9 - !!! Cannot restore two jobs a the same time that were
10 written simultaneously unless they were totally spooled.
11 - Document cleaning up the spool files:
12 db, pid, state, bsr, mail, conmsg, spool
13 - Document the multiple-drive-changer.txt script.
14 - Pruning with Admin job.
15 - Does WildFile match against full name? Doc.
16 - %d and %v only valid on Director, not for ClientRunBefore/After.
17 - During tests with the 260 char fix code, I found one problem:
18 if the system "sees" a long path once, it seems to forget it's
19 working drive (e.g. c:\), which will lead to a problem during
20 the next job (create bootstrap file will fail). Here is the
21 workaround: specify absolute working and pid directory in
22 bacula-fd.conf (e.g. c:\bacula\working instead of
24 - Document techniques for restoring large numbers of files.
25 - Document setting my.cnf to big file usage.
26 - Add example of proper index output to doc. show index from File;
27 - Correct the Include syntax in the m4.xxx files in examples/conf
28 - Document JobStatus and Termination codes.
29 - Fix the error with the "DVI file can't be opened" while
30 building the French PDF.
31 - Document more DVD stuff
39 - Document all the little details of setting up certificates for
40 the Bacula data encryption code.
41 - Document more precisely how to use master keys -- especially
42 for disaster recovery.
46 - Add Where: client:/.... to restore job report.
47 - Ensure that each device in an Autochanger has a different
49 - Add Catalog = to Pool resource so that pools will exist
50 in only one catalog -- currently Pools are "global".
51 - Look at sg_logs -a /dev/sg0 for getting soft errors.
52 - btape "test" command with Offline on Unmount = yes
54 This test is essential to Bacula.
56 I'm going to write one record in file 0,
57 two records in file 1,
58 and three records in file 2
60 02-Feb 11:00 btape: ABORTING due to ERROR in dev.c:715
61 dev.c:714 Bad call to rewind. Device "LTO" (/dev/nst0) not open
62 02-Feb 11:00 btape: Fatal Error because: Bacula interrupted by signal 11: Segmentation violation
63 Kaboom! btape, btape got signal 11. Attempting traceback.
65 - Why the heck doesn't bacula drop root priviledges before connecting to
67 - Ensure that moving a purged Volume in ua_purge.c to the RecyclePool
69 - Why doesn't @"xxx abc" work in a conf file?
70 - Figure out some way to "automatically" backup conf changes.
71 - Look at using posix_fadvise(2) for backups -- see bug #751.
72 Possibly add the code at findlib/bfile.c:795
73 - Add the OS version back to the Win32 client info.
74 - Restarted jobs have a NULL in the from field.
75 - Modify SD status command to indicate when the SD is writing
76 to a DVD (the device is not open -- see bug #732).
77 - Look at the possibility of adding "SET NAMES UTF8" for MySQL,
78 and possibly changing the blobs into varchar.
79 - Check if gnome-console works with TLS.
80 - Ensure that the SD re-reads the Media record if the JobFiles
81 does not match -- it may have been updated by another job.
82 - Look at moving the Storage directive from the Job to the
83 Pool in the default conf files.
84 - Test FIFO backup/restore -- make regression
86 - Test Volume compatibility between machine architectures
87 - Encryption documentation
88 - Wrong jobbytes with query 12 (todo)
89 - bacula-1.38.2-ssl.patch
90 - Bare-metal recovery Windows (todo)
97 - Add doc for bweb -- especially Installation
99 http://www.orangecrate.com/modules.php?name=News&file=article&sid=501
101 - FD-SD quick disconnect
102 - Despool attributes in separate thread
105 - Check why restore repeatedly sends Rechdrs between
106 each data chunk -- according to James Harper 9Jan07.
107 - Building the in memory restore tree is slow.
110 - Full at least once a month, ...
111 - Cancel Inc if Diff/Full running
112 - More intelligent re-run
113 - New/deleted file backup
115 - Incremental backup -- rsync, Stow
121 - Look at mondo/mindi
122 - Don't restore Solaris Door files:
123 #define S_IFDOOR in st_mode.
124 see: http://docs.sun.com/app/docs/doc/816-5173/6mbb8ae23?a=view#indexterm-360
125 - Make Bacula by default not backup tmpfs, procfs, sysfs, ...
126 - Fix hardlinked immutable files when linking a second file, the
127 immutable flag must be removed prior to trying to link it.
128 - Implement Python event for backing up/restoring a file.
129 - Change dbcheck to tell users to use native tools for fixing
130 broken databases, and to ensure they have the proper indexes.
131 - add udev rules for Bacula devices.
132 - If a job terminates, the DIR connection can close before the
133 Volume info is updated, leaving the File count wrong.
134 - Look at why SIGPIPE during connection can cause seg fault in
135 writing the daemon message, when Dir dropped to bacula:bacula
136 - Look at zlib 32 => 64 problems.
137 - Possibly turn on St. Bernard code.
138 - Fix bextract to restore ACLs, or better yet, use common routines.
139 - Do we migrate appendable Volumes?
140 - Remove queue.c code.
141 - Print warning message if LANG environment variable does not specify
143 - New dot commands from Arno.
144 .show device=xxx lists information from one storage device, including
145 devices (I'm not even sure that information exists in the DIR...)
146 .move eject device=xxx mostly the same as 'unmount xxx' but perhaps with
147 better machine-readable output like "Ok" or "Error busy"
148 .move eject device=xxx toslot=yyy the same as above, but with a new
149 target slot. The catalog should be updated accordingly.
150 .move transfer device=xxx fromslot=yyy toslot=zzz
153 - Article: http://www.heise.de/open/news/meldung/83231
154 - Article: http://www.golem.de/0701/49756.html
155 - Article: http://lwn.net/Articles/209809/
156 - Article: http://www.onlamp.com/pub/a/onlamp/2004/01/09/bacula.html
157 - Article: http://www.linuxdevcenter.com/pub/a/linux/2005/04/07/bacula.html
158 - Article: http://www.osreviews.net/reviews/admin/bacula
159 - Article: http://www.debianhelp.co.uk/baculaweb.htm
161 - It appears to me that you have run into some sort of race
162 condition where two threads want to use the same Volume and they
163 were both given access. Normally that is no problem. However,
164 one thread wanted the particular Volume in drive 0, but it was
165 loaded into drive 1 so it decided to unload it from drive 1 and
166 then loaded it into drive 0, while the second thread went on
167 thinking that the Volume could be used in drive 1 not realizing
168 that in between time, it was loaded in drive 0.
169 I'll look at the code to see if there is some way we can avoid
170 this kind of problem. Probably the best solution is to make the
171 first thread simply start using the Volume in drive 1 rather than
172 transferring it to drive 0.
173 - Fix re-read of last block to check if job has actually written
174 a block, and check if block was written by a different job
175 (i.e. multiple simultaneous jobs writing).
176 - Figure out how to configure query.sql. Suggestion to use m4:
177 == changequote.m4 ===
178 changequote(`[',`]')dnl
179 ==== query.sql.in ===
180 :List next 20 volumes to expire
182 Pool.Name AS PoolName,
187 [ FROM_UNIXTIME(UNIX_TIMESTAMP(Media.LastWritten) Media.VolRetention) AS Expire, ])dnl
189 [ media.lastwritten + interval '1 second' * media.volretention as expire, ])dnl
193 ON Media.PoolId=Pool.PoolId
194 WHERE Media.LastWritten>0
198 Command: m4 -DmySQL changequote.m4 query.sql.in >query.sql
200 The problem is that it requires m4, which is not present on all machines
202 - Given all the problems with FIFOs, I think the solution is to do something a
203 little different, though I will look at the code and see if there is not some
204 simple solution (i.e. some bug that was introduced). What might be a better
205 solution would be to use a FIFO as a sort of "key" to tell Bacula to read and
206 write data to a program rather than the FIFO. For example, suppose you
211 Then, I could imagine if you backup and restore this file with a direct
212 reference as is currently done for fifos, instead, during backup Bacula will
215 /home/kern/my-fifo.backup
217 and read the data that my-fifo.backup writes to stdout. For restore, Bacula
220 /home/kern/my-fifo.restore
222 and send the data backed up to stdout. These programs can either be an
223 executable or a shell script and they need only read/write to stdin/stdout.
225 I think this would give a lot of flexibility to the user without making any
226 significant changes to Bacula.
231 select FilenameId from Filename where Name='';
232 # Get list of all directories referenced in a Backup.
