6 - !!! Cannot restore two jobs a the same time that were
7 written simultaneously unless they were totally spooled.
8 - Document cleaning up the spool files:
9 db, pid, state, bsr, mail, conmsg, spool
10 - Document the multiple-drive-changer.txt script.
11 - Pruning with Admin job.
12 - Does WildFile match against full name? Doc.
13 - %d and %v only valid on Director, not for ClientRunBefore/After.
14 - During tests with the 260 char fix code, I found one problem:
15 if the system "sees" a long path once, it seems to forget it's
16 working drive (e.g. c:\), which will lead to a problem during
17 the next job (create bootstrap file will fail). Here is the
18 workaround: specify absolute working and pid directory in
19 bacula-fd.conf (e.g. c:\bacula\working instead of
21 - Document techniques for restoring large numbers of files.
22 - Document setting my.cnf to big file usage.
23 - Add example of proper index output to doc. show index from File;
24 - Correct the Include syntax in the m4.xxx files in examples/conf
25 - Document JobStatus and Termination codes.
26 - Fix the error with the "DVI file can't be opened" while
27 building the French PDF.
28 - Document more DVD stuff
36 - Document all the little details of setting up certificates for
37 the Bacula data encryption code.
38 - Document more precisely how to use master keys -- especially
39 for disaster recovery.
42 - Migration from other vendors
46 - Backup conf/exe (all daemons)
47 - Backup up system state
48 - Detect state change of system (verify)
49 - Synthetic Full, Diff, Inc (Virtual, Reconstructed)
51 - Modules for Databases, Exchange, ...
52 - Novell NSS backup http://www.novell.com/coolsolutions/tools/18952.html
53 - Compliance norms that compare restored code hash code.
54 - When glibc crash, get address with
56 - How to sync remote offices.
58 http://www.microsoft.com/technet/itshowcase/content/exchbkup.mspx
61 - Remove all install temp files in Win32 PLUGINSDIR.
62 - Audit retention periods to make sure everything is 64 bit.
63 - Use E'xxx' to escape PostgreSQL strings.
64 - No where in restore causes kaboom.
65 - Performance: multiple spool files for a single job.
66 - Performance: despool attributes when despooling data (problem
67 multiplexing Dir connection).
68 - Make restore use the in-use volume reservation algorithm.
69 - Look at mincore: http://insights.oetiker.ch/linux/fadvise.html
70 - Unicode input http://en.wikipedia.org/wiki/Byte_Order_Mark
71 - Add TLS to bat (should be done).
72 - When Pool specifies Storage command override does not work.
73 - Implement wait_for_sysop() message display in wait_for_device(), which
74 now prints warnings too often.
75 - Ensure that each device in an Autochanger has a different
77 - Add Catalog = to Pool resource so that pools will exist
78 in only one catalog -- currently Pools are "global".
79 - Look at sg_logs -a /dev/sg0 for getting soft errors.
80 - btape "test" command with Offline on Unmount = yes
82 This test is essential to Bacula.
84 I'm going to write one record in file 0,
85 two records in file 1,
86 and three records in file 2
88 02-Feb 11:00 btape: ABORTING due to ERROR in dev.c:715
89 dev.c:714 Bad call to rewind. Device "LTO" (/dev/nst0) not open
90 02-Feb 11:00 btape: Fatal Error because: Bacula interrupted by signal 11: Segmentation violation
91 Kaboom! btape, btape got signal 11. Attempting traceback.
93 - Encryption -- email from Landon
94 > The backup encryption algorithm is currently not configurable, and is
95 > set to AES_128_CBC in src/filed/backup.c. The encryption code
96 > supports a number of different ciphers (as well as adding arbitrary
97 > new ones) -- only a small bit of code would be required to map a
98 > configuration string value to a CRYPTO_CIPHER_* value, if anyone is
99 > interested in implementing this functionality.
101 - Why doesn't @"xxx abc" work in a conf file?
102 - Figure out some way to "automatically" backup conf changes.
103 - Add the OS version back to the Win32 client info.
104 - Restarted jobs have a NULL in the from field.
105 - Modify SD status command to indicate when the SD is writing
106 to a DVD (the device is not open -- see bug #732).
107 - Look at the possibility of adding "SET NAMES UTF8" for MySQL,
108 and possibly changing the blobs into varchar.
109 - Ensure that the SD re-reads the Media record if the JobFiles
110 does not match -- it may have been updated by another job.
111 - Look at moving the Storage directive from the Job to the
112 Pool in the default conf files.
114 - Test Volume compatibility between machine architectures
115 - Encryption documentation
116 - Wrong jobbytes with query 12 (todo)
117 - Bare-metal recovery Windows (todo)
121 - Average tape size from Eric
122 SELECT COALESCE(media_avg_size.volavg,0) * count(Media.MediaId) AS volmax, GROUP BY Media.MediaType, Media.PoolId, media_avg_size.volavg
123 count(Media.MediaId) AS volnum,
124 sum(Media.VolBytes) AS voltotal,
125 Media.PoolId AS PoolId,
126 Media.MediaType AS MediaType
128 LEFT JOIN (SELECT avg(Media.VolBytes) AS volavg,
129 Media.MediaType AS MediaType
131 WHERE Media.VolStatus = 'Full'
132 GROUP BY Media.MediaType
133 ) AS media_avg_size ON (Media.MediaType = media_avg_size.MediaType)
134 GROUP BY Media.MediaType, Media.PoolId, media_avg_size.volavg
138 - Add doc for bweb -- especially Installation
140 http://www.orangecrate.com/modules.php?name=News&file=article&sid=501
142 - Despool attributes in separate thread
145 - Check why restore repeatedly sends Rechdrs between
146 each data chunk -- according to James Harper 9Jan07.
149 - Full at least once a month, ...
150 - Cancel Inc if Diff/Full running
151 - More intelligent re-run
152 - New/deleted file backup
154 - Incremental backup -- rsync, Stow
158 - Look at mondo/mindi
159 - Don't restore Solaris Door files:
160 #define S_IFDOOR in st_mode.
161 see: http://docs.sun.com/app/docs/doc/816-5173/6mbb8ae23?a=view#indexterm-360
162 - Make Bacula by default not backup tmpfs, procfs, sysfs, ...
163 - Fix hardlinked immutable files when linking a second file, the
164 immutable flag must be removed prior to trying to link it.
165 - Implement Python event for backing up/restoring a file.
166 - Change dbcheck to tell users to use native tools for fixing
167 broken databases, and to ensure they have the proper indexes.
168 - add udev rules for Bacula devices.
169 - If a job terminates, the DIR connection can close before the
170 Volume info is updated, leaving the File count wrong.
171 - Look at why SIGPIPE during connection can cause seg fault in
172 writing the daemon message, when Dir dropped to bacula:bacula
173 - Look at zlib 32 => 64 problems.
174 - Possibly turn on St. Bernard code.
175 - Fix bextract to restore ACLs, or better yet, use common routines.
176 - Do we migrate appendable Volumes?
177 - Remove queue.c code.
178 - Print warning message if LANG environment variable does not specify
180 - New dot commands from Arno.
181 .show device=xxx lists information from one storage device, including
182 devices (I'm not even sure that information exists in the DIR...)
183 .move eject device=xxx mostly the same as 'unmount xxx' but perhaps with
184 better machine-readable output like "Ok" or "Error busy"
185 .move eject device=xxx toslot=yyy the same as above, but with a new
186 target slot. The catalog should be updated accordingly.
187 .move transfer device=xxx fromslot=yyy toslot=zzz
190 - Article: http://www.heise.de/open/news/meldung/83231
191 - Article: http://www.golem.de/0701/49756.html
192 - Article: http://lwn.net/Articles/209809/
193 - Article: http://www.onlamp.com/pub/a/onlamp/2004/01/09/bacula.html
194 - Article: http://www.linuxdevcenter.com/pub/a/linux/2005/04/07/bacula.html
195 - Article: http://www.osreviews.net/reviews/admin/bacula
196 - Article: http://www.debianhelp.co.uk/baculaweb.htm
198 - Wikis mentioning Bacula
199 http://wiki.finkproject.org/index.php/Admin:Backups
200 http://wiki.linuxquestions.org/wiki/Bacula
201 http://www.openpkg.org/product/packages/?package=bacula
202 http://www.iterating.com/products/Bacula
203 http://net-snmp.sourceforge.net/wiki/index.php/Net-snmp_extensions
204 http://www.section6.net/wiki/index.php/Using_Bacula_for_Tape_Backups
205 http://bacula.darwinports.com/
206 http://wiki.mandriva.com/en/Releases/Corporate/Server_4/Notes#Bacula
207 http://en.wikipedia.org/wiki/Bacula
210 http://www.devco.net/pubwiki/Bacula/
211 http://paramount.ind.wpi.edu/wiki/doku.php
212 http://gentoo-wiki.com/HOWTO_Backup
213 http://www.georglutz.de/wiki/Bacula
214 http://www.clarkconnect.com/wiki/index.php?title=Modules_-_LAN_Backup/Recovery
215 http://linuxwiki.de/Bacula (in German)
217 - Possibly allow SD to spool even if a tape is not mounted.
