183 1997 Volume "LT0-002" not in catalog.
184 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Error: Autochanger Volume "LT0-002"
185 Setting InChanger to zero in catalog.
186 05-May 03:50 roxie-dir: Tibs.2006-05-05_03.05.02 Error: Unable to get Media record
188 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: Error getting Volume i
189 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: Job 530 canceled.
190 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: spool.c:249 Fatal appe
191 05-May 03:49 Tibs: Tibs.2006-05-05_03.05.02 Fatal error: c:\cygwin\home\kern\bacula
200 FirstWritten: 2006-05-05 03:11:54
201 LastWritten: 2006-05-05 03:50:23
202 LabelDate: 2005-12-26 16:52:40
213 VolRetention: 31,536,000
225 Note VolStatus is blank!!!!!
232 FirstWritten: 0000-00-00 00:00:00
233 LastWritten: 0000-00-00 00:00:00
234 LabelDate: 2005-12-26 16:52:40
245 VolRetention: 31,536,000
258 Automatically selected Storage: LTO-changer
259 Enter autochanger drive[0]: 0
260 3301 Issuing autochanger "loaded drive 0" command.
261 3302 Autochanger "loaded drive 0", result: nothing loaded.
262 3301 Issuing autochanger "loaded drive 0" command.
263 3302 Autochanger "loaded drive 0", result: nothing loaded.
264 3902 Cannot mount Volume on Storage Device "LTO-Drive1" (/dev/nst0) because:
265 Couldn't rewind device "LTO-Drive1" (/dev/nst0): ERR=dev.c:678 Rewind error on "LTO-Drive1" (/dev/nst0). ERR=No medium found.
267 3905 Device "LTO-Drive1" (/dev/nst0) open but no Bacula volume is mounted.
268 If this is not a blank tape, try unmounting and remounting the Volume.
270 - Add VolumeState (enable, disable, archive)
271 - Add VolumeLock to prevent all but lock holder (SD) from updating
272 the Volume data (with the exception of VolumeState).
273 - The btape fill command does not seem to use the Autochanger
274 - Make Windows installer default to system disk drive.
275 - Look at using ioctl(FIOBMAP, ...) on Linux, and
276 DeviceIoControl(..., FSCTL_QUERY_ALLOCATED_RANGES, ...) on
277 Win32 for sparse files.
278 http://www.flexhex.com/docs/articles/sparse-files.phtml
279 http://www.informatik.uni-frankfurt.de/~loizides/reiserfs/fibmap.html
280 - Directive: at <event> "command"
281 - Command: pycmd "command" generates "command" event. How to
282 attach to a specific job?
283 - Integrate Christopher's St. Bernard code.
284 - run_cmd() returns int should return JobId_t
285 - get_next_jobid_from_list() returns int should return JobId_t
286 - Document export LDFLAGS=-L/usr/lib64
287 - Don't attempt to restore from "Disabled" Volumes.
288 - Network error on Win32 should set Win32 error code.
289 - What happens when you rename a Disk Volume?
290 - Job retention period in a Pool (and hence Volume). The job would
292 - Detect resource deadlock in Migrate when same job wants to read
293 and write the same device.
294 - Queue warning/error messages during restore so that they
295 are reported at the end of the report rather than being
296 hidden in the file listing ...
297 - Look at -D_FORTIFY_SOURCE=2
298 - Add Win32 FileSet definition somewhere
299 - Look at fixing restore status stats in SD.
300 - Make selection of Database used in restore correspond to
302 - Look at using ioctl(FIMAP) and FIGETBSZ for sparse files.
303 http://www.informatik.uni-frankfurt.de/~loizides/reiserfs/fibmap.html
304 - Implement a mode that says when a hard read error is
305 encountered, read many times (as it currently does), and if the
306 block cannot be read, skip to the next block, and try again. If
307 that fails, skip to the next file and try again, ...
309 create table LevelType (LevelType binary(1), LevelTypeLong tinyblob);
310 insert into LevelType (LevelType,LevelTypeLong) values
314 - Add ACL to restore only to original location.
315 - Show files/second in client status output.
316 - Add a recursive mark command (rmark) to restore.
317 - "Minimum Job Interval = nnn" sets minimum interval between Jobs
318 of the same level and does not permit multiple simultaneous
319 running of that Job (i.e. lets any previous invocation finish
320 before doing Interval testing).
321 - Look at simplifying File exclusions.
322 - New directive "Delete purged Volumes"
323 - new pool XXX with ScratchPoolId = MyScratchPool's PoolId and
324 let it fill itself, and RecyclePoolId = XXX's PoolId so I can
325 see if it become stable and I just have to supervise
327 - If I want to remove this pool, I set RecyclePoolId = MyScratchPool's
328 PoolId, and when it is empty remove it.
329 - Figure out how to recycle Scratch volumes back to the Scratch Pool.
331 - Allow Check Labels to be used with Bacula labels.
332 - "Resuming" a failed backup (lost line for example) by using the
333 failed backup as a sort of "base" job.
335 - Email to the user when the tape is about to need changing x
336 days before it needs changing.
337 - Command to show next tape that will be used for a job even
338 if the job is not scheduled.
339 - From: Arunav Mandal <amandal@trolltech.com>
340 1. When jobs are running and bacula for some reason crashes or if I do a
341 restart it remembers and jobs it was running before it crashed or restarted
342 as of now I loose all jobs if I restart it.
344 2. When spooling and in the midway if client is disconnected for instance a
345 laptop bacula completely discard the spool. It will be nice if it can write
346 that spool to tape so there will be some backups for that client if not all.
348 3. We have around 150 clients machines it will be nice to have a option to
349 upgrade all the client machines bacula version automatically.
351 4. Atleast one connection should be reserved for the bconsole so at heavy load
352 I should connect to the director via bconsole which at sometimes I can't
354 5. Another most important feature that is missing, say at 10am I manually
355 started backup of client abc and it was a full backup since client abc has
356 no backup history and at 10.30am bacula again automatically started backup of
357 client abc as that was in the schedule. So now we have 2 multiple Full
358 backups of the same client and if we again try to start a full backup of
359 client backup abc bacula won't complain. That should be fixed.
361 - Fix bpipe.c so that it does not modify results pointer.
362 ***FIXME*** calling sequence should be changed.
363 - For Windows disaster recovery see http://unattended.sf.net/
364 - regardless of the retention period, Bacula will not prune the
365 last Full, Diff, or Inc File data until a month after the
366 retention period for the last Full backup that was done.
367 - update volume=xxx --- add status=Full
368 - Remove old spool files on startup.
369 - Exclude SD spool/working directory.
370 - Refuse to prune last valid Full backup. Same goes for Catalog.
372 - Make a callback when Rerun failed levels is called.
373 - Give Python program access to Scheduled jobs.
374 - Add setting Volume State via Python.
375 - Python script to save with Python, not save, save with Bacula.
376 - Python script to do backup.
378 - Change the Priority, Client, Storage, JobStatus (error)
379 at the start of a job.
380 - Why is SpoolDirectory = /home/bacula/spool; not reported
381 as an error when writing a DVD?