233 select Path.Path from Path,File where File.JobId=nnn and
234 File.FilenameId=(FilenameId-from-above) and File.PathId=Path.PathId
235 order by Path.Path ASC;
237 - Look into using Dart for testing
238 http://public.kitware.com/Dart/HTML/Index.shtml
240 - Look into replacing autotools with cmake
241 http://www.cmake.org/HTML/Index.html
243 === Migration from David ===
244 What I'd like to see:
247 Name = "<poolname>-migrate"
251 Migration Selection Type = LowestUtil | OldestVol | PoolOccupancy |
252 Client | PoolResidence | Volume | JobName | SQLquery
253 Migration Selection Pattern = "regexp"
254 Next Pool = <override>
257 There should be no need for a Level (migration is always Full, since you
258 don't calculate differential/incremental differences for migration),
259 Storage should be determined by the volume types in the pool, and Client
260 is really a selection issue. Migration should always occur to the
261 NextPool defined in the pool definition. If no nextpool is defined, the
262 job should end with a reason of "no place to go". If Next Pool statement
263 is present, we override the check in the pool definition and use the
266 Here's how I'd define Migration Selection Types:
269 Client -- Migrate data from selected client only. Migration Selection
270 Pattern regexp provides pattern to select client names, eg ^FS00* makes
271 all client names starting with FS00 eligible for migration.
273 Jobname -- Migration all jobs matching name. Migration Selection Pattern
274 regexp provides pattern to select jobnames existing in pool.
276 Volume -- Migrate all data on specified volumes. Migration Selection
277 Pattern regexp provides selection criteria for volumes to be migrated.
278 Volumes must exist in pool to be eligible for migration.
282 LowestUtil -- Identify the volume in the pool with the least data on it
283 and empty it. No Migration Selection Pattern required.
285 OldestVol -- Identify the LRU volume with data written, and empty it. No
286 Migration Selection Pattern required.
288 PoolOccupancy -- if pool occupancy exceeds <highmig>, migrate volumes
289 (starting with most full volumes) until pool occupancy drops below
290 <lowmig>. Pool highmig and lowmig values are in pool definition, no
291 Migration Selection Pattern required.
295 SQLQuery -- Migrate all jobuids returned by the supplied SQL query.
296 Migration Selection Pattern contains SQL query to execute; should return
297 a list of 1 or more jobuids to migrate.
299 PoolResidence -- Migrate data sitting in pool for longer than
300 PoolResidence value in pool definition. Migration Selection Pattern
301 optional; if specified, override value in pool definition (value in
305 [ possibly a Python event -- kes ]
307 - Mount on an Autochanger with no tape in the drive causes:
308 Automatically selected Storage: LTO-changer
309 Enter autochanger drive[0]: 0
310 3301 Issuing autochanger "loaded drive 0" command.
311 3302 Autochanger "loaded drive 0", result: nothing loaded.
312 3301 Issuing autochanger "loaded drive 0" command.
313 3302 Autochanger "loaded drive 0", result: nothing loaded.
314 3902 Cannot mount Volume on Storage Device "LTO-Drive1" (/dev/nst0) because:
315 Couldn't rewind device "LTO-Drive1" (/dev/nst0): ERR=dev.c:678 Rewind error on "LTO-Drive1" (/dev/nst0). ERR=No medium found.
316 3905 Device "LTO-Drive1" (/dev/nst0) open but no Bacula volume is mounted.
317 If this is not a blank tape, try unmounting and remounting the Volume.
318 - If Drive 0 is blocked, and drive 1 is set "Autoselect=no", drive 1 will
320 - Autochanger did not change volumes.
321 select * from Storage;
322 +-----------+-------------+-------------+
323 | StorageId | Name | AutoChanger |
324 +-----------+-------------+-------------+
325 | 1 | LTO-changer | 0 |
326 +-----------+-------------+-------------+
327 05-May 03:50 roxie-sd: 3302 Autochanger "loaded drive 0", result is Slot 11.
328 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Warning: Director wanted Volume "LT
329 Current Volume "LT0-002" not acceptable because:
330 1997 Volume "LT0-002" not in catalog.
331 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Error: Autochanger Volume "LT0-002"
332 Setting InChanger to zero in catalog.
333 05-May 03:50 roxie-dir: Tibs.2006-05-05_03.05.02 Error: Unable to get Media record
335 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: Error getting Volume i
336 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: Job 530 canceled.
337 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: spool.c:249 Fatal appe
338 05-May 03:49 Tibs: Tibs.2006-05-05_03.05.02 Fatal error: c:\cygwin\home\kern\bacula
347 FirstWritten: 2006-05-05 03:11:54
348 LastWritten: 2006-05-05 03:50:23
349 LabelDate: 2005-12-26 16:52:40
360 VolRetention: 31,536,000
372 Note VolStatus is blank!!!!!
379 FirstWritten: 0000-00-00 00:00:00
380 LastWritten: 0000-00-00 00:00:00
381 LabelDate: 2005-12-26 16:52:40
392 VolRetention: 31,536,000
405 Automatically selected Storage: LTO-changer
406 Enter autochanger drive[0]: 0
407 3301 Issuing autochanger "loaded drive 0" command.
408 3302 Autochanger "loaded drive 0", result: nothing loaded.
409 3301 Issuing autochanger "loaded drive 0" command.
410 3302 Autochanger "loaded drive 0", result: nothing loaded.
411 3902 Cannot mount Volume on Storage Device "LTO-Drive1" (/dev/nst0) because:
412 Couldn't rewind device "LTO-Drive1" (/dev/nst0): ERR=dev.c:678 Rewind error on "LTO-Drive1" (/dev/nst0). ERR=No medium found.
414 3905 Device "LTO-Drive1" (/dev/nst0) open but no Bacula volume is mounted.
415 If this is not a blank tape, try unmounting and remounting the Volume.
417 - http://www.dwheeler.com/essays/commercial-floss.html
418 - Add VolumeLock to prevent all but lock holder (SD) from updating
419 the Volume data (with the exception of VolumeState).
420 - The btape fill command does not seem to use the Autochanger
421 - Make Windows installer default to system disk drive.
422 - Look at using ioctl(FIOBMAP, ...) on Linux, and
423 DeviceIoControl(..., FSCTL_QUERY_ALLOCATED_RANGES, ...) on
424 Win32 for sparse files.
425 http://www.flexhex.com/docs/articles/sparse-files.phtml
426 http://www.informatik.uni-frankfurt.de/~loizides/reiserfs/fibmap.html
427 - Directive: at <event> "command"
428 - Command: pycmd "command" generates "command" event. How to
429 attach to a specific job?
430 - Integrate Christopher's St. Bernard code.
431 - run_cmd() returns int should return JobId_t
432 - get_next_jobid_from_list() returns int should return JobId_t
433 - Document export LDFLAGS=-L/usr/lib64
434 - Don't attempt to restore from "Disabled" Volumes.
435 - Network error on Win32 should set Win32 error code.
436 - What happens when you rename a Disk Volume?
437 - Job retention period in a Pool (and hence Volume). The job would
439 - Look at -D_FORTIFY_SOURCE=2
440 - Add Win32 FileSet definition somewhere
441 - Look at fixing restore status stats in SD.
442 - Look at using ioctl(FIMAP) and FIGETBSZ for sparse files.
443 http://www.informatik.uni-frankfurt.de/~loizides/reiserfs/fibmap.html
444 - Implement a mode that says when a hard read error is
445 encountered, read many times (as it currently does), and if the
446 block cannot be read, skip to the next block, and try again. If
447 that fails, skip to the next file and try again, ...
449 create table LevelType (LevelType binary(1), LevelTypeLong tinyblob);
450 insert into LevelType (LevelType,LevelTypeLong) values
454 - Show files/second in client status output.
455 - Add a recursive mark command (rmark) to restore.
456 - "Minimum Job Interval = nnn" sets minimum interval between Jobs
457 of the same level and does not permit multiple simultaneous
458 running of that Job (i.e. lets any previous invocation finish
459 before doing Interval testing).
460 - Look at simplifying File exclusions.
461 - New directive "Delete purged Volumes"
462 - new pool XXX with ScratchPoolId = MyScratchPool's PoolId and
463 let it fill itself, and RecyclePoolId = XXX's PoolId so I can
464 see if it become stable and I just have to supervise
466 - If I want to remove this pool, I set RecyclePoolId = MyScratchPool's
467 PoolId, and when it is empty remove it.
468 - Figure out how to recycle Scratch volumes back to the Scratch Pool.
470 - Allow Check Labels to be used with Bacula labels.
471 - "Resuming" a failed backup (lost line for example) by using the
472 failed backup as a sort of "base" job.
474 - Email to the user when the tape is about to need changing x
475 days before it needs changing.
476 - Command to show next tape that will be used for a job even
477 if the job is not scheduled.
478 - From: Arunav Mandal <amandal@trolltech.com>
479 1. When jobs are running and bacula for some reason crashes or if I do a
480 restart it remembers and jobs it was running before it crashed or restarted
481 as of now I loose all jobs if I restart it.
483 2. When spooling and in the midway if client is disconnected for instance a
484 laptop bacula completely discard the spool. It will be nice if it can write
485 that spool to tape so there will be some backups for that client if not all.
