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 - Look at mincore: http://insights.oetiker.ch/linux/fadvise.html
62 - Unicode input http://en.wikipedia.org/wiki/Byte_Order_Mark
63 - How does restore JobId=nnn work? (Dirk)
64 - What does: restore select fileset="TestSet" client="workplay-fd" pool="Default"
65 > storage="File2" before="2007-02-05 23:05:04" do? (Dirk)
67 - Implement update jobid=
68 - Mention Eric and Marc's work + Marc's doc.
70 - When Pool specifies Storage command override does not work.
71 - Implement wait_for_sysop() message display in wait_for_device(), which
72 now prints warnings too often.
74 - Ensure that each device in an Autochanger has a different
76 - Add Catalog = to Pool resource so that pools will exist
77 in only one catalog -- currently Pools are "global".
78 - Look at sg_logs -a /dev/sg0 for getting soft errors.
79 - btape "test" command with Offline on Unmount = yes
81 This test is essential to Bacula.
83 I'm going to write one record in file 0,
84 two records in file 1,
85 and three records in file 2
87 02-Feb 11:00 btape: ABORTING due to ERROR in dev.c:715
88 dev.c:714 Bad call to rewind. Device "LTO" (/dev/nst0) not open
89 02-Feb 11:00 btape: Fatal Error because: Bacula interrupted by signal 11: Segmentation violation
90 Kaboom! btape, btape got signal 11. Attempting traceback.
92 - Encryption -- email from Landon
93 > The backup encryption algorithm is currently not configurable, and is
94 > set to AES_128_CBC in src/filed/backup.c. The encryption code
95 > supports a number of different ciphers (as well as adding arbitrary
96 > new ones) -- only a small bit of code would be required to map a
97 > configuration string value to a CRYPTO_CIPHER_* value, if anyone is
98 > interested in implementing this functionality.
100 - Ensure that moving a purged Volume in ua_purge.c to the RecyclePool
101 does the right thing.
102 - Why doesn't @"xxx abc" work in a conf file?
103 - Figure out some way to "automatically" backup conf changes.
104 - Add the OS version back to the Win32 client info.
105 - Restarted jobs have a NULL in the from field.
106 - Modify SD status command to indicate when the SD is writing
107 to a DVD (the device is not open -- see bug #732).
108 - Look at the possibility of adding "SET NAMES UTF8" for MySQL,
109 and possibly changing the blobs into varchar.
110 - Ensure that the SD re-reads the Media record if the JobFiles
111 does not match -- it may have been updated by another job.
112 - Look at moving the Storage directive from the Job to the
113 Pool in the default conf files.
115 - Test Volume compatibility between machine architectures
116 - Encryption documentation
117 - Wrong jobbytes with query 12 (todo)
118 - bacula-1.38.2-ssl.patch
119 - Bare-metal recovery Windows (todo)
123 - Average tape size from Eric
124 SELECT COALESCE(media_avg_size.volavg,0) * count(Media.MediaId) AS volmax, GROUP BY Media.MediaType, Media.PoolId, media_avg_size.volavg
125 count(Media.MediaId) AS volnum,
126 sum(Media.VolBytes) AS voltotal,
127 Media.PoolId AS PoolId,
128 Media.MediaType AS MediaType
130 LEFT JOIN (SELECT avg(Media.VolBytes) AS volavg,
131 Media.MediaType AS MediaType
133 WHERE Media.VolStatus = 'Full'
134 GROUP BY Media.MediaType
135 ) AS media_avg_size ON (Media.MediaType = media_avg_size.MediaType)
136 GROUP BY Media.MediaType, Media.PoolId, media_avg_size.volavg
140 - Add doc for bweb -- especially Installation
142 http://www.orangecrate.com/modules.php?name=News&file=article&sid=501
144 - FD-SD quick disconnect
145 - Despool attributes in separate thread
148 - Check why restore repeatedly sends Rechdrs between
149 each data chunk -- according to James Harper 9Jan07.
150 - Building the in memory restore tree is slow.
153 - Full at least once a month, ...
154 - Cancel Inc if Diff/Full running
155 - More intelligent re-run
156 - New/deleted file backup
158 - Incremental backup -- rsync, Stow
162 - Look at mondo/mindi
163 - Don't restore Solaris Door files:
164 #define S_IFDOOR in st_mode.