382 - Make bootstrap file handle multiple MediaTypes (SD)
383 - Remove all old Device resource code in Dir and code to pass it
384 back in SD -- better, rework it to pass back device statistics.
385 - Check locking of resources -- be sure to lock devices where previously
386 resources were locked.
387 - The last part is left in the spool dir.
390 - In restore don't compare byte count on a raw device -- directory
391 entry does not contain bytes.
392 - To mark files as deleted, run essentially a Verify to disk, and
393 when a file is found missing (MarkId != JobId), then create
394 a new File record with FileIndex == -1. This could be done
395 by the FD at the same time as the backup.
399 MA = (last_MA * 3 + rate) / 4
400 rate = (bytes - last_bytes) / (runtime - last_runtime)
401 - Max Vols limit in Pool off by one?
402 - Implement Files/Bytes,... stats for restore job.
403 - Implement Total Bytes Written, ... for restore job.
404 - Despool attributes simultaneously with data in a separate
405 thread, rejoined at end of data spooling.
406 - 7. Implement new Console commands to allow offlining/reserving drives,
407 and possibly manipulating the autochanger (much asked for).
408 - Add start/end date editing in messages (%t %T, %e?) ...
409 - Add ClientDefs similar to JobDefs.
410 - Print more info when bextract -p accepts a bad block.
411 - Fix FD JobType to be set before RunBeforeJob in FD.
412 - Look at adding full Volume and Pool information to a Volume
413 label so that bscan can get *all* the info.
414 - If the user puts "Purge Oldest Volume = yes" or "Recycle Oldest Volume = yes"
415 and there is only one volume in the pool, refuse to do it -- otherwise
416 he fills the Volume, then immediately starts reusing it.
417 - Implement copies and stripes.
418 - Add history file to console.
419 - Each file on tape creates a JobMedia record. Peter has 4 million
420 files spread over 10000 tape files and four tapes. A restore takes
421 16 hours to build the restore list.
422 - Add and option to check if the file size changed during backup.
423 - Make sure SD deletes spool files on error exit.
424 - Delete old spool files when SD starts.
425 - When labeling tapes, if you enter 000026, Bacula uses
426 the tape index rather than the Volume name 000026.
427 - Add offline tape command to Bacula console.
429 Enter MediaId or Volume name: 32
430 Enter new Volume name: DLT-20Dec04
431 Automatically selected Pool: Default
432 Connecting to Storage daemon DLTDrive at 192.168.68.104:9103 ...
433 Sending relabel command from "DLT-28Jun03" to "DLT-20Dec04" ...
434 block.c:552 Write error at 0:0 on device /dev/nst0. ERR=Bad file descriptor.
435 Error writing final EOF to tape. This tape may not be readable.
436 dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
437 askdir.c:219 NULL Volume name. This shouldn't happen!!!
438 3912 Failed to label Volume: ERR=dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
439 Label command failed for Volume DLT-20Dec04.
440 Do not forget to mount the drive!!!
441 - Bug: if a job is manually scheduled to run later, it does not appear
442 in any status report and cannot be cancelled.
444 ==== Keeping track of deleted files ====
445 My "trick" for keeping track of deletions is the following.
446 Assuming the user turns on this option, after all the files
447 have been backed up, but before the job has terminated, the
448 FD will make a pass through all the files and send their
449 names to the DIR (*exactly* the same as what a Verify job
450 currently does). This will probably be done at the same
451 time the files are being sent to the SD avoiding a second
452 pass. The DIR will then compare that to what is stored in
453 the catalog. Any files in the catalog but not in what the
454 FD sent will receive a catalog File entry that indicates
455 that at that point in time the file was deleted.
457 During a restore, any file initially picked up by some
458 backup (Full, ...) then subsequently having a File entry
459 marked "delete" will be removed from the tree, so will not
460 be restored. If a file with the same name is later OK it
461 will be inserted in the tree -- this already happens. All
462 will be consistent except for possible changes during the
465 Since I'm on the subject, some of you may be wondering what
466 the utility of the in memory tree is if you are going to
467 restore everything (at least it comes up from time to time
468 on the list). Well, it is still *very* useful because it
469 allows only the last item found for a particular filename
470 (full path) to be entered into the tree, and thus if a file
471 is backed up 10 times, only the last copy will be restored.
472 I recently (last Friday) restored a complete directory, and
473 the Full and all the Differential and Incremental backups
474 spanned 3 Volumes. The first Volume was not even mounted
475 because all the files had been updated and hence backed up
476 since the Full backup was made. In this case, the tree
477 saved me a *lot* of time.
479 Make sure this information is stored on the tape too so
480 that it can be restored directly from the tape.
482 Comments from Martin Simmons (I think they are all covered):
483 Ok, that should cover the basics. There are few issues though:
485 - Restore will depend on the catalog. I think it is better to include the
486 extra data in the backup as well, so it can be seen by bscan and bextract.
488 - I'm not sure if it will preserve multiple hard links to the same inode. Or
489 maybe adding or removing links will cause the data to be dumped again?
491 - I'm not sure if it will handle renamed directories. Possibly it will work
492 by dumping the whole tree under a renamed directory?
494 - It remains to be seen how the backup performance of the DIR's will be
495 affected when comparing the catalog for a large filesystem.
499 How about introducing a Type = MgmtPolicy job type? That job type would
500 be responsible for scanning the Bacula environment looking for specific
501 conditions, and submitting the appropriate jobs for implementing said
505 Name = "Migration-Policy"
507 Policy Selection Job Type = Migrate
508 Scope = "<keyword> <operator> <regexp>"
509 Threshold = "<keyword> <operator> <regexp>"
510 Job Template = <template-name>
513 Where <keyword> is any legal job keyword, <operator> is a comparison
514 operator (=,<,>,!=, logical operators AND/OR/NOT) and <regexp> is a
515 appropriate regexp. I could see an argument for Scope and Threshold
516 being SQL queries if we want to support full flexibility. The
517 Migration-Policy job would then get scheduled as frequently as a site
518 felt necessary (suggested default: every 15 minutes).
523 Name = "Migration-Policy"
525 Policy Selection Job Type = Migration
527 Threshold = "Migration Selection Type = LowestUtil"
528 Job Template = "MigrationTemplate"
531 would select all pools for examination and generate a job based on
532 MigrationTemplate to automatically select the volume with the lowest
533 usage and migrate it's contents to the nextpool defined for that pool.
535 This policy abstraction would be really handy for adjusting the behavior
536 of Bacula according to site-selectable criteria (one thing that pops
537 into mind is Amanda's ability to automatically adjust backup levels
538 depending on various criteria).
544 - Add Pool/Storage override regression test.
545 - Add delete JobId to regression.
546 - Add a regression test for dbcheck.
547 - New test to add bscan to four-concurrent-jobs regression,
548 i.e. after the four-concurrent jobs zap the
549 database as is done in the bscan-test, then use bscan to
550 restore the database, do a restore and compare with the
552 - Add restore of specific JobId to regression (item 3
553 on the restore prompt)
554 - Add IPv6 to regression
555 - Add database test to regression. Test each function like delete,
558 - AntiVir can slow down backups on Win32 systems.
559 - Win32 systems with FAT32 can be much slower than NTFS for
560 more than 1000 files per directory.