218 - It appears to me that you have run into some sort of race
219 condition where two threads want to use the same Volume and they
220 were both given access. Normally that is no problem. However,
221 one thread wanted the particular Volume in drive 0, but it was
222 loaded into drive 1 so it decided to unload it from drive 1 and
223 then loaded it into drive 0, while the second thread went on
224 thinking that the Volume could be used in drive 1 not realizing
225 that in between time, it was loaded in drive 0.
226 I'll look at the code to see if there is some way we can avoid
227 this kind of problem. Probably the best solution is to make the
228 first thread simply start using the Volume in drive 1 rather than
229 transferring it to drive 0.
230 - Fix re-read of last block to check if job has actually written
231 a block, and check if block was written by a different job
232 (i.e. multiple simultaneous jobs writing).
233 - Figure out how to configure query.sql. Suggestion to use m4:
234 == changequote.m4 ===
235 changequote(`[',`]')dnl
236 ==== query.sql.in ===
237 :List next 20 volumes to expire
239 Pool.Name AS PoolName,
244 [ FROM_UNIXTIME(UNIX_TIMESTAMP(Media.LastWritten) Media.VolRetention) AS Expire, ])dnl
246 [ media.lastwritten + interval '1 second' * media.volretention as expire, ])dnl
250 ON Media.PoolId=Pool.PoolId
251 WHERE Media.LastWritten>0
255 Command: m4 -DmySQL changequote.m4 query.sql.in >query.sql
257 The problem is that it requires m4, which is not present on all machines
259 - Given all the problems with FIFOs, I think the solution is to do something a
260 little different, though I will look at the code and see if there is not some
261 simple solution (i.e. some bug that was introduced). What might be a better
262 solution would be to use a FIFO as a sort of "key" to tell Bacula to read and
263 write data to a program rather than the FIFO. For example, suppose you
268 Then, I could imagine if you backup and restore this file with a direct
269 reference as is currently done for fifos, instead, during backup Bacula will
272 /home/kern/my-fifo.backup
274 and read the data that my-fifo.backup writes to stdout. For restore, Bacula
277 /home/kern/my-fifo.restore
279 and send the data backed up to stdout. These programs can either be an
280 executable or a shell script and they need only read/write to stdin/stdout.
282 I think this would give a lot of flexibility to the user without making any
283 significant changes to Bacula.
288 select FilenameId from Filename where Name='';
289 # Get list of all directories referenced in a Backup.
290 select Path.Path from Path,File where File.JobId=nnn and
291 File.FilenameId=(FilenameId-from-above) and File.PathId=Path.PathId
292 order by Path.Path ASC;
294 - Look into using Dart for testing
295 http://public.kitware.com/Dart/HTML/Index.shtml
297 - Look into replacing autotools with cmake
298 http://www.cmake.org/HTML/Index.html
300 === Migration from David ===
301 What I'd like to see:
304 Name = "<poolname>-migrate"
308 Migration Selection Type = LowestUtil | OldestVol | PoolOccupancy |
309 Client | PoolResidence | Volume | JobName | SQLquery
310 Migration Selection Pattern = "regexp"
311 Next Pool = <override>
314 There should be no need for a Level (migration is always Full, since you
315 don't calculate differential/incremental differences for migration),
316 Storage should be determined by the volume types in the pool, and Client
317 is really a selection issue. Migration should always occur to the
318 NextPool defined in the pool definition. If no nextpool is defined, the
319 job should end with a reason of "no place to go". If Next Pool statement
320 is present, we override the check in the pool definition and use the
323 Here's how I'd define Migration Selection Types:
326 Client -- Migrate data from selected client only. Migration Selection
327 Pattern regexp provides pattern to select client names, eg ^FS00* makes
328 all client names starting with FS00 eligible for migration.
330 Jobname -- Migration all jobs matching name. Migration Selection Pattern
331 regexp provides pattern to select jobnames existing in pool.
333 Volume -- Migrate all data on specified volumes. Migration Selection
334 Pattern regexp provides selection criteria for volumes to be migrated.
335 Volumes must exist in pool to be eligible for migration.
339 LowestUtil -- Identify the volume in the pool with the least data on it
340 and empty it. No Migration Selection Pattern required.
342 OldestVol -- Identify the LRU volume with data written, and empty it. No
343 Migration Selection Pattern required.
345 PoolOccupancy -- if pool occupancy exceeds <highmig>, migrate volumes
346 (starting with most full volumes) until pool occupancy drops below
347 <lowmig>. Pool highmig and lowmig values are in pool definition, no
348 Migration Selection Pattern required.
352 SQLQuery -- Migrate all jobuids returned by the supplied SQL query.
353 Migration Selection Pattern contains SQL query to execute; should return
354 a list of 1 or more jobuids to migrate.
356 PoolResidence -- Migrate data sitting in pool for longer than
357 PoolResidence value in pool definition. Migration Selection Pattern
358 optional; if specified, override value in pool definition (value in
362 [ possibly a Python event -- kes ]
364 - Mount on an Autochanger with no tape in the drive causes:
365 Automatically selected Storage: LTO-changer
366 Enter autochanger drive[0]: 0
367 3301 Issuing autochanger "loaded drive 0" command.
368 3302 Autochanger "loaded drive 0", result: nothing loaded.
369 3301 Issuing autochanger "loaded drive 0" command.
370 3302 Autochanger "loaded drive 0", result: nothing loaded.
371 3902 Cannot mount Volume on Storage Device "LTO-Drive1" (/dev/nst0) because:
372 Couldn't rewind device "LTO-Drive1" (/dev/nst0): ERR=dev.c:678 Rewind error on "LTO-Drive1" (/dev/nst0). ERR=No medium found.
373 3905 Device "LTO-Drive1" (/dev/nst0) open but no Bacula volume is mounted.
374 If this is not a blank tape, try unmounting and remounting the Volume.
375 - If Drive 0 is blocked, and drive 1 is set "Autoselect=no", drive 1 will
377 - Autochanger did not change volumes.
378 select * from Storage;
379 +-----------+-------------+-------------+
380 | StorageId | Name | AutoChanger |
381 +-----------+-------------+-------------+
382 | 1 | LTO-changer | 0 |
383 +-----------+-------------+-------------+
384 05-May 03:50 roxie-sd: 3302 Autochanger "loaded drive 0", result is Slot 11.
385 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Warning: Director wanted Volume "LT
386 Current Volume "LT0-002" not acceptable because:
387 1997 Volume "LT0-002" not in catalog.
388 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Error: Autochanger Volume "LT0-002"
389 Setting InChanger to zero in catalog.
390 05-May 03:50 roxie-dir: Tibs.2006-05-05_03.05.02 Error: Unable to get Media record
392 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: Error getting Volume i
393 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: Job 530 canceled.
394 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: spool.c:249 Fatal appe
395 05-May 03:49 Tibs: Tibs.2006-05-05_03.05.02 Fatal error: c:\cygwin\home\kern\bacula
404 FirstWritten: 2006-05-05 03:11:54
405 LastWritten: 2006-05-05 03:50:23
406 LabelDate: 2005-12-26 16:52:40
417 VolRetention: 31,536,000
429 Note VolStatus is blank!!!!!
436 FirstWritten: 0000-00-00 00:00:00
437 LastWritten: 0000-00-00 00:00:00
438 LabelDate: 2005-12-26 16:52:40
449 VolRetention: 31,536,000
462 Automatically selected Storage: LTO-changer
463 Enter autochanger drive[0]: 0
464 3301 Issuing autochanger "loaded drive 0" command.
465 3302 Autochanger "loaded drive 0", result: nothing loaded.
466 3301 Issuing autochanger "loaded drive 0" command.
467 3302 Autochanger "loaded drive 0", result: nothing loaded.
468 3902 Cannot mount Volume on Storage Device "LTO-Drive1" (/dev/nst0) because:
469 Couldn't rewind device "LTO-Drive1" (/dev/nst0): ERR=dev.c:678 Rewind error on "LTO-Drive1" (/dev/nst0). ERR=No medium found.
471 3905 Device "LTO-Drive1" (/dev/nst0) open but no Bacula volume is mounted.
472 If this is not a blank tape, try unmounting and remounting the Volume.
474 - http://www.dwheeler.com/essays/commercial-floss.html
475 - Add VolumeLock to prevent all but lock holder (SD) from updating
476 the Volume data (with the exception of VolumeState).
477 - The btape fill command does not seem to use the Autochanger
478 - Make Windows installer default to system disk drive.
479 - Look at using ioctl(FIOBMAP, ...) on Linux, and
480 DeviceIoControl(..., FSCTL_QUERY_ALLOCATED_RANGES, ...) on
481 Win32 for sparse files.
482 http://www.flexhex.com/docs/articles/sparse-files.phtml
483 http://www.informatik.uni-frankfurt.de/~loizides/reiserfs/fibmap.html
484 - Directive: at <event> "command"
485 - Command: pycmd "command" generates "command" event. How to
486 attach to a specific job?