487 3. We have around 150 clients machines it will be nice to have a option to
488 upgrade all the client machines bacula version automatically.
490 4. Atleast one connection should be reserved for the bconsole so at heavy load
491 I should connect to the director via bconsole which at sometimes I can't
493 5. Another most important feature that is missing, say at 10am I manually
494 started backup of client abc and it was a full backup since client abc has
495 no backup history and at 10.30am bacula again automatically started backup of
496 client abc as that was in the schedule. So now we have 2 multiple Full
497 backups of the same client and if we again try to start a full backup of
498 client backup abc bacula won't complain. That should be fixed.
500 - Fix bpipe.c so that it does not modify results pointer.
501 ***FIXME*** calling sequence should be changed.
502 - For Windows disaster recovery see http://unattended.sf.net/
503 - regardless of the retention period, Bacula will not prune the
504 last Full, Diff, or Inc File data until a month after the
505 retention period for the last Full backup that was done.
506 - update volume=xxx --- add status=Full
507 - Remove old spool files on startup.
508 - Exclude SD spool/working directory.
509 - Refuse to prune last valid Full backup. Same goes for Catalog.
511 - Make a callback when Rerun failed levels is called.
512 - Give Python program access to Scheduled jobs.
513 - Add setting Volume State via Python.
514 - Python script to save with Python, not save, save with Bacula.
515 - Python script to do backup.
517 - Change the Priority, Client, Storage, JobStatus (error)
518 at the start of a job.
519 - Why is SpoolDirectory = /home/bacula/spool; not reported
520 as an error when writing a DVD?
521 - Make bootstrap file handle multiple MediaTypes (SD)
522 - Remove all old Device resource code in Dir and code to pass it
523 back in SD -- better, rework it to pass back device statistics.
524 - Check locking of resources -- be sure to lock devices where previously
525 resources were locked.
526 - The last part is left in the spool dir.
529 - In restore don't compare byte count on a raw device -- directory
530 entry does not contain bytes.
534 MA = (last_MA * 3 + rate) / 4
535 rate = (bytes - last_bytes) / (runtime - last_runtime)
536 - Max Vols limit in Pool off by one?
537 - Implement Files/Bytes,... stats for restore job.
538 - Implement Total Bytes Written, ... for restore job.
539 - Despool attributes simultaneously with data in a separate
540 thread, rejoined at end of data spooling.
541 - 7. Implement new Console commands to allow offlining/reserving drives,
542 and possibly manipulating the autochanger (much asked for).
543 - Add start/end date editing in messages (%t %T, %e?) ...
544 - Add ClientDefs similar to JobDefs.
545 - Print more info when bextract -p accepts a bad block.
546 - Fix FD JobType to be set before RunBeforeJob in FD.
547 - Look at adding full Volume and Pool information to a Volume
548 label so that bscan can get *all* the info.
549 - If the user puts "Purge Oldest Volume = yes" or "Recycle Oldest Volume = yes"
550 and there is only one volume in the pool, refuse to do it -- otherwise
551 he fills the Volume, then immediately starts reusing it.
552 - Implement copies and stripes.
553 - Add history file to console.
554 - Each file on tape creates a JobMedia record. Peter has 4 million
555 files spread over 10000 tape files and four tapes. A restore takes
556 16 hours to build the restore list.
557 - Add and option to check if the file size changed during backup.
558 - Make sure SD deletes spool files on error exit.
559 - Delete old spool files when SD starts.
560 - When labeling tapes, if you enter 000026, Bacula uses
561 the tape index rather than the Volume name 000026.
562 - Add offline tape command to Bacula console.
564 Enter MediaId or Volume name: 32
565 Enter new Volume name: DLT-20Dec04
566 Automatically selected Pool: Default
567 Connecting to Storage daemon DLTDrive at 192.168.68.104:9103 ...
568 Sending relabel command from "DLT-28Jun03" to "DLT-20Dec04" ...
569 block.c:552 Write error at 0:0 on device /dev/nst0. ERR=Bad file descriptor.
570 Error writing final EOF to tape. This tape may not be readable.
571 dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
572 askdir.c:219 NULL Volume name. This shouldn't happen!!!
573 3912 Failed to label Volume: ERR=dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
574 Label command failed for Volume DLT-20Dec04.
575 Do not forget to mount the drive!!!
576 - Bug: if a job is manually scheduled to run later, it does not appear
577 in any status report and cannot be cancelled.
579 ==== Keeping track of deleted/new files ====
580 - To mark files as deleted, run essentially a Verify to disk, and
581 when a file is found missing (MarkId != JobId), then create
582 a new File record with FileIndex == -1. This could be done
583 by the FD at the same time as the backup.
585 My "trick" for keeping track of deletions is the following.
586 Assuming the user turns on this option, after all the files
587 have been backed up, but before the job has terminated, the
588 FD will make a pass through all the files and send their
589 names to the DIR (*exactly* the same as what a Verify job
590 currently does). This will probably be done at the same
591 time the files are being sent to the SD avoiding a second
592 pass. The DIR will then compare that to what is stored in
593 the catalog. Any files in the catalog but not in what the
594 FD sent will receive a catalog File entry that indicates
595 that at that point in time the file was deleted. This
596 either transmitted to the FD or simultaneously computed in
597 the FD, so that the FD can put a record on the tape that
598 indicates that the file has been deleted at this point.
599 A delete file entry could potentially be one with a FileIndex
600 of 0 or perhaps -1 (need to check if FileIndex is used for
601 some other thing as many of the Bacula fields are "overloaded"
604 During a restore, any file initially picked up by some
605 backup (Full, ...) then subsequently having a File entry
606 marked "delete" will be removed from the tree, so will not
607 be restored. If a file with the same name is later OK it
608 will be inserted in the tree -- this already happens. All
609 will be consistent except for possible changes during the
612 Since I'm on the subject, some of you may be wondering what
613 the utility of the in memory tree is if you are going to
614 restore everything (at least it comes up from time to time
615 on the list). Well, it is still *very* useful because it
616 allows only the last item found for a particular filename
617 (full path) to be entered into the tree, and thus if a file
618 is backed up 10 times, only the last copy will be restored.
619 I recently (last Friday) restored a complete directory, and
620 the Full and all the Differential and Incremental backups
621 spanned 3 Volumes. The first Volume was not even mounted
622 because all the files had been updated and hence backed up
623 since the Full backup was made. In this case, the tree
624 saved me a *lot* of time.
626 Make sure this information is stored on the tape too so
627 that it can be restored directly from the tape.
629 All the code (with the exception of formally generating and
630 saving the delete file entries) already exists in the Verify
631 Catalog command. It explicitly recognizes added/deleted files since
632 the last InitCatalog. It is more or less a "simple" matter of
633 taking that code and adapting it slightly to work for backups.
635 Comments from Martin Simmons (I think they are all covered):
636 Ok, that should cover the basics. There are few issues though:
638 - Restore will depend on the catalog. I think it is better to include the
639 extra data in the backup as well, so it can be seen by bscan and bextract.
641 - I'm not sure if it will preserve multiple hard links to the same inode. Or
642 maybe adding or removing links will cause the data to be dumped again?
644 - I'm not sure if it will handle renamed directories. Possibly it will work
645 by dumping the whole tree under a renamed directory?
647 - It remains to be seen how the backup performance of the DIR's will be
648 affected when comparing the catalog for a large filesystem.
652 How about introducing a Type = MgmtPolicy job type? That job type would
653 be responsible for scanning the Bacula environment looking for specific
654 conditions, and submitting the appropriate jobs for implementing said
658 Name = "Migration-Policy"
660 Policy Selection Job Type = Migrate
661 Scope = "<keyword> <operator> <regexp>"
662 Threshold = "<keyword> <operator> <regexp>"
663 Job Template = <template-name>
666 Where <keyword> is any legal job keyword, <operator> is a comparison
667 operator (=,<,>,!=, logical operators AND/OR/NOT) and <regexp> is a
668 appropriate regexp. I could see an argument for Scope and Threshold
669 being SQL queries if we want to support full flexibility. The
670 Migration-Policy job would then get scheduled as frequently as a site
671 felt necessary (suggested default: every 15 minutes).
676 Name = "Migration-Policy"
678 Policy Selection Job Type = Migration
680 Threshold = "Migration Selection Type = LowestUtil"
681 Job Template = "MigrationTemplate"
684 would select all pools for examination and generate a job based on
685 MigrationTemplate to automatically select the volume with the lowest
686 usage and migrate it's contents to the nextpool defined for that pool.
688 This policy abstraction would be really handy for adjusting the behavior
689 of Bacula according to site-selectable criteria (one thing that pops
690 into mind is Amanda's ability to automatically adjust backup levels
691 depending on various criteria).
697 - Add Pool/Storage override regression test.
698 - Add delete JobId to regression.
699 - Add a regression test for dbcheck.