165 see: http://docs.sun.com/app/docs/doc/816-5173/6mbb8ae23?a=view#indexterm-360
166 - Make Bacula by default not backup tmpfs, procfs, sysfs, ...
167 - Fix hardlinked immutable files when linking a second file, the
168 immutable flag must be removed prior to trying to link it.
169 - Implement Python event for backing up/restoring a file.
170 - Change dbcheck to tell users to use native tools for fixing
171 broken databases, and to ensure they have the proper indexes.
172 - add udev rules for Bacula devices.
173 - If a job terminates, the DIR connection can close before the
174 Volume info is updated, leaving the File count wrong.
175 - Look at why SIGPIPE during connection can cause seg fault in
176 writing the daemon message, when Dir dropped to bacula:bacula
177 - Look at zlib 32 => 64 problems.
178 - Possibly turn on St. Bernard code.
179 - Fix bextract to restore ACLs, or better yet, use common routines.
180 - Do we migrate appendable Volumes?
181 - Remove queue.c code.
182 - Print warning message if LANG environment variable does not specify
184 - New dot commands from Arno.
185 .show device=xxx lists information from one storage device, including
186 devices (I'm not even sure that information exists in the DIR...)
187 .move eject device=xxx mostly the same as 'unmount xxx' but perhaps with
188 better machine-readable output like "Ok" or "Error busy"
189 .move eject device=xxx toslot=yyy the same as above, but with a new
190 target slot. The catalog should be updated accordingly.
191 .move transfer device=xxx fromslot=yyy toslot=zzz
194 - Article: http://www.heise.de/open/news/meldung/83231
195 - Article: http://www.golem.de/0701/49756.html
196 - Article: http://lwn.net/Articles/209809/
197 - Article: http://www.onlamp.com/pub/a/onlamp/2004/01/09/bacula.html
198 - Article: http://www.linuxdevcenter.com/pub/a/linux/2005/04/07/bacula.html
199 - Article: http://www.osreviews.net/reviews/admin/bacula
200 - Article: http://www.debianhelp.co.uk/baculaweb.htm
202 - Wikis mentioning Bacula
203 http://wiki.finkproject.org/index.php/Admin:Backups
204 http://wiki.linuxquestions.org/wiki/Bacula
205 http://www.openpkg.org/product/packages/?package=bacula
206 http://www.iterating.com/products/Bacula
207 http://net-snmp.sourceforge.net/wiki/index.php/Net-snmp_extensions
208 http://www.section6.net/wiki/index.php/Using_Bacula_for_Tape_Backups
209 http://bacula.darwinports.com/
210 http://wiki.mandriva.com/en/Releases/Corporate/Server_4/Notes#Bacula
211 http://en.wikipedia.org/wiki/Bacula
214 http://www.devco.net/pubwiki/Bacula/
215 http://paramount.ind.wpi.edu/wiki/doku.php
216 http://gentoo-wiki.com/HOWTO_Backup
217 http://www.georglutz.de/wiki/Bacula
218 http://www.clarkconnect.com/wiki/index.php?title=Modules_-_LAN_Backup/Recovery
219 http://linuxwiki.de/Bacula (in German)
221 - Possibly allow SD to spool even if a tape is not mounted.
222 - It appears to me that you have run into some sort of race
223 condition where two threads want to use the same Volume and they
224 were both given access. Normally that is no problem. However,
225 one thread wanted the particular Volume in drive 0, but it was
226 loaded into drive 1 so it decided to unload it from drive 1 and
227 then loaded it into drive 0, while the second thread went on
228 thinking that the Volume could be used in drive 1 not realizing
229 that in between time, it was loaded in drive 0.
230 I'll look at the code to see if there is some way we can avoid
231 this kind of problem. Probably the best solution is to make the
232 first thread simply start using the Volume in drive 1 rather than
233 transferring it to drive 0.
234 - Fix re-read of last block to check if job has actually written
235 a block, and check if block was written by a different job
236 (i.e. multiple simultaneous jobs writing).