487 - Integrate Christopher's St. Bernard code.
488 - run_cmd() returns int should return JobId_t
489 - get_next_jobid_from_list() returns int should return JobId_t
490 - Document export LDFLAGS=-L/usr/lib64
491 - Don't attempt to restore from "Disabled" Volumes.
492 - Network error on Win32 should set Win32 error code.
493 - What happens when you rename a Disk Volume?
494 - Job retention period in a Pool (and hence Volume). The job would
496 - Look at -D_FORTIFY_SOURCE=2
497 - Add Win32 FileSet definition somewhere
498 - Look at fixing restore status stats in SD.
499 - Look at using ioctl(FIMAP) and FIGETBSZ for sparse files.
500 http://www.informatik.uni-frankfurt.de/~loizides/reiserfs/fibmap.html
501 - Implement a mode that says when a hard read error is
502 encountered, read many times (as it currently does), and if the
503 block cannot be read, skip to the next block, and try again. If
504 that fails, skip to the next file and try again, ...
506 create table LevelType (LevelType binary(1), LevelTypeLong tinyblob);
507 insert into LevelType (LevelType,LevelTypeLong) values
511 - Show files/second in client status output.
512 - Add a recursive mark command (rmark) to restore.
513 - "Minimum Job Interval = nnn" sets minimum interval between Jobs
514 of the same level and does not permit multiple simultaneous
515 running of that Job (i.e. lets any previous invocation finish
516 before doing Interval testing).
517 - Look at simplifying File exclusions.
518 - New directive "Delete purged Volumes"
519 - new pool XXX with ScratchPoolId = MyScratchPool's PoolId and
520 let it fill itself, and RecyclePoolId = XXX's PoolId so I can
521 see if it become stable and I just have to supervise
523 - If I want to remove this pool, I set RecyclePoolId = MyScratchPool's
524 PoolId, and when it is empty remove it.
525 - Figure out how to recycle Scratch volumes back to the Scratch Pool.
527 - Allow Check Labels to be used with Bacula labels.
528 - "Resuming" a failed backup (lost line for example) by using the
529 failed backup as a sort of "base" job.
531 - Email to the user when the tape is about to need changing x
532 days before it needs changing.
533 - Command to show next tape that will be used for a job even
534 if the job is not scheduled.
535 - From: Arunav Mandal <amandal@trolltech.com>
536 1. When jobs are running and bacula for some reason crashes or if I do a
537 restart it remembers and jobs it was running before it crashed or restarted
538 as of now I loose all jobs if I restart it.
540 2. When spooling and in the midway if client is disconnected for instance a
541 laptop bacula completely discard the spool. It will be nice if it can write
542 that spool to tape so there will be some backups for that client if not all.
544 3. We have around 150 clients machines it will be nice to have a option to
545 upgrade all the client machines bacula version automatically.
547 4. Atleast one connection should be reserved for the bconsole so at heavy load
548 I should connect to the director via bconsole which at sometimes I can't
550 5. Another most important feature that is missing, say at 10am I manually
551 started backup of client abc and it was a full backup since client abc has
552 no backup history and at 10.30am bacula again automatically started backup of
553 client abc as that was in the schedule. So now we have 2 multiple Full
554 backups of the same client and if we again try to start a full backup of
555 client backup abc bacula won't complain. That should be fixed.
557 - Fix bpipe.c so that it does not modify results pointer.
558 ***FIXME*** calling sequence should be changed.
559 - For Windows disaster recovery see http://unattended.sf.net/
560 - regardless of the retention period, Bacula will not prune the
561 last Full, Diff, or Inc File data until a month after the
562 retention period for the last Full backup that was done.
563 - update volume=xxx --- add status=Full
564 - Remove old spool files on startup.
565 - Exclude SD spool/working directory.
566 - Refuse to prune last valid Full backup. Same goes for Catalog.
568 - Make a callback when Rerun failed levels is called.
569 - Give Python program access to Scheduled jobs.
570 - Add setting Volume State via Python.
571 - Python script to save with Python, not save, save with Bacula.
572 - Python script to do backup.
574 - Change the Priority, Client, Storage, JobStatus (error)
575 at the start of a job.
576 - Why is SpoolDirectory = /home/bacula/spool; not reported
577 as an error when writing a DVD?
578 - Make bootstrap file handle multiple MediaTypes (SD)
579 - Remove all old Device resource code in Dir and code to pass it
580 back in SD -- better, rework it to pass back device statistics.
581 - Check locking of resources -- be sure to lock devices where previously
582 resources were locked.
583 - The last part is left in the spool dir.
586 - In restore don't compare byte count on a raw device -- directory
587 entry does not contain bytes.
591 MA = (last_MA * 3 + rate) / 4
592 rate = (bytes - last_bytes) / (runtime - last_runtime)
593 - Max Vols limit in Pool off by one?
594 - Implement Files/Bytes,... stats for restore job.
595 - Implement Total Bytes Written, ... for restore job.
596 - Despool attributes simultaneously with data in a separate
597 thread, rejoined at end of data spooling.
598 - 7. Implement new Console commands to allow offlining/reserving drives,
599 and possibly manipulating the autochanger (much asked for).
600 - Add start/end date editing in messages (%t %T, %e?) ...
601 - Add ClientDefs similar to JobDefs.
602 - Print more info when bextract -p accepts a bad block.
603 - Fix FD JobType to be set before RunBeforeJob in FD.
604 - Look at adding full Volume and Pool information to a Volume
605 label so that bscan can get *all* the info.
606 - If the user puts "Purge Oldest Volume = yes" or "Recycle Oldest Volume = yes"
607 and there is only one volume in the pool, refuse to do it -- otherwise
608 he fills the Volume, then immediately starts reusing it.
609 - Implement copies and stripes.
610 - Add history file to console.
611 - Each file on tape creates a JobMedia record. Peter has 4 million
612 files spread over 10000 tape files and four tapes. A restore takes
613 16 hours to build the restore list.
614 - Add and option to check if the file size changed during backup.
615 - Make sure SD deletes spool files on error exit.
616 - Delete old spool files when SD starts.
617 - When labeling tapes, if you enter 000026, Bacula uses
618 the tape index rather than the Volume name 000026.
619 - Add offline tape command to Bacula console.
621 Enter MediaId or Volume name: 32
622 Enter new Volume name: DLT-20Dec04
623 Automatically selected Pool: Default
624 Connecting to Storage daemon DLTDrive at 192.168.68.104:9103 ...
625 Sending relabel command from "DLT-28Jun03" to "DLT-20Dec04" ...
626 block.c:552 Write error at 0:0 on device /dev/nst0. ERR=Bad file descriptor.
627 Error writing final EOF to tape. This tape may not be readable.
628 dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
629 askdir.c:219 NULL Volume name. This shouldn't happen!!!
630 3912 Failed to label Volume: ERR=dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
631 Label command failed for Volume DLT-20Dec04.
632 Do not forget to mount the drive!!!
633 - Bug: if a job is manually scheduled to run later, it does not appear
634 in any status report and cannot be cancelled.
636 ==== Keeping track of deleted/new files ====
637 - To mark files as deleted, run essentially a Verify to disk, and
638 when a file is found missing (MarkId != JobId), then create
639 a new File record with FileIndex == -1. This could be done
640 by the FD at the same time as the backup.
642 My "trick" for keeping track of deletions is the following.
643 Assuming the user turns on this option, after all the files
644 have been backed up, but before the job has terminated, the
645 FD will make a pass through all the files and send their
646 names to the DIR (*exactly* the same as what a Verify job
647 currently does). This will probably be done at the same
648 time the files are being sent to the SD avoiding a second
649 pass. The DIR will then compare that to what is stored in
650 the catalog. Any files in the catalog but not in what the
651 FD sent will receive a catalog File entry that indicates
652 that at that point in time the file was deleted. This
653 either transmitted to the FD or simultaneously computed in
654 the FD, so that the FD can put a record on the tape that
655 indicates that the file has been deleted at this point.
656 A delete file entry could potentially be one with a FileIndex
657 of 0 or perhaps -1 (need to check if FileIndex is used for
658 some other thing as many of the Bacula fields are "overloaded"
661 During a restore, any file initially picked up by some
662 backup (Full, ...) then subsequently having a File entry
663 marked "delete" will be removed from the tree, so will not
664 be restored. If a file with the same name is later OK it
665 will be inserted in the tree -- this already happens. All
666 will be consistent except for possible changes during the
669 Since I'm on the subject, some of you may be wondering what
670 the utility of the in memory tree is if you are going to
671 restore everything (at least it comes up from time to time
672 on the list). Well, it is still *very* useful because it
673 allows only the last item found for a particular filename
674 (full path) to be entered into the tree, and thus if a file
675 is backed up 10 times, only the last copy will be restored.