700 - New test to add bscan to four-concurrent-jobs regression,
701 i.e. after the four-concurrent jobs zap the
702 database as is done in the bscan-test, then use bscan to
703 restore the database, do a restore and compare with the
705 - Add restore of specific JobId to regression (item 3
706 on the restore prompt)
707 - Add IPv6 to regression
708 - Add database test to regression. Test each function like delete,
711 - AntiVir can slow down backups on Win32 systems.
712 - Win32 systems with FAT32 can be much slower than NTFS for
713 more than 1000 files per directory.
717 - A HOLD command to stop all jobs from starting.
718 - A PAUSE command to pause all running jobs ==> release the
720 - Media Type = LTO,LTO-2,LTO-3
721 Media Type Read = LTO,LTO2,LTO3
722 Media Type Write = LTO2, LTO3
724 === From Carsten Menke <bootsy52@gmx.net>
726 Following is a list of what I think in the situations where I'm faced with,
727 could be a usefull enhancement to bacula, which I'm certain other users will
728 benefit from as well.
730 1. NextJob/NextJobs Directive within a Job Resource in the form of
731 NextJobs = job1,job2.
734 I currently solved the problem with running multiple jobs each after each
735 by setting the Max Wait Time for a job to 8 hours, and give
736 the jobs different Priorities. However, there scenarios where
737 1 Job is directly depending on another job, so if the former job fails,
738 the job after it needn't to be run
739 while maybe other jobs should run despite of that
742 A Backup Job and a Verify job, if the backup job fails there is no need to run
743 the verify job, as the backup job already failed. However, one may like
744 to backup the Catalog to disk despite of that the main backup job failed.
747 I see that this is related to the Event Handlers which are on the ToDo
748 list, also it is maybe a good idea to check for the return value and
749 execute different actions based on the return value
752 3. offline capability to bconsole
755 Currently I use a script which I execute within the last Job via the
756 RunAfterJob Directive, to release and eject the tape.
757 So I have to call bconsole "release=Storage-Name" and afterwards
758 mt -f /dev/nst0 eject to get the tape out.
760 If I have multiple Storage Devices, than these may not be /dev/nst0 and
761 I have to modify the script or call it with parameters etc.
762 This would actually not be needed, as everything is already defined
763 in bacula-sd.conf and if I can invoke bconsole with the
764 storage name via $1 in the script than I'm done and information is
767 4. %s for Storage Name added to the chars being substituted in "RunAfterJob"
771 For the reason mentioned in 3. to have the ability to call a
772 script with /scripts/foobar %s and in the script use $1
773 to pass the Storage Name to bconsole
775 5. Setting Volume State within a Job Resource
778 Instead of using "Maximum Volume Jobs" in the Pool Resource,
779 I would have the possibilty to define
780 in a Job Resource that after this certain job is run, the Volume State
781 should be set to "Volume State = Used", this give more flexibility (IMHO).
783 6. Localization of Bacula Messages
786 Unfortunatley many,many people I work with don't speak english very well.
787 So if at least the Reporting messages would be localized then they
788 would understand that they have to change the tape,etc. etc.
790 I volunteer to do the german translations, and if I can convince my wife also
791 french and Morre (western african language).
793 7. OK, this is evil, probably bound to security risks and maybe not possible
794 due to the design of bacula.
796 Implementation of Backtics ( `command` ) for shell comand execution to
797 the "Label Format" Directive.
801 Currently I have defined BACULA_DAY_OF_WEEK="day1|day2..." resulting in
802 Label Format = "HolyBackup-${BACULA_DAY_OF_WEEK[${WeekDay}]}". If I could
803 use backticks than I could use "Label Format = HolyBackup-`date +%A` to have
804 the localized name for the day of the week appended to the
805 format string. Then I have the tape labeled automatically with weekday
806 name in the correct language.
808 - Yes, that is surely the case. I probably should turn those into Warning
809 errors. In addition, you just made me think that it might not be bad to
810 add an option to check the file size after backing up the file and
811 report if it changes. This would be done as an option because it would
814 Kern, good idea. If you do do that, mention in the output: file
815 shrunk, or file expanded, just to make it obvious to the user
816 (without having to the refer to file size), just how the file size
819 Would this option be for all file, or just one file? Or a fileset?
820 - Make output from status use html table tags for nicely
821 presenting in a browser.
822 - Can one write tapes faster with 8192 byte block sizes?
823 - Document security problems with the same password for everyone in
824 rpm and Win32 releases.
825 - Browse generations of files.
826 - I've seen an error when my catalog's File table fills up. I
827 then have to recreate the File table with a larger maximum row
828 size. Relevant information is at
829 http://dev.mysql.com/doc/mysql/en/Full_table.html ; I think the
830 "Installing and Configuring MySQL" chapter should talk a bit
831 about this potential problem, and recommend a solution.
832 - For Solaris must use POSIX awk.
833 - Want speed of writing to tape while despooling.
834 - Supported autochanger:
842 Wangtek 6525ES (SCSI-1 QIC drive, 525MB), under Linux 2.4.something,
843 bacula 1.36.0/1 works with blocksize 16k INSIDE bacula-sd.conf.
844 - Add regex from http://www.pcre.org to Bacula for Win32.
845 - Use only shell tools no make in CDROM package.
846 - Include within include does it work?
847 - Implement a Pool of type Cleaning?
848 - Implement VolReadTime and VolWriteTime in SD
849 - Modify Backing up Your Database to include a bootstrap file.
850 - Think about making certain database errors fatal.
851 - Look at correcting the time jump in the scheduler for daylight
852 savings time changes.
853 - Add a "real" timer to network connections.
854 - Promote to Full = Time period
855 - Check dates entered by user for correctness (month/day/... ranges)
856 - Compress restore Volume listing by date and first file.
857 - Look at patches/bacula_db.b2z postgresql that loops during restore.
859 - Perhaps add read/write programs and/or plugins to FileSets.
860 - How to handle backing up portables ...
861 - Add some sort of guaranteed Interval for upgrading jobs.
862 - Can we write the state file after every job terminates? On Win32
863 the system crashes and the state file is not updated.
866 Documentation to do: (any release a little bit at a time)
867 - Doc to do unmount before removing magazine.
868 - Alternative to static linking "ldd prog" save all binaries listed,
869 restore them and point LD_LIBRARY_PATH to them.
870 - Document add "</dev/null >/dev/null 2>&1" to the bacula-fd command line
871 - Document query file format.
872 - Add more documentation for bsr files.
873 - Document problems with Verify and pruning.
874 - Document how to use multiple databases.
875 - VXA drives have a "cleaning required"
876 indicator, but Exabyte recommends preventive cleaning after every 75
879 In this context, it should be noted that Exabyte has a command-line
880 vxatool utility available for free download. (The current version is
881 vxatool-3.72.) It can get diagnostic info, read, write and erase tapes,
882 test the drive, unload tapes, change drive settings, flash new firmware,
884 Of particular interest in this context is that vxatool <device> -i will
885 report, among other details, the time since last cleaning in tape motion
886 minutes. This information can be retrieved (and settings changed, for
887 that matter) through the generic-SCSI device even when Bacula has the
888 regular tape device locked. (Needless to say, I don't recommend
889 changing tape settings while a job is running.)
890 - Lookup HP cleaning recommendations.
891 - Lookup HP tape replacement recommendations (see trouble shooting autochanger)
892 - Document doing table repair
895 ===================================
896 - Add macro expansions in JobDefs.
897 Run Before Job = "SomeFile %{Level} %{Client}"
898 Write Bootstrap="/some/dir/%{JobName}_%{Client}.bsr"
899 - Use non-blocking network I/O but if no data is available, use
901 - Use gather write() for network I/O.
902 - Autorestart on crash.
903 - Add bandwidth limiting.
904 - Add acks every once and a while from the SD to keep
905 the line from timing out.
906 - When an error in input occurs and conio beeps, you can back
907 up through the prompt.
908 - Detect fixed tape block mode during positioning by looking at
909 block numbers in btape "test". Possibly adjust in Bacula.
910 - Fix list volumes to output volume retention in some other
911 units, perhaps via a directive.
912 - Allow Simultaneous Priorities = yes => run up to Max concurrent jobs even
913 with multiple priorities.
914 - If you use restore replace=never, the directory attributes for
915 non-existent directories will not be restored properly.
917 - see lzma401.zip in others directory for new compression
919 - Allow the user to select JobType for manual pruning/purging.
920 - bscan does not put first of two volumes back with all info in
922 - Implement the FreeBSD nodump flag in chflags.
923 - Figure out how to make named console messages go only to that
924 console and to the non-restricted console (new console class?).
925 - Make restricted console prompt for password if *ask* is set or
926 perhaps if password is undefined.