564 - A HOLD command to stop all jobs from starting.
565 - A PAUSE command to pause all running jobs ==> release the
567 - Media Type = LTO,LTO-2,LTO-3
568 Media Type Read = LTO,LTO2,LTO3
569 Media Type Write = LTO2, LTO3
571 === From Carsten Menke <bootsy52@gmx.net>
573 Following is a list of what I think in the situations where I'm faced with,
574 could be a usefull enhancement to bacula, which I'm certain other users will
575 benefit from as well.
577 1. NextJob/NextJobs Directive within a Job Resource in the form of
578 NextJobs = job1,job2.
581 I currently solved the problem with running multiple jobs each after each
582 by setting the Max Wait Time for a job to 8 hours, and give
583 the jobs different Priorities. However, there scenarios where
584 1 Job is directly depending on another job, so if the former job fails,
585 the job after it needn't to be run
586 while maybe other jobs should run despite of that
589 A Backup Job and a Verify job, if the backup job fails there is no need to run
590 the verify job, as the backup job already failed. However, one may like
591 to backup the Catalog to disk despite of that the main backup job failed.
594 I see that this is related to the Event Handlers which are on the ToDo
595 list, also it is maybe a good idea to check for the return value and
596 execute different actions based on the return value
599 3. offline capability to bconsole
602 Currently I use a script which I execute within the last Job via the
603 RunAfterJob Directive, to release and eject the tape.
604 So I have to call bconsole "release=Storage-Name" and afterwards
605 mt -f /dev/nst0 eject to get the tape out.
607 If I have multiple Storage Devices, than these may not be /dev/nst0 and
608 I have to modify the script or call it with parameters etc.
609 This would actually not be needed, as everything is already defined
610 in bacula-sd.conf and if I can invoke bconsole with the
611 storage name via $1 in the script than I'm done and information is
614 4. %s for Storage Name added to the chars being substituted in "RunAfterJob"
618 For the reason mentioned in 3. to have the ability to call a
619 script with /scripts/foobar %s and in the script use $1
620 to pass the Storage Name to bconsole
622 5. Setting Volume State within a Job Resource
625 Instead of using "Maximum Volume Jobs" in the Pool Resource,
626 I would have the possibilty to define
627 in a Job Resource that after this certain job is run, the Volume State
628 should be set to "Volume State = Used", this give more flexibility (IMHO).
630 6. Localization of Bacula Messages
633 Unfortunatley many,many people I work with don't speak english very well.
634 So if at least the Reporting messages would be localized then they
635 would understand that they have to change the tape,etc. etc.
637 I volunteer to do the german translations, and if I can convince my wife also
638 french and Morre (western african language).
640 7. OK, this is evil, probably bound to security risks and maybe not possible
641 due to the design of bacula.
643 Implementation of Backtics ( `command` ) for shell comand execution to
644 the "Label Format" Directive.
648 Currently I have defined BACULA_DAY_OF_WEEK="day1|day2..." resulting in
649 Label Format = "HolyBackup-${BACULA_DAY_OF_WEEK[${WeekDay}]}". If I could
650 use backticks than I could use "Label Format = HolyBackup-`date +%A` to have
651 the localized name for the day of the week appended to the
652 format string. Then I have the tape labeled automatically with weekday
653 name in the correct language.
655 - Yes, that is surely the case. I probably should turn those into Warning
656 errors. In addition, you just made me think that it might not be bad to
657 add an option to check the file size after backing up the file and
658 report if it changes. This would be done as an option because it would
661 Kern, good idea. If you do do that, mention in the output: file
662 shrunk, or file expanded, just to make it obvious to the user
663 (without having to the refer to file size), just how the file size
666 Would this option be for all file, or just one file? Or a fileset?
667 - Make output from status use html table tags for nicely
668 presenting in a browser.
669 - Can one write tapes faster with 8192 byte block sizes?
670 - Document security problems with the same password for everyone in
671 rpm and Win32 releases.
672 - Browse generations of files.
673 - I've seen an error when my catalog's File table fills up. I
674 then have to recreate the File table with a larger maximum row
675 size. Relevant information is at
676 http://dev.mysql.com/doc/mysql/en/Full_table.html ; I think the
677 "Installing and Configuring MySQL" chapter should talk a bit
678 about this potential problem, and recommend a solution.
679 - For Solaris must use POSIX awk.
680 - Want speed of writing to tape while despooling.
681 - Supported autochanger:
689 Wangtek 6525ES (SCSI-1 QIC drive, 525MB), under Linux 2.4.something,
690 bacula 1.36.0/1 works with blocksize 16k INSIDE bacula-sd.conf.
691 - Add regex from http://www.pcre.org to Bacula for Win32.
692 - Use only shell tools no make in CDROM package.
693 - Include within include does it work?
694 - Implement a Pool of type Cleaning?
695 - Implement VolReadTime and VolWriteTime in SD
696 - Modify Backing up Your Database to include a bootstrap file.
697 - Think about making certain database errors fatal.
698 - Look at correcting the time jump in the scheduler for daylight
699 savings time changes.
700 - Add a "real" timer to network connections.
701 - Promote to Full = Time period
702 - Check dates entered by user for correctness (month/day/... ranges)
703 - Compress restore Volume listing by date and first file.
704 - Look at patches/bacula_db.b2z postgresql that loops during restore.
706 - Perhaps add read/write programs and/or plugins to FileSets.
707 - How to handle backing up portables ...
708 - Add some sort of guaranteed Interval for upgrading jobs.
709 - Can we write the state file after every job terminates? On Win32
710 the system crashes and the state file is not updated.
713 Documentation to do: (any release a little bit at a time)
714 - Doc to do unmount before removing magazine.
715 - Alternative to static linking "ldd prog" save all binaries listed,
716 restore them and point LD_LIBRARY_PATH to them.
717 - Document add "</dev/null >/dev/null 2>&1" to the bacula-fd command line
718 - Document query file format.
719 - Add more documentation for bsr files.
720 - Document problems with Verify and pruning.
721 - Document how to use multiple databases.
722 - VXA drives have a "cleaning required"
723 indicator, but Exabyte recommends preventive cleaning after every 75
726 In this context, it should be noted that Exabyte has a command-line
727 vxatool utility available for free download. (The current version is
728 vxatool-3.72.) It can get diagnostic info, read, write and erase tapes,
729 test the drive, unload tapes, change drive settings, flash new firmware,
731 Of particular interest in this context is that vxatool <device> -i will
732 report, among other details, the time since last cleaning in tape motion
733 minutes. This information can be retrieved (and settings changed, for
734 that matter) through the generic-SCSI device even when Bacula has the
735 regular tape device locked. (Needless to say, I don't recommend
736 changing tape settings while a job is running.)
737 - Lookup HP cleaning recommendations.
738 - Lookup HP tape replacement recommendations (see trouble shooting autochanger)
739 - Document doing table repair
742 ===================================
743 - Add macro expansions in JobDefs.
744 Run Before Job = "SomeFile %{Level} %{Client}"
745 Write Bootstrap="/some/dir/%{JobName}_%{Client}.bsr"
746 - Use non-blocking network I/O but if no data is available, use
748 - Use gather write() for network I/O.