676 I recently (last Friday) restored a complete directory, and
677 the Full and all the Differential and Incremental backups
678 spanned 3 Volumes. The first Volume was not even mounted
679 because all the files had been updated and hence backed up
680 since the Full backup was made. In this case, the tree
681 saved me a *lot* of time.
683 Make sure this information is stored on the tape too so
684 that it can be restored directly from the tape.
686 All the code (with the exception of formally generating and
687 saving the delete file entries) already exists in the Verify
688 Catalog command. It explicitly recognizes added/deleted files since
689 the last InitCatalog. It is more or less a "simple" matter of
690 taking that code and adapting it slightly to work for backups.
692 Comments from Martin Simmons (I think they are all covered):
693 Ok, that should cover the basics. There are few issues though:
695 - Restore will depend on the catalog. I think it is better to include the
696 extra data in the backup as well, so it can be seen by bscan and bextract.
698 - I'm not sure if it will preserve multiple hard links to the same inode. Or
699 maybe adding or removing links will cause the data to be dumped again?
701 - I'm not sure if it will handle renamed directories. Possibly it will work
702 by dumping the whole tree under a renamed directory?
704 - It remains to be seen how the backup performance of the DIR's will be
705 affected when comparing the catalog for a large filesystem.
709 How about introducing a Type = MgmtPolicy job type? That job type would
710 be responsible for scanning the Bacula environment looking for specific
711 conditions, and submitting the appropriate jobs for implementing said
715 Name = "Migration-Policy"
717 Policy Selection Job Type = Migrate
718 Scope = "<keyword> <operator> <regexp>"
719 Threshold = "<keyword> <operator> <regexp>"
720 Job Template = <template-name>
723 Where <keyword> is any legal job keyword, <operator> is a comparison
724 operator (=,<,>,!=, logical operators AND/OR/NOT) and <regexp> is a
725 appropriate regexp. I could see an argument for Scope and Threshold
726 being SQL queries if we want to support full flexibility. The
727 Migration-Policy job would then get scheduled as frequently as a site
728 felt necessary (suggested default: every 15 minutes).
733 Name = "Migration-Policy"
735 Policy Selection Job Type = Migration
737 Threshold = "Migration Selection Type = LowestUtil"
738 Job Template = "MigrationTemplate"
741 would select all pools for examination and generate a job based on
742 MigrationTemplate to automatically select the volume with the lowest
743 usage and migrate it's contents to the nextpool defined for that pool.
745 This policy abstraction would be really handy for adjusting the behavior
746 of Bacula according to site-selectable criteria (one thing that pops
747 into mind is Amanda's ability to automatically adjust backup levels
748 depending on various criteria).
754 - Add Pool/Storage override regression test.
755 - Add delete JobId to regression.
756 - Add a regression test for dbcheck.
757 - New test to add bscan to four-concurrent-jobs regression,
758 i.e. after the four-concurrent jobs zap the
759 database as is done in the bscan-test, then use bscan to
760 restore the database, do a restore and compare with the
762 - Add restore of specific JobId to regression (item 3
763 on the restore prompt)
764 - Add IPv6 to regression
765 - Add database test to regression. Test each function like delete,
768 - AntiVir can slow down backups on Win32 systems.
769 - Win32 systems with FAT32 can be much slower than NTFS for
770 more than 1000 files per directory.
774 - A HOLD command to stop all jobs from starting.
775 - A PAUSE command to pause all running jobs ==> release the
777 - Media Type = LTO,LTO-2,LTO-3
778 Media Type Read = LTO,LTO2,LTO3
779 Media Type Write = LTO2, LTO3
781 === From Carsten Menke <bootsy52@gmx.net>
783 Following is a list of what I think in the situations where I'm faced with,
784 could be a usefull enhancement to bacula, which I'm certain other users will
785 benefit from as well.
787 1. NextJob/NextJobs Directive within a Job Resource in the form of
788 NextJobs = job1,job2.
791 I currently solved the problem with running multiple jobs each after each
792 by setting the Max Wait Time for a job to 8 hours, and give
793 the jobs different Priorities. However, there scenarios where
794 1 Job is directly depending on another job, so if the former job fails,
795 the job after it needn't to be run
796 while maybe other jobs should run despite of that
799 A Backup Job and a Verify job, if the backup job fails there is no need to run
800 the verify job, as the backup job already failed. However, one may like
801 to backup the Catalog to disk despite of that the main backup job failed.
804 I see that this is related to the Event Handlers which are on the ToDo
805 list, also it is maybe a good idea to check for the return value and
806 execute different actions based on the return value
809 3. offline capability to bconsole
812 Currently I use a script which I execute within the last Job via the
813 RunAfterJob Directive, to release and eject the tape.
814 So I have to call bconsole "release=Storage-Name" and afterwards
815 mt -f /dev/nst0 eject to get the tape out.
817 If I have multiple Storage Devices, than these may not be /dev/nst0 and
818 I have to modify the script or call it with parameters etc.
819 This would actually not be needed, as everything is already defined
820 in bacula-sd.conf and if I can invoke bconsole with the
821 storage name via $1 in the script than I'm done and information is
824 4. %s for Storage Name added to the chars being substituted in "RunAfterJob"
828 For the reason mentioned in 3. to have the ability to call a
829 script with /scripts/foobar %s and in the script use $1
830 to pass the Storage Name to bconsole
832 5. Setting Volume State within a Job Resource
835 Instead of using "Maximum Volume Jobs" in the Pool Resource,
836 I would have the possibilty to define
837 in a Job Resource that after this certain job is run, the Volume State
838 should be set to "Volume State = Used", this give more flexibility (IMHO).
840 6. Localization of Bacula Messages
843 Unfortunatley many,many people I work with don't speak english very well.
844 So if at least the Reporting messages would be localized then they
845 would understand that they have to change the tape,etc. etc.
847 I volunteer to do the german translations, and if I can convince my wife also
848 french and Morre (western african language).
850 7. OK, this is evil, probably bound to security risks and maybe not possible
851 due to the design of bacula.
853 Implementation of Backtics ( `command` ) for shell comand execution to
854 the "Label Format" Directive.
858 Currently I have defined BACULA_DAY_OF_WEEK="day1|day2..." resulting in
859 Label Format = "HolyBackup-${BACULA_DAY_OF_WEEK[${WeekDay}]}". If I could
860 use backticks than I could use "Label Format = HolyBackup-`date +%A` to have
861 the localized name for the day of the week appended to the
862 format string. Then I have the tape labeled automatically with weekday
863 name in the correct language.
865 - Yes, that is surely the case. I probably should turn those into Warning
866 errors. In addition, you just made me think that it might not be bad to
867 add an option to check the file size after backing up the file and
868 report if it changes. This would be done as an option because it would
871 Kern, good idea. If you do do that, mention in the output: file
872 shrunk, or file expanded, just to make it obvious to the user
873 (without having to the refer to file size), just how the file size
876 Would this option be for all file, or just one file? Or a fileset?
877 - Make output from status use html table tags for nicely
878 presenting in a browser.
879 - Can one write tapes faster with 8192 byte block sizes?
880 - Document security problems with the same password for everyone in
881 rpm and Win32 releases.
882 - Browse generations of files.
883 - I've seen an error when my catalog's File table fills up. I
884 then have to recreate the File table with a larger maximum row
885 size. Relevant information is at
886 http://dev.mysql.com/doc/mysql/en/Full_table.html ; I think the
887 "Installing and Configuring MySQL" chapter should talk a bit
888 about this potential problem, and recommend a solution.
889 - For Solaris must use POSIX awk.
890 - Want speed of writing to tape while despooling.
891 - Supported autochanger:
899 Wangtek 6525ES (SCSI-1 QIC drive, 525MB), under Linux 2.4.something,
900 bacula 1.36.0/1 works with blocksize 16k INSIDE bacula-sd.conf.
901 - Add regex from http://www.pcre.org to Bacula for Win32.
902 - Use only shell tools no make in CDROM package.
903 - Include within include does it work?
904 - Implement a Pool of type Cleaning?
905 - Implement VolReadTime and VolWriteTime in SD
906 - Modify Backing up Your Database to include a bootstrap file.
907 - Think about making certain database errors fatal.
908 - Look at correcting the time jump in the scheduler for daylight
909 savings time changes.
910 - Add a "real" timer to network connections.
911 - Promote to Full = Time period
912 - Check dates entered by user for correctness (month/day/... ranges)
913 - Compress restore Volume listing by date and first file.
914 - Look at patches/bacula_db.b2z postgresql that loops during restore.
916 - Perhaps add read/write programs and/or plugins to FileSets.
917 - How to handle backing up portables ...
918 - Add some sort of guaranteed Interval for upgrading jobs.
919 - Can we write the state file after every job terminates? On Win32
920 the system crashes and the state file is not updated.
923 Documentation to do: (any release a little bit at a time)
924 - Doc to do unmount before removing magazine.
925 - Alternative to static linking "ldd prog" save all binaries listed,
926 restore them and point LD_LIBRARY_PATH to them.