927 - Implement "from ISO-date/time every x hours/days/weeks/months" in
930 ==== from Marc Schoechlin
931 - the help-command should be more verbose
932 (it should explain the paramters of the different
934 -> its time-comsuming to consult the manual anytime
935 you need a special parameter
936 -> maybe its more easy to maintain this, if the
937 descriptions of that commands are outsourced to
939 - the cd-command should allow complete paths
940 i.e. cd /foo/bar/foo/bar
941 -> if a customer mails me the path to a certain file,
942 its faster to enter the specified directory
943 - if the password is not configured in bconsole.conf
944 you should be asked for it.
945 -> sometimes you like to do restore on a customer-machine
946 which shouldnt know the password for bacula.
947 -> adding the password to the file favours admins
948 to forget to remove the password after usage
950 the protection of that file is less important
951 - long-listed-output of commands should be scrollable
952 like the unix more/less-command does
953 -> if someone runs 200 and more machines, the lists could
954 be a little long and complex
955 - command-output should be shown column by column
956 to reduce scrolling and to increase clarity
958 - lsmark should list the selected files with full
960 - wildcards for selecting and file and directories would be nice
961 - any actions should be interuptable with STRG+C
962 - command-expansion would be pretty cool
964 - When the replace Never option is set, new directory permissions
965 are not restored. See bug 213. To fix this requires creating a
966 list of newly restored directories so that those directory
967 permissions *can* be restored.
968 - Add prune all command
969 - Document fact that purge can destroy a part of a restore by purging
970 one volume while others remain valid -- perhaps mark Jobs.
971 - Add multiple-media-types.txt
972 - look at mxt-changer.html
973 - Make ? do a help command (no return needed).
974 - Implement restore directory.
975 - Document streams and how to implement them.
976 - Try not to re-backup a file if a new hard link is added.
977 - Add feature to backup hard links only, but not the data.
978 - Fix stream handling to be simpler.
979 - Add Priority and Bootstrap to Run a Job.
980 - Eliminate Restore "Run Restore Job" prompt by allowing new "run command
982 - Remove View FileSet button from Run a Job dialog.
983 - Handle prompt for restore job at end of Restore command.
984 - Add display of total selected files to Restore window.
985 - Add tree pane to left of window.
986 - Add progress meter.
987 - Max wait time or max run time causes seg fault -- see runtime-bug.txt
988 - Add message to user to check for fixed block size when the forward
989 space test fails in btape.
990 - When unmarking a directory check if all files below are unmarked and
991 then remove the + flag -- in the restore tree.
992 - Possibly implement: Action = Unmount Device="TapeDrive1" in Admin jobs.
993 - Setup lrrd graphs: (http://www.linpro.no/projects/lrrd/) Mike Acar.
994 - Revisit the question of multiple Volumes (disk) on a single device.
995 - Add a block copy option to bcopy.
996 - Finish work on Gnome restore GUI.
997 - Fix "llist jobid=xx" where no fileset or client exists.
998 - For each job type (Admin, Restore, ...) require only the really necessary
999 fields.- Pass Director resource name as an option to the Console.
1000 - Add a "batch" mode to the Console (no unsolicited queries, ...).
1001 - Add a .list all files in the restore tree (probably also a list all files)
1002 Do both a long and short form.
1003 - Allow browsing the catalog to see all versions of a file (with
1004 stat data on each file).
1005 - Restore attributes of directory if replace=never set but directory
1007 - Use SHA1 on authentication if possible.
1008 - See comtest-xxx.zip for Windows code to talk to USB.
1009 - Add John's appended files:
1010 Appended = { /files/server/logs/http/*log }
1011 and such files would be treated as follows.On a FULL backup, they would
1012 be backed up like any other file.On an INCREMENTAL backup, where a
1013 previous INCREMENTAL or FULL was already in thecatalogue and the length
1014 of the file wasgreater than the length of the last backup, only thedata
1015 added since the last backup will be dumped.On an INCREMENTAL backup, if
1016 the length of the file is less than thelength of the file with the same
1017 name last backed up, the completefile is dumped.On Windows systems, with
1018 creation date of files, we can be evensmarter about this and not count
1019 entirely upon the length.On a restore, the full and all incrementals
1020 since it will beapplied in sequence to restore the file.
1021 - Check new HAVE_WIN32 open bits.
1022 - Check if the tape has moved before writing.
1023 - Handling removable disks -- see below:
1024 - Keep track of tape use time, and report when cleaning is necessary.
1025 - Add FromClient and ToClient keywords on restore command (or
1026 BackupClient RestoreClient).
1027 - Implement a JobSet, which groups any number of jobs. If the
1028 JobSet is started, all the jobs are started together.
1029 Allow Pool, Level, and Schedule overrides.
1030 - Enhance cancel to timeout BSOCK packets after a specific delay.
1031 - Do scheduling by UTC using gmtime_r() in run_conf, scheduler, and
1032 ua_status.!!! Thanks to Alan Brown for this tip.
1033 - Look at updating Volume Jobs so that Max Volume Jobs = 1 will work
1034 correctly for multiple simultaneous jobs.
1035 - Correct code so that FileSet MD5 is calculated for < and | filename
1037 - Implement the Media record flag that indicates that the Volume does disk
1039 - Implement VolAddr, which is used when Volume is addressed like a disk,
1040 and form it from VolFile and VolBlock.
1041 - Make multiple restore jobs for multiple media types specifying
1042 the proper storage type.
1043 - Fix fast block rejection (stored/read_record.c:118). It passes a null
1044 pointer (rec) to try_repositioning().
1045 - Look at extracting Win data from BackupRead.
1046 - Implement RestoreJobRetention? Maybe better "JobRetention" in a Job,
1047 which would take precidence over the Catalog "JobRetention".
1048 - Implement Label Format in Add and Label console commands.
1049 - Possibly up network buffers to 65K. Put on variable.
1050 - Put email tape request delays on one or more variables. User wants
1051 to cancel the job after a certain time interval. Maximum Mount Wait?
1052 - Job, Client, Device, Pool, or Volume?
1053 Is it possible to make this a directive which is *optional* in multiple
1054 resources, like Level? If so, I think I'd make it an optional directive
1055 in Job, Client, and Pool, with precedence such that Job overrides Client
1056 which in turn overrides Pool.
1058 - New Storage specifications:
1059 - Want to write to multiple storage devices simultaneously
1060 - Want to write to multiple storage devices sequentially (in one job)
1061 - Want to read/write simultaneously
1062 - Key is MediaType -- it must match
1064 Passed to SD as a sort of BSR record called Storage Specification
1068 MediaType -> Next MediaType
1070 Device -> Next Device
1072 Allow multiple Storage specifications
1080 Allow Multiple Pool specifications (note, Pool currently
1082 Allow Multiple MediaType specifications in Dir conf
1083 Allow Multiple Device specifications in Dir conf
1084 Perhaps keep this in a single SSR
1085 Tie a Volume to a specific device by using a MediaType that
1086 is contained in only one device.
1087 In SD allow Device to have Multiple MediaTypes
1089 - Ideas from Jerry Scharf:
1090 First let's point out some big pluses that bacula has for this
1092 more importantly it's active. Thank you so much for that
1093 even more important, it's not flaky
1094 it has an open access catalog, opening many possibilities
1095 it's pushing toward heterogeneous systems capability
1097 Macintosh file client
1098 macs are an interesting niche, but I fear a server is a rathole
1099 working bare iron recovery for windows
1100 the option for inc/diff backups not reset on fileset revision
1101 a) use both change and inode update time against base time
1102 b) do the full catalog check (expensive but accurate)
1103 sizing guide (how much system is needed to back up N systems/files)
1104 consultants on using bacula in building a disaster recovery system
1105 an integration guide
1106 or how to get at fancy things that one could do with bacula
1107 logwatch code for bacula logs (or similar)
1108 linux distro inclusion of bacula (brings good and bad, but necessary)
1109 win2k/XP server capability (icky but you asked)
1110 support for Oracle database ??
1112 - Look at adding SQL server and Exchange support for Windows.
1113 - Make dev->file and dev->block_num signed integers so that -1 can
1114 be an invalid value which happens with BSR.
1115 - Create VolAddr for disk files in place of VolFile and VolBlock. This
1116 is needed to properly specify ranges.
1117 - Add progress of files/bytes to SD and FD.
1118 - Print warning message if FileId > 4 billion
1119 - do a "messages" before the first prompt in Console
1120 - Client does not show busy during Estimate command.
1121 - Implement Console mtx commands.
1122 - Implement a Mount Command and an Unmount Command where
1123 the users could specify a system command to be performed
1124 to do the mount, after which Bacula could attempt to
1125 read the device. This is for Removeable media such as a CDROM.
1126 - Most likely, this mount command would be invoked explicitly
1127 by the user using the current Console "mount" and "unmount"
1128 commands -- the Storage Daemon would do the right thing
1129 depending on the exact nature of the device.