749 - Autorestart on crash.
750 - Add bandwidth limiting.
751 - Add acks every once and a while from the SD to keep
752 the line from timing out.
753 - When an error in input occurs and conio beeps, you can back
754 up through the prompt.
755 - Detect fixed tape block mode during positioning by looking at
756 block numbers in btape "test". Possibly adjust in Bacula.
757 - Fix list volumes to output volume retention in some other
758 units, perhaps via a directive.
759 - If opening a tape in read/write mode fails attempt to open
760 it in read-only mode, and mark the tape for read only.
761 - Allow Simultaneous Priorities = yes => run up to Max concurrent jobs even
762 with multiple priorities.
763 - If you use restore replace=never, the directory attributes for
764 non-existent directories will not be restored properly.
766 - see lzma401.zip in others directory for new compression
768 - Minimal autochanger handling in Bacula and in btape.
769 - Look into how tar does not save sockets and the possiblity of
770 not saving them in Bacula (Martin Simmons reported this).
771 - Fix restore jobs so that multiple jobs can run if they
772 are not using the same tape(s).
773 - Allow the user to select JobType for manual pruning/purging.
774 - bscan does not put first of two volumes back with all info in
776 - Implement the FreeBSD nodump flag in chflags.
777 - Figure out how to make named console messages go only to that
778 console and to the non-restricted console (new console class?).
779 - Make restricted console prompt for password if *ask* is set or
780 perhaps if password is undefined.
781 - Implement "from ISO-date/time every x hours/days/weeks/months" in
784 ==== from Marc Schoechlin
785 - the help-command should be more verbose
786 (it should explain the paramters of the different
788 -> its time-comsuming to consult the manual anytime
789 you need a special parameter
790 -> maybe its more easy to maintain this, if the
791 descriptions of that commands are outsourced to
793 - the cd-command should allow complete paths
794 i.e. cd /foo/bar/foo/bar
795 -> if a customer mails me the path to a certain file,
796 its faster to enter the specified directory
797 - if the password is not configured in bconsole.conf
798 you should be asked for it.
799 -> sometimes you like to do restore on a customer-machine
800 which shouldnt know the password for bacula.
801 -> adding the password to the file favours admins
802 to forget to remove the password after usage
804 the protection of that file is less important
805 - long-listed-output of commands should be scrollable
806 like the unix more/less-command does
807 -> if someone runs 200 and more machines, the lists could
808 be a little long and complex
809 - command-output should be shown column by column
810 to reduce scrolling and to increase clarity
812 - lsmark should list the selected files with full
814 - wildcards for selecting and file and directories would be nice
815 - any actions should be interuptable with STRG+C
816 - command-expansion would be pretty cool
818 - When the replace Never option is set, new directory permissions
819 are not restored. See bug 213. To fix this requires creating a
820 list of newly restored directories so that those directory
821 permissions *can* be restored.
822 - Compaction of Disk space by "migrating" Volumes that have pruned
823 Jobs (what criteria? size, #jobs, time).
824 - Add prune all command
825 - Document fact that purge can destroy a part of a restore by purging
826 one volume while others remain valid -- perhaps mark Jobs.
827 - Add multiple-media-types.txt
828 - look at mxt-changer.html
829 - Make ? do a help command (no return needed).
830 - Implement restore directory.
831 - Document streams and how to implement them.
832 - Try not to re-backup a file if a new hard link is added.
833 - Add feature to backup hard links only, but not the data.
834 - Fix stream handling to be simpler.
835 - Add Priority and Bootstrap to Run a Job.
836 - Eliminate Restore "Run Restore Job" prompt by allowing new "run command
838 - Remove View FileSet button from Run a Job dialog.
839 - Handle prompt for restore job at end of Restore command.
840 - Add display of total selected files to Restore window.
841 - Add tree pane to left of window.
842 - Add progress meter.
843 - Max wait time or max run time causes seg fault -- see runtime-bug.txt
844 - Document writing to a CD/DVD with Bacula.
845 - Add a "base" package to the window installer for pthreadsVCE.dll
846 which is needed by all packages.
847 - Add message to user to check for fixed block size when the forward
848 space test fails in btape.
849 - When unmarking a directory check if all files below are unmarked and
850 then remove the + flag -- in the restore tree.
851 - Possibly implement: Action = Unmount Device="TapeDrive1" in Admin jobs.
852 - Setup lrrd graphs: (http://www.linpro.no/projects/lrrd/) Mike Acar.
853 - Revisit the question of multiple Volumes (disk) on a single device.
854 - Add a block copy option to bcopy.
855 - Investigate adding Mac Resource Forks.
856 - Finish work on Gnome restore GUI.
857 - Fix "llist jobid=xx" where no fileset or client exists.
858 - For each job type (Admin, Restore, ...) require only the really necessary
859 fields.- Pass Director resource name as an option to the Console.
860 - Add a "batch" mode to the Console (no unsolicited queries, ...).
861 - Add a .list all files in the restore tree (probably also a list all files)
862 Do both a long and short form.
863 - Allow browsing the catalog to see all versions of a file (with
864 stat data on each file).
865 - Restore attributes of directory if replace=never set but directory
867 - Use SHA1 on authentication if possible.
868 - See comtest-xxx.zip for Windows code to talk to USB.
869 - Add John's appended files:
870 Appended = { /files/server/logs/http/*log }
871 and such files would be treated as follows.On a FULL backup, they would
872 be backed up like any other file.On an INCREMENTAL backup, where a
873 previous INCREMENTAL or FULL was already in thecatalogue and the length
874 of the file wasgreater than the length of the last backup, only thedata
875 added since the last backup will be dumped.On an INCREMENTAL backup, if
876 the length of the file is less than thelength of the file with the same
877 name last backed up, the completefile is dumped.On Windows systems, with
878 creation date of files, we can be evensmarter about this and not count
879 entirely upon the length.On a restore, the full and all incrementals
880 since it will beapplied in sequence to restore the file.
881 - Check new HAVE_WIN32 open bits.
882 - Check if the tape has moved before writing.
883 - Handling removable disks -- see below:
884 - Keep track of tape use time, and report when cleaning is necessary.
885 - Add FromClient and ToClient keywords on restore command (or
886 BackupClient RestoreClient).
887 - Implement a JobSet, which groups any number of jobs. If the
888 JobSet is started, all the jobs are started together.
889 Allow Pool, Level, and Schedule overrides.
890 - Enhance cancel to timeout BSOCK packets after a specific delay.
891 - Do scheduling by UTC using gmtime_r() in run_conf, scheduler, and
892 ua_status.!!! Thanks to Alan Brown for this tip.
893 - Look at updating Volume Jobs so that Max Volume Jobs = 1 will work
894 correctly for multiple simultaneous jobs.
895 - Correct code so that FileSet MD5 is calculated for < and | filename
897 - Implement the Media record flag that indicates that the Volume does disk
899 - Implement VolAddr, which is used when Volume is addressed like a disk,
900 and form it from VolFile and VolBlock.
901 - Make multiple restore jobs for multiple media types specifying
902 the proper storage type.