927 - Document add "</dev/null >/dev/null 2>&1" to the bacula-fd command line
928 - Document query file format.
929 - Add more documentation for bsr files.
930 - Document problems with Verify and pruning.
931 - Document how to use multiple databases.
932 - VXA drives have a "cleaning required"
933 indicator, but Exabyte recommends preventive cleaning after every 75
936 In this context, it should be noted that Exabyte has a command-line
937 vxatool utility available for free download. (The current version is
938 vxatool-3.72.) It can get diagnostic info, read, write and erase tapes,
939 test the drive, unload tapes, change drive settings, flash new firmware,
941 Of particular interest in this context is that vxatool <device> -i will
942 report, among other details, the time since last cleaning in tape motion
943 minutes. This information can be retrieved (and settings changed, for
944 that matter) through the generic-SCSI device even when Bacula has the
945 regular tape device locked. (Needless to say, I don't recommend
946 changing tape settings while a job is running.)
947 - Lookup HP cleaning recommendations.
948 - Lookup HP tape replacement recommendations (see trouble shooting autochanger)
949 - Document doing table repair
952 ===================================
953 - Add macro expansions in JobDefs.
954 Run Before Job = "SomeFile %{Level} %{Client}"
955 Write Bootstrap="/some/dir/%{JobName}_%{Client}.bsr"
956 - Use non-blocking network I/O but if no data is available, use
958 - Use gather write() for network I/O.
959 - Autorestart on crash.
960 - Add bandwidth limiting.
961 - Add acks every once and a while from the SD to keep
962 the line from timing out.
963 - When an error in input occurs and conio beeps, you can back
964 up through the prompt.
965 - Detect fixed tape block mode during positioning by looking at
966 block numbers in btape "test". Possibly adjust in Bacula.
967 - Fix list volumes to output volume retention in some other
968 units, perhaps via a directive.
969 - Allow Simultaneous Priorities = yes => run up to Max concurrent jobs even
970 with multiple priorities.
971 - If you use restore replace=never, the directory attributes for
972 non-existent directories will not be restored properly.
974 - see lzma401.zip in others directory for new compression
976 - Allow the user to select JobType for manual pruning/purging.
977 - bscan does not put first of two volumes back with all info in
979 - Implement the FreeBSD nodump flag in chflags.
980 - Figure out how to make named console messages go only to that
981 console and to the non-restricted console (new console class?).
982 - Make restricted console prompt for password if *ask* is set or
983 perhaps if password is undefined.
984 - Implement "from ISO-date/time every x hours/days/weeks/months" in
987 ==== from Marc Schoechlin
988 - the help-command should be more verbose
989 (it should explain the paramters of the different
991 -> its time-comsuming to consult the manual anytime
992 you need a special parameter
993 -> maybe its more easy to maintain this, if the
994 descriptions of that commands are outsourced to
996 - the cd-command should allow complete paths
997 i.e. cd /foo/bar/foo/bar
998 -> if a customer mails me the path to a certain file,
999 its faster to enter the specified directory
1000 - if the password is not configured in bconsole.conf
1001 you should be asked for it.
1002 -> sometimes you like to do restore on a customer-machine
1003 which shouldnt know the password for bacula.
1004 -> adding the password to the file favours admins
1005 to forget to remove the password after usage
1007 the protection of that file is less important
1008 - long-listed-output of commands should be scrollable
1009 like the unix more/less-command does
1010 -> if someone runs 200 and more machines, the lists could
1011 be a little long and complex
1012 - command-output should be shown column by column
1013 to reduce scrolling and to increase clarity
1015 - lsmark should list the selected files with full
1017 - wildcards for selecting and file and directories would be nice
1018 - any actions should be interuptable with STRG+C
1019 - command-expansion would be pretty cool
1021 - When the replace Never option is set, new directory permissions
1022 are not restored. See bug 213. To fix this requires creating a
1023 list of newly restored directories so that those directory
1024 permissions *can* be restored.
1025 - Add prune all command
1026 - Document fact that purge can destroy a part of a restore by purging
1027 one volume while others remain valid -- perhaps mark Jobs.
1028 - Add multiple-media-types.txt
1029 - look at mxt-changer.html
1030 - Make ? do a help command (no return needed).
1031 - Implement restore directory.
1032 - Document streams and how to implement them.
1033 - Try not to re-backup a file if a new hard link is added.
1034 - Add feature to backup hard links only, but not the data.
1035 - Fix stream handling to be simpler.
1036 - Add Priority and Bootstrap to Run a Job.
1037 - Eliminate Restore "Run Restore Job" prompt by allowing new "run command
1039 - Remove View FileSet button from Run a Job dialog.
1040 - Handle prompt for restore job at end of Restore command.
1041 - Add display of total selected files to Restore window.
1042 - Add tree pane to left of window.
1043 - Add progress meter.
1044 - Max wait time or max run time causes seg fault -- see runtime-bug.txt
1045 - Add message to user to check for fixed block size when the forward
1046 space test fails in btape.
1047 - When unmarking a directory check if all files below are unmarked and
1048 then remove the + flag -- in the restore tree.
1049 - Possibly implement: Action = Unmount Device="TapeDrive1" in Admin jobs.
1050 - Setup lrrd graphs: (http://www.linpro.no/projects/lrrd/) Mike Acar.
1051 - Revisit the question of multiple Volumes (disk) on a single device.
1052 - Add a block copy option to bcopy.
1053 - Finish work on Gnome restore GUI.
1054 - Fix "llist jobid=xx" where no fileset or client exists.
1055 - For each job type (Admin, Restore, ...) require only the really necessary
1056 fields.- Pass Director resource name as an option to the Console.
1057 - Add a "batch" mode to the Console (no unsolicited queries, ...).
1058 - Add a .list all files in the restore tree (probably also a list all files)
1059 Do both a long and short form.
1060 - Allow browsing the catalog to see all versions of a file (with
1061 stat data on each file).
1062 - Restore attributes of directory if replace=never set but directory
1064 - Use SHA1 on authentication if possible.
1065 - See comtest-xxx.zip for Windows code to talk to USB.
1066 - Add John's appended files:
1067 Appended = { /files/server/logs/http/*log }
1068 and such files would be treated as follows.On a FULL backup, they would
1069 be backed up like any other file.On an INCREMENTAL backup, where a
1070 previous INCREMENTAL or FULL was already in thecatalogue and the length
1071 of the file wasgreater than the length of the last backup, only thedata
1072 added since the last backup will be dumped.On an INCREMENTAL backup, if
1073 the length of the file is less than thelength of the file with the same
1074 name last backed up, the completefile is dumped.On Windows systems, with
1075 creation date of files, we can be evensmarter about this and not count
1076 entirely upon the length.On a restore, the full and all incrementals
1077 since it will beapplied in sequence to restore the file.
1078 - Check new HAVE_WIN32 open bits.
1079 - Check if the tape has moved before writing.
1080 - Handling removable disks -- see below:
1081 - Keep track of tape use time, and report when cleaning is necessary.
1082 - Add FromClient and ToClient keywords on restore command (or
1083 BackupClient RestoreClient).
1084 - Implement a JobSet, which groups any number of jobs. If the
1085 JobSet is started, all the jobs are started together.
1086 Allow Pool, Level, and Schedule overrides.
1087 - Enhance cancel to timeout BSOCK packets after a specific delay.
1088 - Do scheduling by UTC using gmtime_r() in run_conf, scheduler, and
1089 ua_status.!!! Thanks to Alan Brown for this tip.
1090 - Look at updating Volume Jobs so that Max Volume Jobs = 1 will work
1091 correctly for multiple simultaneous jobs.
1092 - Correct code so that FileSet MD5 is calculated for < and | filename
1094 - Implement the Media record flag that indicates that the Volume does disk
1096 - Implement VolAddr, which is used when Volume is addressed like a disk,
1097 and form it from VolFile and VolBlock.
1098 - Make multiple restore jobs for multiple media types specifying
1099 the proper storage type.
1100 - Fix fast block rejection (stored/read_record.c:118). It passes a null
1101 pointer (rec) to try_repositioning().
1102 - Look at extracting Win data from BackupRead.
1103 - Implement RestoreJobRetention? Maybe better "JobRetention" in a Job,
1104 which would take precidence over the Catalog "JobRetention".
1105 - Implement Label Format in Add and Label console commands.
1106 - Possibly up network buffers to 65K. Put on variable.
1107 - Put email tape request delays on one or more variables. User wants
1108 to cancel the job after a certain time interval. Maximum Mount Wait?
1109 - Job, Client, Device, Pool, or Volume?
1110 Is it possible to make this a directive which is *optional* in multiple
1111 resources, like Level? If so, I think I'd make it an optional directive
1112 in Job, Client, and Pool, with precedence such that Job overrides Client
1113 which in turn overrides Pool.