1130 - As with tape drives, when Bacula wanted a new removable
1131 disk mounted, it would unmount the old one, and send a message
1132 to the user, who would then use "mount" as described above
1133 once he had actually inserted the disk.
1134 - Implement dump/print label to UA
1135 - Spool to disk only when the tape is full, then when a tape is hung move
1137 - bextract is sending everything to the log file ****FIXME****
1138 - Allow multiple Storage specifications (or multiple names on
1139 a single Storage specification) in the Job record. Thus a job
1140 can be backed up to a number of storage devices.
1141 - Implement some way for the File daemon to contact the Director
1142 to start a job or pass its DHCP obtained IP number.
1143 - Implement a query tape prompt/replace feature for a console
1144 - Copy console @ code to gnome2-console
1145 - Make tree walk routines like cd, ls, ... more user friendly
1146 by handling spaces better.
1147 - Make sure that Bacula rechecks the tape after the 20 min wait.
1148 - Set IO_NOWAIT on Bacula TCP/IP packets.
1149 - Try doing a raw partition backup and restore by mounting a
1151 - From Lars Kellers:
1152 Yes, it would allow to highly automatic the request for new tapes. If a
1153 tape is empty, bacula reads the barcodes (native or simulated), and if
1154 an unused tape is found, it runs the label command with all the
1155 necessary parameters.
1157 By the way can bacula automatically "move" an empty/purged volume say
1158 in the "short" pool to the "long" pool if this pool runs out of volume
1160 - What to do about "list files job=xxx".
1161 - Look at how fuser works and /proc/PID/fd that is how Nic found the
1162 file descriptor leak in Bacula.
1163 - Implement WrapCounters in Counters.
1164 - Add heartbeat from FD to SD if hb interval expires.
1165 - Can we dynamically change FileSets?
1166 - If pool specified to label command and Label Format is specified,
1167 automatically generate the Volume name.
1168 - Why can't SQL do the filename sort for restore?
1169 - Add ExhautiveRestoreSearch
1170 - Look at the possibility of loading only the necessary
1171 data into the restore tree (i.e. do it one directory at a
1172 time as the user walks through the tree).
1173 - Possibly use the hash code if the user selects all for a restore command.
1174 - Fix "restore all" to bypass building the tree.
1175 - Prohibit backing up archive device (findlib/find_one.c:128)
1176 - Implement Release Device in the Job resource to unmount a drive.
1177 - Implement Acquire Device in the Job resource to mount a drive,
1178 be sure this works with admin jobs so that the user can get
1179 prompted to insert the correct tape. Possibly some way to say to
1180 run the job but don't save the files.
1181 - Make things like list where a file is saved case independent for
1183 - Use autochanger to handle multiple devices.
1184 - Implement a Recycle command
1185 - Start working on Base jobs.
1186 - Implement UnsavedFiles DB record.
1187 - From Phil Stracchino:
1188 It would probably be a per-client option, and would be called
1189 something like, say, "Automatically purge obsoleted jobs". What it
1190 would do is, when you successfully complete a Differential backup of a
1191 client, it would automatically purge all Incremental backups for that
1192 client that are rendered redundant by that Differential. Likewise,
1193 when a Full backup on a client completed, it would automatically purge
1194 all Differential and Incremental jobs obsoleted by that Full backup.
1195 This would let people minimize the number of tapes they're keeping on
1196 hand without having to master the art of retention times.
1197 - When doing a Backup send all attributes back to the Director, who
1198 would then figure out what files have been deleted.
1199 - Currently in mount.c:236 the SD simply creates a Volume. It should have
1200 explicit permission to do so. It should also mark the tape in error
1201 if there is an error.
1202 - Cancel waiting for Client connect in SD if FD goes away.
1204 - Implement timeout in response() when it should come quickly.
1205 - Implement a Slot priority (loaded/not loaded).
1206 - Implement "vacation" Incremental only saves.
1207 - Implement create "FileSet"?
1208 - Add prefixlinks to where or not where absolute links to FD.
1209 - Issue message to mount a new tape before the rewind.
1210 - Simplified client job initiation for portables.
1211 - If SD cannot open a drive, make it periodically retry.
1212 - Add more of the config info to the tape label.
1214 - Refine SD waiting output:
1215 Device is being positioned
1216 > Device is being positioned for append
1217 > Device is being positioned to file x
1219 - Figure out some way to estimate output size and to avoid splitting
1220 a backup across two Volumes -- this could be useful for writing CDROMs
1221 where you really prefer not to have it split -- not serious.
1222 - Have SD compute MD5 or SHA1 and compare to what FD computes.
1223 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
1224 actual data on the Volume and compare it.
1225 - Implement Bacula plugins -- design API
1226 - Make bcopy read through bad tape records.
1227 - Program files (i.e. execute a program to read/write files).
1228 Pass read date of last backup, size of file last time.
1229 - Add Signature type to File DB record.
1230 - CD into subdirectory when open()ing files for backup to
1231 speed up things. Test with testfind().
1232 - Priority job to go to top of list.
1233 - Why are save/restore of device different sizes (sparse?) Yup! Fix it.
1234 - Implement some way for the Console to dynamically create a job.
1235 - Solaris -I on tar for include list
1236 - Need a verbose mode in restore, perhaps to bsr.
1237 - bscan without -v is too quiet -- perhaps show jobs.
1238 - Add code to reject whole blocks if not wanted on restore.
1239 - Check if we can increase Bacula FD priorty in Win2000
1240 - Make sure the MaxVolFiles is fully implemented in SD
1241 - Check if both CatalogFiles and UseCatalog are set to SD.
1242 - Possibly add email to Watchdog if drive is unmounted too
1243 long and a job is waiting on the drive.
1244 - After unmount, if restore job started, ask to mount.
1245 - Add UA rc and history files.
1246 - put termcap (used by console) in ./configure and
1247 allow -with-termcap-dir.
1248 - Fix Autoprune for Volumes to respect need for full save.
1249 - Compare tape to Client files (attributes, or attributes and data)
1250 - Make all database Ids 64 bit.
1251 - Allow console commands to detach or run in background.
1252 - Add SD message variables to control operator wait time
1253 - Maximum Operator Wait
1254 - Minimum Message Interval
1255 - Maximum Message Interval
1256 - Send Operator message when cannot read tape label.
1257 - Verify level=Volume (scan only), level=Data (compare of data to file).
1258 Verify level=Catalog, level=InitCatalog
1260 - Add keyword search to show command in Console.
1261 - Events : tape has more than xxx bytes.
1262 - Complete code in Bacula Resources -- this will permit
1263 reading a new config file at any time.
1264 - Handle ctl-c in Console
1265 - Implement script driven addition of File daemon to config files.
1266 - Think about how to make Bacula work better with File (non-tape) archives.
1267 - Write Unix emulator for Windows.
1268 - Put memory utilization in Status output of each daemon
1269 if full status requested or if some level of debug on.
1270 - Make database type selectable by .conf files i.e. at runtime
1271 - Set flag for uname -a. Add to Volume label.
1272 - Restore files modified after date
1273 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
1274 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
1275 - Timout a job or terminate if link goes down, or reopen link and query.
1276 - Concept of precious tapes (cannot be reused).
1277 - Make bcopy copy with a single tape drive.
1278 - Permit changing ownership during restore.
1281 > My suggestion: Add a feature on the systray menu-icon menu to request
1282 > an immediate backup now. This would be useful for laptop users who may
1283 > not be on the network when the regular scheduled backup is run.
1285 > My wife's suggestion: Add a setting to the win32 client to allow it to
1286 > shut down the machine after backup is complete (after, of course,
1287 > displaying a "System will shut down in one minute, click here to cancel"
1288 > warning dialog). This would be useful for sites that want user
1289 > woorkstations to be shut down overnight to save power.
1292 - Autolabel should be specified by DIR instead of SD.
1294 - Add media capacity
1295 - AutoScan (check checksum of tape)
1296 - Format command = "format /dev/nst0"
1300 - Seek resolution (usually corresponds to buffer size)
1301 - EODErrorCode=ENOSPC or code
1302 - Partial Read error code
1303 - Partial write error code
1304 - Nonformatted read error
1305 - Nonformatted write error
1306 - WriteProtected error
1310 - IgnoreCloseErrors=yes
1320 - FD sends unsaved file list to Director at end of job (see
1322 - File daemon should build list of files skipped, and then
1323 at end of save retry and report any errors.
1324 - Write a Storage daemon that uses pipes and
1325 standard Unix programs to write to the tape.
1327 - Need something that monitors the JCR queue and
1328 times out jobs by asking the deamons where they are.
1329 - Enhance Jmsg code to permit buffering and saving to disk.
1330 - device driver = "xxxx" for drives.
1331 - Verify from Volume
1332 - Ensure that /dev/null works
1333 - Need report class for messages. Perhaps
1334 report resource where report=group of messages
1335 - enhance scan_attrib and rename scan_jobtype, and
1336 fill in code for "since" option
1337 - Director needs a time after which the report status is sent
1338 anyway -- or better yet, a retry time for the job.