903 - Fix fast block rejection (stored/read_record.c:118). It passes a null
904 pointer (rec) to try_repositioning().
905 - Look at extracting Win data from BackupRead.
906 - Implement RestoreJobRetention? Maybe better "JobRetention" in a Job,
907 which would take precidence over the Catalog "JobRetention".
908 - Implement Label Format in Add and Label console commands.
909 - Possibly up network buffers to 65K. Put on variable.
910 - Put email tape request delays on one or more variables. User wants
911 to cancel the job after a certain time interval. Maximum Mount Wait?
912 - Job, Client, Device, Pool, or Volume?
913 Is it possible to make this a directive which is *optional* in multiple
914 resources, like Level? If so, I think I'd make it an optional directive
915 in Job, Client, and Pool, with precedence such that Job overrides Client
916 which in turn overrides Pool.
918 - New Storage specifications:
919 - Want to write to multiple storage devices simultaneously
920 - Want to write to multiple storage devices sequentially (in one job)
921 - Want to read/write simultaneously
922 - Key is MediaType -- it must match
924 Passed to SD as a sort of BSR record called Storage Specification
928 MediaType -> Next MediaType
930 Device -> Next Device
932 Allow multiple Storage specifications
940 Allow Multiple Pool specifications (note, Pool currently
942 Allow Multiple MediaType specifications in Dir conf
943 Allow Multiple Device specifications in Dir conf
944 Perhaps keep this in a single SSR
945 Tie a Volume to a specific device by using a MediaType that
946 is contained in only one device.
947 In SD allow Device to have Multiple MediaTypes
949 - Ideas from Jerry Scharf:
950 First let's point out some big pluses that bacula has for this
952 more importantly it's active. Thank you so much for that
953 even more important, it's not flaky
954 it has an open access catalog, opening many possibilities
955 it's pushing toward heterogeneous systems capability
957 Macintosh file client
958 macs are an interesting niche, but I fear a server is a rathole
959 working bare iron recovery for windows
960 the option for inc/diff backups not reset on fileset revision
961 a) use both change and inode update time against base time
962 b) do the full catalog check (expensive but accurate)
963 sizing guide (how much system is needed to back up N systems/files)
964 consultants on using bacula in building a disaster recovery system
966 or how to get at fancy things that one could do with bacula
967 logwatch code for bacula logs (or similar)
968 linux distro inclusion of bacula (brings good and bad, but necessary)
969 win2k/XP server capability (icky but you asked)
970 support for Oracle database ??
972 - Look at adding SQL server and Exchange support for Windows.
973 - Make dev->file and dev->block_num signed integers so that -1 can
974 be an invalid value which happens with BSR.
975 - Create VolAddr for disk files in place of VolFile and VolBlock. This
976 is needed to properly specify ranges.
977 - Add progress of files/bytes to SD and FD.
978 - Print warning message if FileId > 4 billion
979 - do a "messages" before the first prompt in Console
980 - Client does not show busy during Estimate command.
981 - Implement Console mtx commands.
982 - Implement a Mount Command and an Unmount Command where
983 the users could specify a system command to be performed
984 to do the mount, after which Bacula could attempt to
985 read the device. This is for Removeable media such as a CDROM.
986 - Most likely, this mount command would be invoked explicitly
987 by the user using the current Console "mount" and "unmount"
988 commands -- the Storage Daemon would do the right thing
989 depending on the exact nature of the device.
990 - As with tape drives, when Bacula wanted a new removable
991 disk mounted, it would unmount the old one, and send a message
992 to the user, who would then use "mount" as described above
993 once he had actually inserted the disk.
994 - Implement dump/print label to UA
995 - Spool to disk only when the tape is full, then when a tape is hung move
997 - bextract is sending everything to the log file ****FIXME****
998 - Allow multiple Storage specifications (or multiple names on
999 a single Storage specification) in the Job record. Thus a job
1000 can be backed up to a number of storage devices.
1001 - Implement some way for the File daemon to contact the Director
1002 to start a job or pass its DHCP obtained IP number.
1003 - Implement a query tape prompt/replace feature for a console
1004 - Copy console @ code to gnome2-console
1005 - Make AES the only encryption algorithm see
1006 http://csrc.nist.gov/CryptoToolkit/aes/). It's
1007 an officially adopted standard, has survived peer
1008 review, and provides keys up to 256 bits.
1009 - Take a careful look at SetACL http://setacl.sourceforge.net
1010 - Make tree walk routines like cd, ls, ... more user friendly
1011 by handling spaces better.
1012 - Make sure that Bacula rechecks the tape after the 20 min wait.
1013 - Set IO_NOWAIT on Bacula TCP/IP packets.
1014 - Try doing a raw partition backup and restore by mounting a
1016 - From Lars Kellers:
1017 Yes, it would allow to highly automatic the request for new tapes. If a
1018 tape is empty, bacula reads the barcodes (native or simulated), and if
1019 an unused tape is found, it runs the label command with all the
1020 necessary parameters.
1022 By the way can bacula automatically "move" an empty/purged volume say
1023 in the "short" pool to the "long" pool if this pool runs out of volume
1025 - What to do about "list files job=xxx".
1026 - Get and test MySQL 4.0
1027 - Look at how fuser works and /proc/PID/fd that is how Nic found the
1028 file descriptor leak in Bacula.
1029 - Implement WrapCounters in Counters.
1030 - Add heartbeat from FD to SD if hb interval expires.
1031 - Can we dynamically change FileSets?
1032 - If pool specified to label command and Label Format is specified,
1033 automatically generate the Volume name.
1034 - Why can't SQL do the filename sort for restore?
1035 - Add ExhautiveRestoreSearch
1036 - Look at the possibility of loading only the necessary
1037 data into the restore tree (i.e. do it one directory at a
1038 time as the user walks through the tree).
1039 - Possibly use the hash code if the user selects all for a restore command.
1040 - Fix "restore all" to bypass building the tree.
1041 - Prohibit backing up archive device (findlib/find_one.c:128)
1042 - Implement Release Device in the Job resource to unmount a drive.
1043 - Implement Acquire Device in the Job resource to mount a drive,
1044 be sure this works with admin jobs so that the user can get
1045 prompted to insert the correct tape. Possibly some way to say to
1046 run the job but don't save the files.
1047 - Make things like list where a file is saved case independent for
1050 - Use autochanger to handle multiple devices.
1051 - On Windows with very long path names, it may be impossible to create
1052 a file (and thus restore it) because the total length is too long.
1053 We must cd into the directory then create the file without the
1055 - Implement a Recycle command
1056 - Test a second language e.g. french.
1057 - Start working on Base jobs.
1058 - Implement UnsavedFiles DB record.
1059 - From Phil Stracchino:
1060 It would probably be a per-client option, and would be called
1061 something like, say, "Automatically purge obsoleted jobs". What it
1062 would do is, when you successfully complete a Differential backup of a
1063 client, it would automatically purge all Incremental backups for that
1064 client that are rendered redundant by that Differential. Likewise,
1065 when a Full backup on a client completed, it would automatically purge
1066 all Differential and Incremental jobs obsoleted by that Full backup.
1067 This would let people minimize the number of tapes they're keeping on
1068 hand without having to master the art of retention times.