1115 - New Storage specifications:
1116 - Want to write to multiple storage devices simultaneously
1117 - Want to write to multiple storage devices sequentially (in one job)
1118 - Want to read/write simultaneously
1119 - Key is MediaType -- it must match
1121 Passed to SD as a sort of BSR record called Storage Specification
1125 MediaType -> Next MediaType
1127 Device -> Next Device
1129 Allow multiple Storage specifications
1137 Allow Multiple Pool specifications (note, Pool currently
1139 Allow Multiple MediaType specifications in Dir conf
1140 Allow Multiple Device specifications in Dir conf
1141 Perhaps keep this in a single SSR
1142 Tie a Volume to a specific device by using a MediaType that
1143 is contained in only one device.
1144 In SD allow Device to have Multiple MediaTypes
1146 - Ideas from Jerry Scharf:
1147 First let's point out some big pluses that bacula has for this
1149 more importantly it's active. Thank you so much for that
1150 even more important, it's not flaky
1151 it has an open access catalog, opening many possibilities
1152 it's pushing toward heterogeneous systems capability
1154 Macintosh file client
1155 macs are an interesting niche, but I fear a server is a rathole
1156 working bare iron recovery for windows
1157 the option for inc/diff backups not reset on fileset revision
1158 a) use both change and inode update time against base time
1159 b) do the full catalog check (expensive but accurate)
1160 sizing guide (how much system is needed to back up N systems/files)
1161 consultants on using bacula in building a disaster recovery system
1162 an integration guide
1163 or how to get at fancy things that one could do with bacula
1164 logwatch code for bacula logs (or similar)
1165 linux distro inclusion of bacula (brings good and bad, but necessary)
1166 win2k/XP server capability (icky but you asked)
1167 support for Oracle database ??
1169 - Look at adding SQL server and Exchange support for Windows.
1170 - Make dev->file and dev->block_num signed integers so that -1 can
1171 be an invalid value which happens with BSR.
1172 - Create VolAddr for disk files in place of VolFile and VolBlock. This
1173 is needed to properly specify ranges.
1174 - Add progress of files/bytes to SD and FD.
1175 - Print warning message if FileId > 4 billion
1176 - do a "messages" before the first prompt in Console
1177 - Client does not show busy during Estimate command.
1178 - Implement Console mtx commands.
1179 - Implement a Mount Command and an Unmount Command where
1180 the users could specify a system command to be performed
1181 to do the mount, after which Bacula could attempt to
1182 read the device. This is for Removeable media such as a CDROM.
1183 - Most likely, this mount command would be invoked explicitly
1184 by the user using the current Console "mount" and "unmount"
1185 commands -- the Storage Daemon would do the right thing
1186 depending on the exact nature of the device.
1187 - As with tape drives, when Bacula wanted a new removable
1188 disk mounted, it would unmount the old one, and send a message
1189 to the user, who would then use "mount" as described above
1190 once he had actually inserted the disk.
1191 - Implement dump/print label to UA
1192 - Spool to disk only when the tape is full, then when a tape is hung move
1194 - bextract is sending everything to the log file ****FIXME****
1195 - Allow multiple Storage specifications (or multiple names on
1196 a single Storage specification) in the Job record. Thus a job
1197 can be backed up to a number of storage devices.
1198 - Implement some way for the File daemon to contact the Director
1199 to start a job or pass its DHCP obtained IP number.
1200 - Implement a query tape prompt/replace feature for a console
1201 - Copy console @ code to gnome2-console
1202 - Make tree walk routines like cd, ls, ... more user friendly
1203 by handling spaces better.
1204 - Make sure that Bacula rechecks the tape after the 20 min wait.
1205 - Set IO_NOWAIT on Bacula TCP/IP packets.
1206 - Try doing a raw partition backup and restore by mounting a
1208 - From Lars Kellers:
1209 Yes, it would allow to highly automatic the request for new tapes. If a
1210 tape is empty, bacula reads the barcodes (native or simulated), and if
1211 an unused tape is found, it runs the label command with all the
1212 necessary parameters.
1214 By the way can bacula automatically "move" an empty/purged volume say
1215 in the "short" pool to the "long" pool if this pool runs out of volume
1217 - What to do about "list files job=xxx".
1218 - Look at how fuser works and /proc/PID/fd that is how Nic found the
1219 file descriptor leak in Bacula.
1220 - Implement WrapCounters in Counters.
1221 - Add heartbeat from FD to SD if hb interval expires.
1222 - Can we dynamically change FileSets?
1223 - If pool specified to label command and Label Format is specified,
1224 automatically generate the Volume name.
1225 - Why can't SQL do the filename sort for restore?
1226 - Add ExhautiveRestoreSearch
1227 - Look at the possibility of loading only the necessary
1228 data into the restore tree (i.e. do it one directory at a
1229 time as the user walks through the tree).
1230 - Possibly use the hash code if the user selects all for a restore command.
1231 - Fix "restore all" to bypass building the tree.
1232 - Prohibit backing up archive device (findlib/find_one.c:128)
1233 - Implement Release Device in the Job resource to unmount a drive.
1234 - Implement Acquire Device in the Job resource to mount a drive,
1235 be sure this works with admin jobs so that the user can get
1236 prompted to insert the correct tape. Possibly some way to say to
1237 run the job but don't save the files.
1238 - Make things like list where a file is saved case independent for
1240 - Use autochanger to handle multiple devices.
1241 - Implement a Recycle command
1242 - Start working on Base jobs.
1243 - Implement UnsavedFiles DB record.
1244 - From Phil Stracchino:
1245 It would probably be a per-client option, and would be called
1246 something like, say, "Automatically purge obsoleted jobs". What it
1247 would do is, when you successfully complete a Differential backup of a
1248 client, it would automatically purge all Incremental backups for that
1249 client that are rendered redundant by that Differential. Likewise,
1250 when a Full backup on a client completed, it would automatically purge
1251 all Differential and Incremental jobs obsoleted by that Full backup.
1252 This would let people minimize the number of tapes they're keeping on
1253 hand without having to master the art of retention times.
1254 - When doing a Backup send all attributes back to the Director, who
1255 would then figure out what files have been deleted.
1256 - Currently in mount.c:236 the SD simply creates a Volume. It should have
1257 explicit permission to do so. It should also mark the tape in error
1258 if there is an error.
1259 - Cancel waiting for Client connect in SD if FD goes away.
1261 - Implement timeout in response() when it should come quickly.
1262 - Implement a Slot priority (loaded/not loaded).
1263 - Implement "vacation" Incremental only saves.
1264 - Implement create "FileSet"?
1265 - Add prefixlinks to where or not where absolute links to FD.
1266 - Issue message to mount a new tape before the rewind.
1267 - Simplified client job initiation for portables.
1268 - If SD cannot open a drive, make it periodically retry.
1269 - Add more of the config info to the tape label.
1271 - Refine SD waiting output:
1272 Device is being positioned
1273 > Device is being positioned for append
1274 > Device is being positioned to file x
1276 - Figure out some way to estimate output size and to avoid splitting
1277 a backup across two Volumes -- this could be useful for writing CDROMs
1278 where you really prefer not to have it split -- not serious.
1279 - Have SD compute MD5 or SHA1 and compare to what FD computes.
1280 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
1281 actual data on the Volume and compare it.
1282 - Implement Bacula plugins -- design API
1283 - Make bcopy read through bad tape records.
1284 - Program files (i.e. execute a program to read/write files).
1285 Pass read date of last backup, size of file last time.
1286 - Add Signature type to File DB record.
1287 - CD into subdirectory when open()ing files for backup to
1288 speed up things. Test with testfind().
1289 - Priority job to go to top of list.
1290 - Why are save/restore of device different sizes (sparse?) Yup! Fix it.
1291 - Implement some way for the Console to dynamically create a job.
1292 - Solaris -I on tar for include list
1293 - Need a verbose mode in restore, perhaps to bsr.
1294 - bscan without -v is too quiet -- perhaps show jobs.
1295 - Add code to reject whole blocks if not wanted on restore.
1296 - Check if we can increase Bacula FD priorty in Win2000
1297 - Make sure the MaxVolFiles is fully implemented in SD
1298 - Check if both CatalogFiles and UseCatalog are set to SD.
1299 - Possibly add email to Watchdog if drive is unmounted too
1300 long and a job is waiting on the drive.
1301 - After unmount, if restore job started, ask to mount.
1302 - Add UA rc and history files.
1303 - put termcap (used by console) in ./configure and
1304 allow -with-termcap-dir.
1305 - Fix Autoprune for Volumes to respect need for full save.
1306 - Compare tape to Client files (attributes, or attributes and data)
1307 - Make all database Ids 64 bit.
1308 - Allow console commands to detach or run in background.
1309 - Add SD message variables to control operator wait time
1310 - Maximum Operator Wait
1311 - Minimum Message Interval
1312 - Maximum Message Interval
1313 - Send Operator message when cannot read tape label.
1314 - Verify level=Volume (scan only), level=Data (compare of data to file).
1315 Verify level=Catalog, level=InitCatalog
1317 - Add keyword search to show command in Console.