1339 - Don't reschedule a job if previous incarnation is still running.
1340 - Some way to automatically backup everything is needed????
1341 - Need a structure for pending actions:
1343 - termination status (part of buffered msgs?)
1345 Read, Write, Clean, Delete
1346 - Login to Bacula; Bacula users with different permissions:
1347 owner, group, user, quotas
1348 - Store info on each file system type (probably in the job header on tape.
1349 This could be the output of df; or perhaps some sort of /etc/mtab record.
1351 ========= ideas ===============
1352 From: "Jerry K. Schieffer" <jerry@skylinetechnology.com>
1353 To: <kern@sibbald.com>
1354 Subject: RE: [Bacula-users] future large programming jobs
1355 Date: Thu, 26 Feb 2004 11:34:54 -0600
1357 I noticed the subject thread and thought I would offer the following
1358 merely as sources of ideas, i.e. something to think about, not even as
1359 strong as a request. In my former life (before retiring) I often
1360 dealt with backups and storage management issues/products as a
1361 developer and as a consultant. I am currently migrating my personal
1362 network from amanda to bacula specifically because of the ability to
1363 cross media boundaries during storing backups.
1364 Are you familiar with the commercial product called ADSM (I think IBM
1365 now sells it under the Tivoli label)? It has a couple of interesting
1366 ideas that may apply to the following topics.
1368 1. Migration: Consider that when you need to restore a system, there
1369 may be pressure to hurry. If all the information for a single client
1370 can eventually end up on the same media (and in chronological order),
1371 the restore is facillitated by not having to search past information
1372 from other clients. ADSM has the concept of "client affinity" that
1373 may be associated with it's storage pools. It seems to me that this
1374 concept (as an optional feature) might fit in your architecture for
1377 ADSM also has the concept of defining one or more storage pools as
1378 "copy pools" (almost mirrors, but only in the sense of contents).
1379 These pools provide the ability to have duplicte data stored both
1380 onsite and offsite. The copy process can be scheduled to be handled
1381 by their storage manager during periods when there is no backup
1382 activity. Again, the migration process might be a place to consider
1383 implementing something like this.
1386 > It strikes me that it would be very nice to be able to do things
1388 > have the Job(s) backing up the machines run, and once they have all
1389 > completed, start a migration job to copy the data from disks Volumes
1391 > a tape library and then to offsite storage. Maybe this can already
1393 > done with some careful scheduling and Job prioritzation; the events
1394 > mechanism described below would probably make it very easy.
1396 This is the goal. In the first step (before events), you simply
1398 the Migration to tape later.
1400 2. Base jobs: In ADSM, each copy of each stored file is tracked in
1401 the database. Once a file (unique by path and metadata such as dates,
1402 size, ownership, etc.) is in a copy pool, no more copies are made. In
1403 other words, when you start ADSM, it begins like your concept of a
1404 base job. After that it is in the "incremental" mode. You can
1405 configure the number of "generations" of files to be retained, plus a
1406 retention date after which even old generations are purged. The
1407 database tracks the contents of media and projects the percentage of
1408 each volume that is valid. When the valid content of a volume drops
1409 below a configured percentage, the valid data are migrated to another
1410 volume and the old volume is marked as empty. Note, this requires
1411 ADSM to have an idea of the contents of a client, i.e. marking the
1412 database when an existing file was deleted, but this would solve your
1413 issue of restoring a client without restoring deleted files.
1415 This is pretty far from what bacula now does, but if you are going to
1416 rip things up for Base jobs,.....
1417 Also, the benefits of this are huge for very large shops, especially
1418 with media robots, but are a pain for shops with manual media
1422 > Base jobs sound pretty useful, but I'm not dying for them.
1424 Nobody is dying for them, but when you see what it does, you will die
1427 3. Restoring deleted files: Since I think my comments in (2) above
1428 have low probability of implementation, I'll also suggest that you
1429 could approach the issue of deleted files by a mechanism of having the
1430 fd report to the dir, a list of all files on the client for every
1431 backup job. The dir could note in the database entry for each file
1432 the date that the file was seen. Then if a restore as of date X takes
1433 place, only files that exist from before X until after X would be
1434 restored. Probably the major cost here is the extra date container in
1435 each row of the files table.
1437 Thanks for "listening". I hope some of this helps. If you want to
1438 contact me, please send me an email - I read some but not all of the
1439 mailing list traffic and might miss a reply there.
1441 Please accept my compliments for bacula. It is doing a great job for
1442 me!! I sympathize with you in the need to wrestle with excelence in
1443 execution vs. excelence in feature inclusion.
1448 ==============================
1451 - Design at hierarchial storage for Bacula. Migration and Clone.
1452 - Implement FSM (File System Modules).
1453 - Audit M_ error codes to ensure they are correct and consistent.
1454 - Add variable break characters to lex analyzer.
1455 Either a bit mask or a string of chars so that
1456 the caller can change the break characters.
1457 - Make a single T_BREAK to replace T_COMMA, etc.
1458 - Ensure that File daemon and Storage daemon can
1459 continue a save if the Director goes down (this
1460 is NOT currently the case). Must detect socket error,
1461 buffer messages for later.
1462 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
1463 - Add ability to backup to two Storage devices (two SD sessions) at
1464 the same time -- e.g. onsite, offsite.
1465 - Add the ability to consolidate old backup sets (basically do a restore
1466 to tape and appropriately update the catalog). Compress Volume sets.
1467 Might need to spool via file is only one drive is available.
1468 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
1469 someway to do so with every volume that has less than x% valid
1473 Migration: Move a backup from one Volume to another
1474 Clone: Copy a backup -- two Volumes
1477 ======================================================
1479 It is somewhat like a Full save becomes an incremental since
1480 the Base job (or jobs) plus other non-base files.
1482 - A Base backup is same as Full backup, just different type.
1483 - New BaseFiles table that contains:
1485 BaseJobId - Base JobId referenced for this FileId (needed ???)
1486 JobId - JobId currently running
1487 FileId - File not backed up, exists in Base Job
1488 FileIndex - FileIndex from Base Job.
1489 i.e. for each base file that exists but is not saved because
1490 it has not changed, the File daemon sends the JobId, BaseId,
1491 FileId, FileIndex back to the Director who creates the DB entry.
1492 - To initiate a Base save, the Director sends the FD
1493 the FileId, and full filename for each file in the Base.
1494 - When the FD finds a Base file, he requests the Director to
1495 send him the full File entry (stat packet plus MD5), or
1496 conversely, the FD sends it to the Director and the Director
1497 says yes or no. This can be quite rapid if the FileId is kept
1498 by the FD for each Base Filename.
1499 - It is probably better to have the comparison done by the FD
1500 despite the fact that the File entry must be sent across the
1502 - An alternative would be to send the FD the whole File entry
1503 from the start. The disadvantage is that it requires a lot of
1504 space. The advantage is that it requires less communications
1506 - The Job record must be updated to indicate that one or more
1508 - At end of Job, FD returns:
1509 1. Count of base files/bytes not written to tape (i.e. matches)
1510 2. Count of base file that were saved i.e. had changed.
1511 - No tape record would be written for a Base file that matches, in the
1512 same way that no tape record is written for Incremental jobs where
1513 the file is not saved because it is unchanged.
1514 - On a restore, all the Base file records must explicitly be
1515 found from the BaseFile tape. I.e. for each Full save that is marked
1516 to have one or more Base Jobs, search the BaseFile for all occurrences
1518 - An optimization might be to make the BaseFile have:
1524 This would avoid the need to explicitly fetch each File record for
1525 the Base job. The Base Job record will be fetched to get the
1526 VolSessionId and VolSessionTime.
1527 =========================================================
1530 ==========================================================
1532 For each Incremental job that is run, there may be files that
1533 were found but not saved because they were locked (this applies
1534 only to Windows). Such a system could send back to the Director
1535 a list of Unsaved files.
1537 - New UnSavedFiles table that contains:
1541 - Then in the next Incremental job, the list of Unsaved Files will be
1542 feed to the FD, who will ensure that they are explicitly chosen even
1543 if standard date/time check would not have selected them.
1544 =============================================================
1548 Multiple drive autochanger data: see Alan Brown
1549 mtx -f xxx unloadStorage Element 1 is Already Full(drive 0 was empty)
1550 Unloading Data Transfer Element into Storage Element 1...source Element
1551 Address 480 is Empty
1553 (drive 0 was empty and so was slot 1)
1554 > mtx -f xxx load 15 0
1555 no response, just returns to the command prompt when complete.