1069 - When doing a Backup send all attributes back to the Director, who
1070 would then figure out what files have been deleted.
1071 - Currently in mount.c:236 the SD simply creates a Volume. It should have
1072 explicit permission to do so. It should also mark the tape in error
1073 if there is an error.
1074 - Cancel waiting for Client connect in SD if FD goes away.
1076 - Implement timeout in response() when it should come quickly.
1077 - Implement a Slot priority (loaded/not loaded).
1078 - Implement "vacation" Incremental only saves.
1079 - Implement create "FileSet"?
1080 - Add prefixlinks to where or not where absolute links to FD.
1081 - Issue message to mount a new tape before the rewind.
1082 - Simplified client job initiation for portables.
1083 - If SD cannot open a drive, make it periodically retry.
1084 - Add more of the config info to the tape label.
1086 - If tape is marked read-only, then try opening it read-only rather than
1087 failing, and remember that it cannot be written.
1088 - Refine SD waiting output:
1089 Device is being positioned
1090 > Device is being positioned for append
1091 > Device is being positioned to file x
1093 - Figure out some way to estimate output size and to avoid splitting
1094 a backup across two Volumes -- this could be useful for writing CDROMs
1095 where you really prefer not to have it split -- not serious.
1096 - Have SD compute MD5 or SHA1 and compare to what FD computes.
1097 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
1098 actual data on the Volume and compare it.
1099 - Implement Bacula plugins -- design API
1100 - Make bcopy read through bad tape records.
1101 - Program files (i.e. execute a program to read/write files).
1102 Pass read date of last backup, size of file last time.
1103 - Add Signature type to File DB record.
1104 - CD into subdirectory when open()ing files for backup to
1105 speed up things. Test with testfind().
1106 - Priority job to go to top of list.
1107 - Why are save/restore of device different sizes (sparse?) Yup! Fix it.
1108 - Implement some way for the Console to dynamically create a job.
1109 - Solaris -I on tar for include list
1110 - Need a verbose mode in restore, perhaps to bsr.
1111 - bscan without -v is too quiet -- perhaps show jobs.
1112 - Add code to reject whole blocks if not wanted on restore.
1113 - Check if we can increase Bacula FD priorty in Win2000
1114 - Make sure the MaxVolFiles is fully implemented in SD
1115 - Check if both CatalogFiles and UseCatalog are set to SD.
1116 - Possibly add email to Watchdog if drive is unmounted too
1117 long and a job is waiting on the drive.
1118 - After unmount, if restore job started, ask to mount.
1119 - Add UA rc and history files.
1120 - put termcap (used by console) in ./configure and
1121 allow -with-termcap-dir.
1122 - Fix Autoprune for Volumes to respect need for full save.
1123 - Compare tape to Client files (attributes, or attributes and data)
1124 - Make all database Ids 64 bit.
1125 - Allow console commands to detach or run in background.
1126 - Fix status delay on storage daemon during rewind.
1127 - Add SD message variables to control operator wait time
1128 - Maximum Operator Wait
1129 - Minimum Message Interval
1130 - Maximum Message Interval
1131 - Send Operator message when cannot read tape label.
1132 - Verify level=Volume (scan only), level=Data (compare of data to file).
1133 Verify level=Catalog, level=InitCatalog
1135 - Add keyword search to show command in Console.
1136 - Events : tape has more than xxx bytes.
1137 - Complete code in Bacula Resources -- this will permit
1138 reading a new config file at any time.
1139 - Handle ctl-c in Console
1140 - Implement script driven addition of File daemon to config files.
1141 - Think about how to make Bacula work better with File (non-tape) archives.
1142 - Write Unix emulator for Windows.
1143 - Put memory utilization in Status output of each daemon
1144 if full status requested or if some level of debug on.
1145 - Make database type selectable by .conf files i.e. at runtime
1146 - Set flag for uname -a. Add to Volume label.
1147 - Restore files modified after date
1148 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
1149 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
1150 - Timout a job or terminate if link goes down, or reopen link and query.
1151 - Concept of precious tapes (cannot be reused).
1152 - Make bcopy copy with a single tape drive.
1153 - Permit changing ownership during restore.
1156 > My suggestion: Add a feature on the systray menu-icon menu to request
1157 > an immediate backup now. This would be useful for laptop users who may
1158 > not be on the network when the regular scheduled backup is run.
1160 > My wife's suggestion: Add a setting to the win32 client to allow it to
1161 > shut down the machine after backup is complete (after, of course,
1162 > displaying a "System will shut down in one minute, click here to cancel"
1163 > warning dialog). This would be useful for sites that want user
1164 > woorkstations to be shut down overnight to save power.
1167 - Autolabel should be specified by DIR instead of SD.
1169 - Add media capacity
1170 - AutoScan (check checksum of tape)
1171 - Format command = "format /dev/nst0"
1175 - Seek resolution (usually corresponds to buffer size)
1176 - EODErrorCode=ENOSPC or code
1177 - Partial Read error code
1178 - Partial write error code
1179 - Nonformatted read error
1180 - Nonformatted write error
1181 - WriteProtected error
1185 - IgnoreCloseErrors=yes
1195 - FD sends unsaved file list to Director at end of job (see
1197 - File daemon should build list of files skipped, and then
1198 at end of save retry and report any errors.
1199 - Write a Storage daemon that uses pipes and
1200 standard Unix programs to write to the tape.
1202 - Need something that monitors the JCR queue and
1203 times out jobs by asking the deamons where they are.
1204 - Enhance Jmsg code to permit buffering and saving to disk.
1205 - device driver = "xxxx" for drives.
1206 - Verify from Volume
1207 - Ensure that /dev/null works
1208 - Need report class for messages. Perhaps
1209 report resource where report=group of messages
1210 - enhance scan_attrib and rename scan_jobtype, and
1211 fill in code for "since" option
1212 - Director needs a time after which the report status is sent
1213 anyway -- or better yet, a retry time for the job.
1214 - Don't reschedule a job if previous incarnation is still running.
1215 - Some way to automatically backup everything is needed????
1216 - Need a structure for pending actions:
1218 - termination status (part of buffered msgs?)
1220 Read, Write, Clean, Delete
1221 - Login to Bacula; Bacula users with different permissions:
1222 owner, group, user, quotas
1223 - Store info on each file system type (probably in the job header on tape.
1224 This could be the output of df; or perhaps some sort of /etc/mtab record.
1226 ========= ideas ===============
1227 From: "Jerry K. Schieffer" <jerry@skylinetechnology.com>
1228 To: <kern@sibbald.com>
1229 Subject: RE: [Bacula-users] future large programming jobs
1230 Date: Thu, 26 Feb 2004 11:34:54 -0600
1232 I noticed the subject thread and thought I would offer the following
1233 merely as sources of ideas, i.e. something to think about, not even as
1234 strong as a request. In my former life (before retiring) I often
1235 dealt with backups and storage management issues/products as a
1236 developer and as a consultant. I am currently migrating my personal
1237 network from amanda to bacula specifically because of the ability to
1238 cross media boundaries during storing backups.