1318 - Events : tape has more than xxx bytes.
1319 - Complete code in Bacula Resources -- this will permit
1320 reading a new config file at any time.
1321 - Handle ctl-c in Console
1322 - Implement script driven addition of File daemon to config files.
1323 - Think about how to make Bacula work better with File (non-tape) archives.
1324 - Write Unix emulator for Windows.
1325 - Put memory utilization in Status output of each daemon
1326 if full status requested or if some level of debug on.
1327 - Make database type selectable by .conf files i.e. at runtime
1328 - Set flag for uname -a. Add to Volume label.
1329 - Restore files modified after date
1330 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
1331 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
1332 - Timout a job or terminate if link goes down, or reopen link and query.
1333 - Concept of precious tapes (cannot be reused).
1334 - Make bcopy copy with a single tape drive.
1335 - Permit changing ownership during restore.
1338 > My suggestion: Add a feature on the systray menu-icon menu to request
1339 > an immediate backup now. This would be useful for laptop users who may
1340 > not be on the network when the regular scheduled backup is run.
1342 > My wife's suggestion: Add a setting to the win32 client to allow it to
1343 > shut down the machine after backup is complete (after, of course,
1344 > displaying a "System will shut down in one minute, click here to cancel"
1345 > warning dialog). This would be useful for sites that want user
1346 > woorkstations to be shut down overnight to save power.
1349 - Autolabel should be specified by DIR instead of SD.
1351 - Add media capacity
1352 - AutoScan (check checksum of tape)
1353 - Format command = "format /dev/nst0"
1357 - Seek resolution (usually corresponds to buffer size)
1358 - EODErrorCode=ENOSPC or code
1359 - Partial Read error code
1360 - Partial write error code
1361 - Nonformatted read error
1362 - Nonformatted write error
1363 - WriteProtected error
1367 - IgnoreCloseErrors=yes
1377 - FD sends unsaved file list to Director at end of job (see
1379 - File daemon should build list of files skipped, and then
1380 at end of save retry and report any errors.
1381 - Write a Storage daemon that uses pipes and
1382 standard Unix programs to write to the tape.
1384 - Need something that monitors the JCR queue and
1385 times out jobs by asking the deamons where they are.
1386 - Enhance Jmsg code to permit buffering and saving to disk.
1387 - device driver = "xxxx" for drives.
1388 - Verify from Volume
1389 - Ensure that /dev/null works
1390 - Need report class for messages. Perhaps
1391 report resource where report=group of messages
1392 - enhance scan_attrib and rename scan_jobtype, and
1393 fill in code for "since" option
1394 - Director needs a time after which the report status is sent
1395 anyway -- or better yet, a retry time for the job.
1396 - Don't reschedule a job if previous incarnation is still running.
1397 - Some way to automatically backup everything is needed????
1398 - Need a structure for pending actions:
1400 - termination status (part of buffered msgs?)
1402 Read, Write, Clean, Delete
1403 - Login to Bacula; Bacula users with different permissions:
1404 owner, group, user, quotas
1405 - Store info on each file system type (probably in the job header on tape.
1406 This could be the output of df; or perhaps some sort of /etc/mtab record.
1408 ========= ideas ===============
1409 From: "Jerry K. Schieffer" <jerry@skylinetechnology.com>
1410 To: <kern@sibbald.com>
1411 Subject: RE: [Bacula-users] future large programming jobs
1412 Date: Thu, 26 Feb 2004 11:34:54 -0600
1414 I noticed the subject thread and thought I would offer the following
1415 merely as sources of ideas, i.e. something to think about, not even as
1416 strong as a request. In my former life (before retiring) I often
1417 dealt with backups and storage management issues/products as a
1418 developer and as a consultant. I am currently migrating my personal
1419 network from amanda to bacula specifically because of the ability to
1420 cross media boundaries during storing backups.
1421 Are you familiar with the commercial product called ADSM (I think IBM
1422 now sells it under the Tivoli label)? It has a couple of interesting
1423 ideas that may apply to the following topics.
1425 1. Migration: Consider that when you need to restore a system, there
1426 may be pressure to hurry. If all the information for a single client
1427 can eventually end up on the same media (and in chronological order),
1428 the restore is facillitated by not having to search past information
1429 from other clients. ADSM has the concept of "client affinity" that
1430 may be associated with it's storage pools. It seems to me that this
1431 concept (as an optional feature) might fit in your architecture for
1434 ADSM also has the concept of defining one or more storage pools as
1435 "copy pools" (almost mirrors, but only in the sense of contents).
1436 These pools provide the ability to have duplicte data stored both
1437 onsite and offsite. The copy process can be scheduled to be handled
1438 by their storage manager during periods when there is no backup
1439 activity. Again, the migration process might be a place to consider
1440 implementing something like this.
1443 > It strikes me that it would be very nice to be able to do things
1445 > have the Job(s) backing up the machines run, and once they have all
1446 > completed, start a migration job to copy the data from disks Volumes
1448 > a tape library and then to offsite storage. Maybe this can already
1450 > done with some careful scheduling and Job prioritzation; the events
1451 > mechanism described below would probably make it very easy.
1453 This is the goal. In the first step (before events), you simply
1455 the Migration to tape later.
1457 2. Base jobs: In ADSM, each copy of each stored file is tracked in
1458 the database. Once a file (unique by path and metadata such as dates,
1459 size, ownership, etc.) is in a copy pool, no more copies are made. In
1460 other words, when you start ADSM, it begins like your concept of a
1461 base job. After that it is in the "incremental" mode. You can
1462 configure the number of "generations" of files to be retained, plus a
1463 retention date after which even old generations are purged. The
1464 database tracks the contents of media and projects the percentage of
1465 each volume that is valid. When the valid content of a volume drops
1466 below a configured percentage, the valid data are migrated to another
1467 volume and the old volume is marked as empty. Note, this requires
1468 ADSM to have an idea of the contents of a client, i.e. marking the
1469 database when an existing file was deleted, but this would solve your
1470 issue of restoring a client without restoring deleted files.
1472 This is pretty far from what bacula now does, but if you are going to
1473 rip things up for Base jobs,.....
1474 Also, the benefits of this are huge for very large shops, especially
1475 with media robots, but are a pain for shops with manual media
1479 > Base jobs sound pretty useful, but I'm not dying for them.
1481 Nobody is dying for them, but when you see what it does, you will die
1484 3. Restoring deleted files: Since I think my comments in (2) above
1485 have low probability of implementation, I'll also suggest that you
1486 could approach the issue of deleted files by a mechanism of having the
1487 fd report to the dir, a list of all files on the client for every
1488 backup job. The dir could note in the database entry for each file
1489 the date that the file was seen. Then if a restore as of date X takes
1490 place, only files that exist from before X until after X would be
1491 restored. Probably the major cost here is the extra date container in
1492 each row of the files table.
1494 Thanks for "listening". I hope some of this helps. If you want to
1495 contact me, please send me an email - I read some but not all of the
1496 mailing list traffic and might miss a reply there.
1498 Please accept my compliments for bacula. It is doing a great job for
1499 me!! I sympathize with you in the need to wrestle with excelence in
1500 execution vs. excelence in feature inclusion.
1505 ==============================
1508 - Design at hierarchial storage for Bacula. Migration and Clone.
1509 - Implement FSM (File System Modules).
1510 - Audit M_ error codes to ensure they are correct and consistent.
1511 - Add variable break characters to lex analyzer.
1512 Either a bit mask or a string of chars so that
1513 the caller can change the break characters.
1514 - Make a single T_BREAK to replace T_COMMA, etc.
1515 - Ensure that File daemon and Storage daemon can
1516 continue a save if the Director goes down (this
1517 is NOT currently the case). Must detect socket error,
1518 buffer messages for later.
1519 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
1520 - Add ability to backup to two Storage devices (two SD sessions) at
1521 the same time -- e.g. onsite, offsite.
1522 - Add the ability to consolidate old backup sets (basically do a restore
1523 to tape and appropriately update the catalog). Compress Volume sets.
1524 Might need to spool via file is only one drive is available.
1525 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
1526 someway to do so with every volume that has less than x% valid
1530 Migration: Move a backup from one Volume to another
1531 Clone: Copy a backup -- two Volumes
1534 ======================================================
1536 It is somewhat like a Full save becomes an incremental since
1537 the Base job (or jobs) plus other non-base files.
1539 - A Base backup is same as Full backup, just different type.
1540 - New BaseFiles table that contains:
1542 BaseJobId - Base JobId referenced for this FileId (needed ???)
1543 JobId - JobId currently running
1544 FileId - File not backed up, exists in Base Job
1545 FileIndex - FileIndex from Base Job.
1546 i.e. for each base file that exists but is not saved because
1547 it has not changed, the File daemon sends the JobId, BaseId,
1548 FileId, FileIndex back to the Director who creates the DB entry.
1549 - To initiate a Base save, the Director sends the FD
1550 the FileId, and full filename for each file in the Base.