1556 > mtx -f xxx status Storage Changer /dev/changer:2 Drives, 60 Slots ( 2 Import/Export )
1557 Data Transfer Element 0:Full (Storage Element 15 Loaded):VolumeTag = HX001
1558 Data Transfer Element 1:Empty
1559 Storage Element 1:Empty
1560 Storage Element 2:Full :VolumeTag=HX002
1561 Storage Element 3:Full :VolumeTag=HX003
1562 Storage Element 4:Full :VolumeTag=HX004
1563 Storage Element 5:Full :VolumeTag=HX005
1564 Storage Element 6:Full :VolumeTag=HX006
1565 Storage Element 7:Full :VolumeTag=HX007
1566 Storage Element 8:Full :VolumeTag=HX008
1567 Storage Element 9:Full :VolumeTag=HX009
1568 Storage Element 10:Full :VolumeTag=HX010
1569 Storage Element 11:Empty
1570 Storage Element 12:Empty
1571 Storage Element 13:Empty
1572 Storage Element 14:Empty
1573 Storage Element 15:Empty
1574 Storage Element 16:Empty....
1575 Storage Element 28:Empty
1576 Storage Element 29:Full :VolumeTag=CLNU01L1
1577 Storage Element 30:Empty....
1578 Storage Element 57:Empty
1579 Storage Element 58:Full :VolumeTag=NEX261L2
1580 Storage Element 59 IMPORT/EXPORT:Empty
1581 Storage Element 60 IMPORT/EXPORT:Empty
1583 Unloading Data Transfer Element into Storage Element 15...done
1585 (just to verify it remembers where it came from, however it can be
1586 overrriden with mtx unload {slotnumber} to go to any storage slot.)
1588 There needs to be a table of drive # to devices somewhere - If there are
1589 multiple changers or drives there may not be a 1:1 correspondance between
1590 changer drive number and system device name - and depending on the way the
1591 drives are hooked up to scsi busses, they may not be linearly numbered
1592 from an offset point either.something like
1594 Autochanger drives = 2
1595 Autochanger drive 0 = /dev/nst1
1596 Autochanger drive 1 = /dev/nst2
1597 IMHO, it would be _safest_ to use explicit mtx unload commands at all
1598 times, not just for multidrive changers. For a 1 drive changer, that's
1604 MTX's manpage (1.2.15):
1605 unload [<slotnum>] [ <drivenum> ]
1606 Unloads media from drive <drivenum> into slot
1607 <slotnum>. If <drivenum> is omitted, defaults to
1608 drive 0 (as do all commands). If <slotnum> is
1609 omitted, defaults to the slot that the drive was
1610 loaded from. Note that there's currently no way
1611 to say 'unload drive 1's media to the slot it
1612 came from', other than to explicitly use that
1613 slot number as the destination.AB
1619 undef# camcontrol devlist
1620 <WANGTEK 51000 SCSI M74H 12B3> at scbus0 target 2 lun 0 (pass0,sa0)
1621 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 0 (pass1,sa1)
1622 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 1 (pass2)
1624 tapeinfo -f /dev/sg0 with a bad tape in drive 1:
1625 [kern@rufus mtx-1.2.17kes]$ ./tapeinfo -f /dev/sg0
1626 Product Type: Tape Drive
1628 Product ID: 'C5713A '
1630 Attached Changer: No
1631 TapeAlert[3]: Hard Error: Uncorrectable read/write error.
1632 TapeAlert[20]: Clean Now: The tape drive neads cleaning NOW.
1639 Medium Type: Not Loaded
1642 DataCompEnabled: yes
1643 DataCompCapable: yes
1644 DataDeCompEnabled: yes
1651 Handling removable disks
1653 From: Karl Cunningham <karlc@keckec.com>
1655 My backups are only to hard disk these days, in removable bays. This is my
1656 idea of how a backup to hard disk would work more smoothly. Some of these
1657 things Bacula does already, but I mention them for completeness. If others
1658 have better ways to do this, I'd like to hear about it.
1660 1. Accommodate several disks, rotated similar to how tapes are. Identified
1661 by partition volume ID or perhaps by the name of a subdirectory.
1662 2. Abort & notify the admin if the wrong disk is in the bay.
1663 3. Write backups to different subdirectories for each machine to be backed
1665 4. Volumes (files) get created as needed in the proper subdirectory, one
1667 5. When a disk is recycled, remove or zero all old backup files. This is
1668 important as the disk being recycled may be close to full. This may be
1669 better done manually since the backup files for many machines may be
1670 scattered in many subdirectories.
1675 - Make sure that all do_prompt() calls in Dir check for
1676 -1 (error) and -2 (cancel) returns.
1677 - Fix foreach_jcr() to have free_jcr() inside next().
1678 jcr=jcr_walk_start();
1679 for ( ; jcr; (jcr=jcr_walk_next(jcr)) )
1682 - A Volume taken from Scratch should take on the retention period
1684 - Correct doc for Maximum Changer Wait (and others) accepting only
1686 - Implement status that shows why a job is being held in reserve, or
1687 rather why none of the drives are suitable.
1688 - Implement a way to disable a drive (so you can use the second
1689 drive of an autochanger, and the first one will not be used or
1691 - Make sure Maximum Volumes is respected in Pools when adding
1692 Volumes (e.g. when pulling a Scratch volume).
1693 - Keep same dcr when switching device ...
1694 - Implement code that makes the Dir aware that a drive is an
1695 autochanger (so the user doesn't need to use the Autochanger = yes
1697 - Make catalog respect ACL.
1698 - Add recycle count to Media record.
1699 - Add initial write date to Media record.
1700 - Fix store_yesno to be store_bitmask.
1701 --- create_file.c.orig Fri Jul 8 12:13:05 2005
1702 +++ create_file.c Fri Jul 8 12:13:07 2005
1704 attr->ofname, be.strerror());
1707 + } else if(S_ISSOCK(attr->statp.st_mode)) {
1708 + Dmsg1(200, "Skipping socket: %s\n", attr->ofname);
1710 Dmsg1(200, "Restore node: %s\n", attr->ofname);
1711 if (mknod(attr->ofname, attr->statp.st_mode, attr->statp.st_rdev) != 0 && errno != EEXIST) {
1712 - Add true/false to conf same as yes/no
1713 - Reserve blocks other restore jobs when first cannot connect to SD.
1714 - Fix Maximum Changer Wait, Maximum Open Wait, Maximum Rewind Wait to
1715 accept time qualifiers.
1716 - Does ClientRunAfterJob fail the job on a bad return code?
1717 - Make hardlink code at line 240 of find_one.c use binary search.
1718 - Add ACL error messages in src/filed/acl.c.
1719 - Make authentication failures single threaded.
1720 - Make Dir and SD authentication errors single threaded.
1721 - Fix catreq.c digestbuf at line 411 in src/dird/catreq.c
1722 - Make base64.c (bin_to_base64) take a buffer length
1723 argument to avoid overruns.
1724 and verify that other buffers cannot overrun.
1725 - Implement VolumeState as discussed with Arno.
1726 - Add LocationId to update volume
1734 - Add Comment to Media record
1735 - Fix auth compatibility with 1.38
1736 - Update dbcheck to include Log table
1737 - Update llist to include new fields.
1738 - Make unmount unload autochanger. Make mount load slot.
1739 - Fix bscan to report the JobType when restoring a job.
1740 - Fix wx-console scanning problem with commas in names.
1741 - Add manpages to the list of directories for make install. Notify
1743 - Add bconsole option to use stdin/out instead of conio.
1744 - Fix ClientRunBefore/AfterJob compatibility.
1745 - Ensure that connection to daemon failure always indicates what
1746 daemon it was trying to connect to.
1747 - Freespace on DVD requested over and over even with no intervening
1749 - .update volume [enabled|disabled|*see below]
1750 > However, I could easily imagine an option to "update slots" that says
1751 > "enable=yes|no" that would automatically enable or disable all the Volumes
1752 > found in the autochanger. This will permit the user to optionally mark all
1753 > the Volumes in the magazine disabled prior to taking them offsite, and mark
1754 > them all enabled when bringing them back on site. Coupled with the options
1755 > to the slots keyword, you can apply the enable/disable to any or all volumes.
1756 - Restricted consoles start in the Default catalog even if it
1758 - When reading through parts on the DVD, the DVD is mounted and
1759 unmounted for each part.
1760 - Make sure that the restore options don't permit "seeing" other
1762 - Restore of a raw drive should not try to check the volume size.
1763 - Lock tape drive door when open()
1764 - Make release unload any autochanger.
1765 - Arno's reservation deadlock.
1767 - Make sure the new level=Full syntax is used in all
1768 example conf files (especially in the manual).
1769 - Fix prog copyright (SD) all other files.
1770 - Document need for UTF-8 format
1771 - Try turning on disk seek code.
1772 - Some users claim that they must do two prune commands to get a
1773 Volume marked as purged.
1774 - Document fact that CatalogACL now needed for Tray monitor (fixed).
1775 - If you have two Catalogs, it will take the first one.
1776 - Migration Volume span bug