1239 Are you familiar with the commercial product called ADSM (I think IBM
1240 now sells it under the Tivoli label)? It has a couple of interesting
1241 ideas that may apply to the following topics.
1243 1. Migration: Consider that when you need to restore a system, there
1244 may be pressure to hurry. If all the information for a single client
1245 can eventually end up on the same media (and in chronological order),
1246 the restore is facillitated by not having to search past information
1247 from other clients. ADSM has the concept of "client affinity" that
1248 may be associated with it's storage pools. It seems to me that this
1249 concept (as an optional feature) might fit in your architecture for
1252 ADSM also has the concept of defining one or more storage pools as
1253 "copy pools" (almost mirrors, but only in the sense of contents).
1254 These pools provide the ability to have duplicte data stored both
1255 onsite and offsite. The copy process can be scheduled to be handled
1256 by their storage manager during periods when there is no backup
1257 activity. Again, the migration process might be a place to consider
1258 implementing something like this.
1261 > It strikes me that it would be very nice to be able to do things
1263 > have the Job(s) backing up the machines run, and once they have all
1264 > completed, start a migration job to copy the data from disks Volumes
1266 > a tape library and then to offsite storage. Maybe this can already
1268 > done with some careful scheduling and Job prioritzation; the events
1269 > mechanism described below would probably make it very easy.
1271 This is the goal. In the first step (before events), you simply
1273 the Migration to tape later.
1275 2. Base jobs: In ADSM, each copy of each stored file is tracked in
1276 the database. Once a file (unique by path and metadata such as dates,
1277 size, ownership, etc.) is in a copy pool, no more copies are made. In
1278 other words, when you start ADSM, it begins like your concept of a
1279 base job. After that it is in the "incremental" mode. You can
1280 configure the number of "generations" of files to be retained, plus a
1281 retention date after which even old generations are purged. The
1282 database tracks the contents of media and projects the percentage of
1283 each volume that is valid. When the valid content of a volume drops
1284 below a configured percentage, the valid data are migrated to another
1285 volume and the old volume is marked as empty. Note, this requires
1286 ADSM to have an idea of the contents of a client, i.e. marking the
1287 database when an existing file was deleted, but this would solve your
1288 issue of restoring a client without restoring deleted files.
1290 This is pretty far from what bacula now does, but if you are going to
1291 rip things up for Base jobs,.....
1292 Also, the benefits of this are huge for very large shops, especially
1293 with media robots, but are a pain for shops with manual media
1297 > Base jobs sound pretty useful, but I'm not dying for them.
1299 Nobody is dying for them, but when you see what it does, you will die
1302 3. Restoring deleted files: Since I think my comments in (2) above
1303 have low probability of implementation, I'll also suggest that you
1304 could approach the issue of deleted files by a mechanism of having the
1305 fd report to the dir, a list of all files on the client for every
1306 backup job. The dir could note in the database entry for each file
1307 the date that the file was seen. Then if a restore as of date X takes
1308 place, only files that exist from before X until after X would be
1309 restored. Probably the major cost here is the extra date container in
1310 each row of the files table.
1312 Thanks for "listening". I hope some of this helps. If you want to
1313 contact me, please send me an email - I read some but not all of the
1314 mailing list traffic and might miss a reply there.
1316 Please accept my compliments for bacula. It is doing a great job for
1317 me!! I sympathize with you in the need to wrestle with excelence in
1318 execution vs. excelence in feature inclusion.
1323 ==============================
1326 - Design at hierarchial storage for Bacula. Migration and Clone.
1327 - Implement FSM (File System Modules).
1328 - Audit M_ error codes to ensure they are correct and consistent.
1329 - Add variable break characters to lex analyzer.
1330 Either a bit mask or a string of chars so that
1331 the caller can change the break characters.
1332 - Make a single T_BREAK to replace T_COMMA, etc.
1333 - Ensure that File daemon and Storage daemon can
1334 continue a save if the Director goes down (this
1335 is NOT currently the case). Must detect socket error,
1336 buffer messages for later.
1337 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
1338 - Add ability to backup to two Storage devices (two SD sessions) at
1339 the same time -- e.g. onsite, offsite.
1340 - Add the ability to consolidate old backup sets (basically do a restore
1341 to tape and appropriately update the catalog). Compress Volume sets.
1342 Might need to spool via file is only one drive is available.
1343 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
1344 someway to do so with every volume that has less than x% valid
1348 Migration: Move a backup from one Volume to another
1349 Clone: Copy a backup -- two Volumes
1351 Bacula Migration is based on Jobs (apparently Networker is file by file).
1353 Migration triggered by:
1357 Highwater mark (keep total size)
1362 ======================================================
1364 It is somewhat like a Full save becomes an incremental since
1365 the Base job (or jobs) plus other non-base files.
1367 - A Base backup is same as Full backup, just different type.
1368 - New BaseFiles table that contains:
1370 BaseJobId - Base JobId referenced for this FileId (needed ???)
1371 JobId - JobId currently running
1372 FileId - File not backed up, exists in Base Job
1373 FileIndex - FileIndex from Base Job.
1374 i.e. for each base file that exists but is not saved because
1375 it has not changed, the File daemon sends the JobId, BaseId,
1376 FileId, FileIndex back to the Director who creates the DB entry.
1377 - To initiate a Base save, the Director sends the FD
1378 the FileId, and full filename for each file in the Base.
1379 - When the FD finds a Base file, he requests the Director to
1380 send him the full File entry (stat packet plus MD5), or
1381 conversely, the FD sends it to the Director and the Director
1382 says yes or no. This can be quite rapid if the FileId is kept
1383 by the FD for each Base Filename.
1384 - It is probably better to have the comparison done by the FD
1385 despite the fact that the File entry must be sent across the
1387 - An alternative would be to send the FD the whole File entry
1388 from the start. The disadvantage is that it requires a lot of
1389 space. The advantage is that it requires less communications
1391 - The Job record must be updated to indicate that one or more
1393 - At end of Job, FD returns:
1394 1. Count of base files/bytes not written to tape (i.e. matches)
1395 2. Count of base file that were saved i.e. had changed.
1396 - No tape record would be written for a Base file that matches, in the
1397 same way that no tape record is written for Incremental jobs where
1398 the file is not saved because it is unchanged.
1399 - On a restore, all the Base file records must explicitly be
1400 found from the BaseFile tape. I.e. for each Full save that is marked
1401 to have one or more Base Jobs, search the BaseFile for all occurrences
1403 - An optimization might be to make the BaseFile have:
1409 This would avoid the need to explicitly fetch each File record for
1410 the Base job. The Base Job record will be fetched to get the
1411 VolSessionId and VolSessionTime.
1412 =========================================================
1415 ==========================================================
1417 For each Incremental job that is run, there may be files that
1418 were found but not saved because they were locked (this applies
1419 only to Windows). Such a system could send back to the Director
1420 a list of Unsaved files.
1422 - New UnSavedFiles table that contains:
1426 - Then in the next Incremental job, the list of Unsaved Files will be
1427 feed to the FD, who will ensure that they are explicitly chosen even
1428 if standard date/time check would not have selected them.