1551 - When the FD finds a Base file, he requests the Director to
1552 send him the full File entry (stat packet plus MD5), or
1553 conversely, the FD sends it to the Director and the Director
1554 says yes or no. This can be quite rapid if the FileId is kept
1555 by the FD for each Base Filename.
1556 - It is probably better to have the comparison done by the FD
1557 despite the fact that the File entry must be sent across the
1559 - An alternative would be to send the FD the whole File entry
1560 from the start. The disadvantage is that it requires a lot of
1561 space. The advantage is that it requires less communications
1563 - The Job record must be updated to indicate that one or more
1565 - At end of Job, FD returns:
1566 1. Count of base files/bytes not written to tape (i.e. matches)
1567 2. Count of base file that were saved i.e. had changed.
1568 - No tape record would be written for a Base file that matches, in the
1569 same way that no tape record is written for Incremental jobs where
1570 the file is not saved because it is unchanged.
1571 - On a restore, all the Base file records must explicitly be
1572 found from the BaseFile tape. I.e. for each Full save that is marked
1573 to have one or more Base Jobs, search the BaseFile for all occurrences
1575 - An optimization might be to make the BaseFile have:
1581 This would avoid the need to explicitly fetch each File record for
1582 the Base job. The Base Job record will be fetched to get the
1583 VolSessionId and VolSessionTime.
1584 =========================================================
1587 ==========================================================
1589 For each Incremental job that is run, there may be files that
1590 were found but not saved because they were locked (this applies
1591 only to Windows). Such a system could send back to the Director
1592 a list of Unsaved files.
1594 - New UnSavedFiles table that contains:
1598 - Then in the next Incremental job, the list of Unsaved Files will be
1599 feed to the FD, who will ensure that they are explicitly chosen even
1600 if standard date/time check would not have selected them.
1601 =============================================================
1605 Multiple drive autochanger data: see Alan Brown
1606 mtx -f xxx unloadStorage Element 1 is Already Full(drive 0 was empty)
1607 Unloading Data Transfer Element into Storage Element 1...source Element
1608 Address 480 is Empty
1610 (drive 0 was empty and so was slot 1)
1611 > mtx -f xxx load 15 0
1612 no response, just returns to the command prompt when complete.
1613 > mtx -f xxx status Storage Changer /dev/changer:2 Drives, 60 Slots ( 2 Import/Export )
1614 Data Transfer Element 0:Full (Storage Element 15 Loaded):VolumeTag = HX001
1615 Data Transfer Element 1:Empty
1616 Storage Element 1:Empty
1617 Storage Element 2:Full :VolumeTag=HX002
1618 Storage Element 3:Full :VolumeTag=HX003
1619 Storage Element 4:Full :VolumeTag=HX004
1620 Storage Element 5:Full :VolumeTag=HX005
1621 Storage Element 6:Full :VolumeTag=HX006
1622 Storage Element 7:Full :VolumeTag=HX007
1623 Storage Element 8:Full :VolumeTag=HX008
1624 Storage Element 9:Full :VolumeTag=HX009
1625 Storage Element 10:Full :VolumeTag=HX010
1626 Storage Element 11:Empty
1627 Storage Element 12:Empty
1628 Storage Element 13:Empty
1629 Storage Element 14:Empty
1630 Storage Element 15:Empty
1631 Storage Element 16:Empty....
1632 Storage Element 28:Empty
1633 Storage Element 29:Full :VolumeTag=CLNU01L1
1634 Storage Element 30:Empty....
1635 Storage Element 57:Empty
1636 Storage Element 58:Full :VolumeTag=NEX261L2
1637 Storage Element 59 IMPORT/EXPORT:Empty
1638 Storage Element 60 IMPORT/EXPORT:Empty
1640 Unloading Data Transfer Element into Storage Element 15...done
1642 (just to verify it remembers where it came from, however it can be
1643 overrriden with mtx unload {slotnumber} to go to any storage slot.)
1645 There needs to be a table of drive # to devices somewhere - If there are
1646 multiple changers or drives there may not be a 1:1 correspondance between
1647 changer drive number and system device name - and depending on the way the
1648 drives are hooked up to scsi busses, they may not be linearly numbered
1649 from an offset point either.something like
1651 Autochanger drives = 2
1652 Autochanger drive 0 = /dev/nst1
1653 Autochanger drive 1 = /dev/nst2
1654 IMHO, it would be _safest_ to use explicit mtx unload commands at all
1655 times, not just for multidrive changers. For a 1 drive changer, that's
1661 MTX's manpage (1.2.15):
1662 unload [<slotnum>] [ <drivenum> ]
1663 Unloads media from drive <drivenum> into slot
1664 <slotnum>. If <drivenum> is omitted, defaults to
1665 drive 0 (as do all commands). If <slotnum> is
1666 omitted, defaults to the slot that the drive was
1667 loaded from. Note that there's currently no way
1668 to say 'unload drive 1's media to the slot it
1669 came from', other than to explicitly use that
1670 slot number as the destination.AB
1676 undef# camcontrol devlist
1677 <WANGTEK 51000 SCSI M74H 12B3> at scbus0 target 2 lun 0 (pass0,sa0)
1678 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 0 (pass1,sa1)
1679 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 1 (pass2)
1681 tapeinfo -f /dev/sg0 with a bad tape in drive 1:
1682 [kern@rufus mtx-1.2.17kes]$ ./tapeinfo -f /dev/sg0
1683 Product Type: Tape Drive
1685 Product ID: 'C5713A '
1687 Attached Changer: No
1688 TapeAlert[3]: Hard Error: Uncorrectable read/write error.
1689 TapeAlert[20]: Clean Now: The tape drive neads cleaning NOW.
1696 Medium Type: Not Loaded
1699 DataCompEnabled: yes
1700 DataCompCapable: yes
1701 DataDeCompEnabled: yes
1708 Handling removable disks
1710 From: Karl Cunningham <karlc@keckec.com>
1712 My backups are only to hard disk these days, in removable bays. This is my
1713 idea of how a backup to hard disk would work more smoothly. Some of these
1714 things Bacula does already, but I mention them for completeness. If others
1715 have better ways to do this, I'd like to hear about it.
1717 1. Accommodate several disks, rotated similar to how tapes are. Identified
1718 by partition volume ID or perhaps by the name of a subdirectory.
1719 2. Abort & notify the admin if the wrong disk is in the bay.
1720 3. Write backups to different subdirectories for each machine to be backed
1722 4. Volumes (files) get created as needed in the proper subdirectory, one
1724 5. When a disk is recycled, remove or zero all old backup files. This is
1725 important as the disk being recycled may be close to full. This may be
1726 better done manually since the backup files for many machines may be
1727 scattered in many subdirectories.
1732 - Why the heck doesn't bacula drop root priviledges before connecting to
1734 - Look at using posix_fadvise(2) for backups -- see bug #751.
1735 Possibly add the code at findlib/bfile.c:795
1736 /* TCP socket options */
1737 #define TCP_KEEPIDLE 4 /* Start keeplives after this period */
1738 - Fix bnet_connect() code to set a timer and to use time to
1740 - Implement 4th argument to make_catalog_backup that passes hostname.
1741 - Test FIFO backup/restore -- make regression
1742 - Please mount volume "xxx" on Storage device ... should also list
1743 Pool and MediaType in case user needs to create a new volume.
1744 - On restore add Restore Client, Original Client.
1745 01-Apr 00:42 rufus-dir: Start Backup JobId 55, Job=kernsave.2007-04-01_00.42.48
1746 01-Apr 00:42 rufus-sd: Python SD JobStart: JobId=55 Client=Rufus
1747 01-Apr 00:42 rufus-dir: Created new Volume "Full0001" in catalog.
1748 01-Apr 00:42 rufus-dir: Using Device "File"
1749 01-Apr 00:42 rufus-sd: kernsave.2007-04-01_00.42.48 Warning: Device "File" (/tmp) not configured to autolabel Volumes.
1750 01-Apr 00:42 rufus-sd: kernsave.2007-04-01_00.42.48 Warning: Device "File" (/tmp) not configured to autolabel Volumes.
1751 01-Apr 00:42 rufus-sd: Please mount Volume "Full0001" on Storage Device "File" (/tmp) for Job kernsave.2007-04-01_00.42.48
1752 01-Apr 00:44 rufus-sd: Wrote label to prelabeled Volume "Full0001" on device "File" (/tmp)
1753 - Check if gnome-console works with TLS.
1754 - the director seg faulted when I omitted the pool directive from a
1755 job resource. I was experimenting and thought it redundant that I had
1756 specified Pool, Full Backup Pool. and Differential Backup Pool. but
1757 apparently not. This happened when I removed the pool directive and
1758 started the director.
1759 - Add Where: client:/.... to restore job report.
1760 - Ensure that moving a purged Volume in ua_purge.c to the RecyclePool
1761 does the right thing.
1762 - FD-SD quick disconnect
1763 - Building the in memory restore tree is slow.