1429 =============================================================
1433 Multiple drive autochanger data: see Alan Brown
1434 mtx -f xxx unloadStorage Element 1 is Already Full(drive 0 was empty)
1435 Unloading Data Transfer Element into Storage Element 1...source Element
1436 Address 480 is Empty
1438 (drive 0 was empty and so was slot 1)
1439 > mtx -f xxx load 15 0
1440 no response, just returns to the command prompt when complete.
1441 > mtx -f xxx status Storage Changer /dev/changer:2 Drives, 60 Slots ( 2 Import/Export )
1442 Data Transfer Element 0:Full (Storage Element 15 Loaded):VolumeTag = HX001
1443 Data Transfer Element 1:Empty
1444 Storage Element 1:Empty
1445 Storage Element 2:Full :VolumeTag=HX002
1446 Storage Element 3:Full :VolumeTag=HX003
1447 Storage Element 4:Full :VolumeTag=HX004
1448 Storage Element 5:Full :VolumeTag=HX005
1449 Storage Element 6:Full :VolumeTag=HX006
1450 Storage Element 7:Full :VolumeTag=HX007
1451 Storage Element 8:Full :VolumeTag=HX008
1452 Storage Element 9:Full :VolumeTag=HX009
1453 Storage Element 10:Full :VolumeTag=HX010
1454 Storage Element 11:Empty
1455 Storage Element 12:Empty
1456 Storage Element 13:Empty
1457 Storage Element 14:Empty
1458 Storage Element 15:Empty
1459 Storage Element 16:Empty....
1460 Storage Element 28:Empty
1461 Storage Element 29:Full :VolumeTag=CLNU01L1
1462 Storage Element 30:Empty....
1463 Storage Element 57:Empty
1464 Storage Element 58:Full :VolumeTag=NEX261L2
1465 Storage Element 59 IMPORT/EXPORT:Empty
1466 Storage Element 60 IMPORT/EXPORT:Empty
1468 Unloading Data Transfer Element into Storage Element 15...done
1470 (just to verify it remembers where it came from, however it can be
1471 overrriden with mtx unload {slotnumber} to go to any storage slot.)
1473 There needs to be a table of drive # to devices somewhere - If there are
1474 multiple changers or drives there may not be a 1:1 correspondance between
1475 changer drive number and system device name - and depending on the way the
1476 drives are hooked up to scsi busses, they may not be linearly numbered
1477 from an offset point either.something like
1479 Autochanger drives = 2
1480 Autochanger drive 0 = /dev/nst1
1481 Autochanger drive 1 = /dev/nst2
1482 IMHO, it would be _safest_ to use explicit mtx unload commands at all
1483 times, not just for multidrive changers. For a 1 drive changer, that's
1489 MTX's manpage (1.2.15):
1490 unload [<slotnum>] [ <drivenum> ]
1491 Unloads media from drive <drivenum> into slot
1492 <slotnum>. If <drivenum> is omitted, defaults to
1493 drive 0 (as do all commands). If <slotnum> is
1494 omitted, defaults to the slot that the drive was
1495 loaded from. Note that there's currently no way
1496 to say 'unload drive 1's media to the slot it
1497 came from', other than to explicitly use that
1498 slot number as the destination.AB
1504 undef# camcontrol devlist
1505 <WANGTEK 51000 SCSI M74H 12B3> at scbus0 target 2 lun 0 (pass0,sa0)
1506 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 0 (pass1,sa1)
1507 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 1 (pass2)
1509 tapeinfo -f /dev/sg0 with a bad tape in drive 1:
1510 [kern@rufus mtx-1.2.17kes]$ ./tapeinfo -f /dev/sg0
1511 Product Type: Tape Drive
1513 Product ID: 'C5713A '
1515 Attached Changer: No
1516 TapeAlert[3]: Hard Error: Uncorrectable read/write error.
1517 TapeAlert[20]: Clean Now: The tape drive neads cleaning NOW.
1524 Medium Type: Not Loaded
1527 DataCompEnabled: yes
1528 DataCompCapable: yes
1529 DataDeCompEnabled: yes
1536 Handling removable disks
1538 From: Karl Cunningham <karlc@keckec.com>
1540 My backups are only to hard disk these days, in removable bays. This is my
1541 idea of how a backup to hard disk would work more smoothly. Some of these
1542 things Bacula does already, but I mention them for completeness. If others
1543 have better ways to do this, I'd like to hear about it.
1545 1. Accommodate several disks, rotated similar to how tapes are. Identified
1546 by partition volume ID or perhaps by the name of a subdirectory.
1547 2. Abort & notify the admin if the wrong disk is in the bay.
1548 3. Write backups to different subdirectories for each machine to be backed
1550 4. Volumes (files) get created as needed in the proper subdirectory, one
1552 5. When a disk is recycled, remove or zero all old backup files. This is
1553 important as the disk being recycled may be close to full. This may be
1554 better done manually since the backup files for many machines may be
1555 scattered in many subdirectories.
1560 - Make sure that all do_prompt() calls in Dir check for
1561 -1 (error) and -2 (cancel) returns.
1562 - Fix foreach_jcr() to have free_jcr() inside next().
1563 jcr=jcr_walk_start();
1564 for ( ; jcr; (jcr=jcr_walk_next(jcr)) )
1567 - A Volume taken from Scratch should take on the retention period
1569 - Correct doc for Maximum Changer Wait (and others) accepting only
1571 - Implement status that shows why a job is being held in reserve, or
1572 rather why none of the drives are suitable.
1573 - Implement a way to disable a drive (so you can use the second
1574 drive of an autochanger, and the first one will not be used or
1576 - Make sure Maximum Volumes is respected in Pools when adding
1577 Volumes (e.g. when pulling a Scratch volume).
1578 - Keep same dcr when switching device ...
1579 - Implement code that makes the Dir aware that a drive is an
1580 autochanger (so the user doesn't need to use the Autochanger = yes
1582 - Make catalog respect ACL.
1583 - Add recycle count to Media record.
1584 - Add initial write date to Media record.
1585 - Fix store_yesno to be store_bitmask.
1586 --- create_file.c.orig Fri Jul 8 12:13:05 2005
1587 +++ create_file.c Fri Jul 8 12:13:07 2005
1589 attr->ofname, be.strerror());
1592 + } else if(S_ISSOCK(attr->statp.st_mode)) {
1593 + Dmsg1(200, "Skipping socket: %s\n", attr->ofname);
1595 Dmsg1(200, "Restore node: %s\n", attr->ofname);
1596 if (mknod(attr->ofname, attr->statp.st_mode, attr->statp.st_rdev) != 0 && errno != EEXIST) {
1597 - Add true/false to conf same as yes/no
1598 - Reserve blocks other restore jobs when first cannot connect to SD.
1599 - Fix Maximum Changer Wait, Maximum Open Wait, Maximum Rewind Wait to
1600 accept time qualifiers.
1601 - Does ClientRunAfterJob fail the job on a bad return code?
1602 - Make hardlink code at line 240 of find_one.c use binary search.
1603 - Add ACL error messages in src/filed/acl.c.
1604 - Make authentication failures single threaded.
1605 - Make Dir and SD authentication errors single threaded.
1607 - Fix catreq.c digestbuf at line 411 in src/dird/catreq.c