237 - Figure out how to configure query.sql. Suggestion to use m4:
238 == changequote.m4 ===
239 changequote(`[',`]')dnl
240 ==== query.sql.in ===
241 :List next 20 volumes to expire
243 Pool.Name AS PoolName,
248 [ FROM_UNIXTIME(UNIX_TIMESTAMP(Media.LastWritten) Media.VolRetention) AS Expire, ])dnl
250 [ media.lastwritten + interval '1 second' * media.volretention as expire, ])dnl
254 ON Media.PoolId=Pool.PoolId
255 WHERE Media.LastWritten>0
259 Command: m4 -DmySQL changequote.m4 query.sql.in >query.sql
261 The problem is that it requires m4, which is not present on all machines
263 - Given all the problems with FIFOs, I think the solution is to do something a
264 little different, though I will look at the code and see if there is not some
265 simple solution (i.e. some bug that was introduced). What might be a better
266 solution would be to use a FIFO as a sort of "key" to tell Bacula to read and
267 write data to a program rather than the FIFO. For example, suppose you
272 Then, I could imagine if you backup and restore this file with a direct
273 reference as is currently done for fifos, instead, during backup Bacula will
276 /home/kern/my-fifo.backup
278 and read the data that my-fifo.backup writes to stdout. For restore, Bacula
281 /home/kern/my-fifo.restore
283 and send the data backed up to stdout. These programs can either be an
284 executable or a shell script and they need only read/write to stdin/stdout.
286 I think this would give a lot of flexibility to the user without making any
287 significant changes to Bacula.
292 select FilenameId from Filename where Name='';
293 # Get list of all directories referenced in a Backup.
294 select Path.Path from Path,File where File.JobId=nnn and
295 File.FilenameId=(FilenameId-from-above) and File.PathId=Path.PathId
296 order by Path.Path ASC;
298 - Look into using Dart for testing
299 http://public.kitware.com/Dart/HTML/Index.shtml
301 - Look into replacing autotools with cmake
302 http://www.cmake.org/HTML/Index.html
304 === Migration from David ===
305 What I'd like to see:
308 Name = "<poolname>-migrate"
312 Migration Selection Type = LowestUtil | OldestVol | PoolOccupancy |
313 Client | PoolResidence | Volume | JobName | SQLquery
314 Migration Selection Pattern = "regexp"
315 Next Pool = <override>
318 There should be no need for a Level (migration is always Full, since you
319 don't calculate differential/incremental differences for migration),
320 Storage should be determined by the volume types in the pool, and Client
321 is really a selection issue. Migration should always occur to the
322 NextPool defined in the pool definition. If no nextpool is defined, the
323 job should end with a reason of "no place to go". If Next Pool statement
324 is present, we override the check in the pool definition and use the
327 Here's how I'd define Migration Selection Types:
330 Client -- Migrate data from selected client only. Migration Selection
331 Pattern regexp provides pattern to select client names, eg ^FS00* makes
332 all client names starting with FS00 eligible for migration.
334 Jobname -- Migration all jobs matching name. Migration Selection Pattern
335 regexp provides pattern to select jobnames existing in pool.
337 Volume -- Migrate all data on specified volumes. Migration Selection
338 Pattern regexp provides selection criteria for volumes to be migrated.
339 Volumes must exist in pool to be eligible for migration.
343 LowestUtil -- Identify the volume in the pool with the least data on it
344 and empty it. No Migration Selection Pattern required.
346 OldestVol -- Identify the LRU volume with data written, and empty it. No
347 Migration Selection Pattern required.
349 PoolOccupancy -- if pool occupancy exceeds <highmig>, migrate volumes
350 (starting with most full volumes) until pool occupancy drops below
351 <lowmig>. Pool highmig and lowmig values are in pool definition, no
352 Migration Selection Pattern required.
356 SQLQuery -- Migrate all jobuids returned by the supplied SQL query.
357 Migration Selection Pattern contains SQL query to execute; should return
358 a list of 1 or more jobuids to migrate.
360 PoolResidence -- Migrate data sitting in pool for longer than
361 PoolResidence value in pool definition. Migration Selection Pattern
362 optional; if specified, override value in pool definition (value in
366 [ possibly a Python event -- kes ]
368 - Mount on an Autochanger with no tape in the drive causes:
369 Automatically selected Storage: LTO-changer
370 Enter autochanger drive[0]: 0
371 3301 Issuing autochanger "loaded drive 0" command.
372 3302 Autochanger "loaded drive 0", result: nothing loaded.
373 3301 Issuing autochanger "loaded drive 0" command.
374 3302 Autochanger "loaded drive 0", result: nothing loaded.
375 3902 Cannot mount Volume on Storage Device "LTO-Drive1" (/dev/nst0) because:
376 Couldn't rewind device "LTO-Drive1" (/dev/nst0): ERR=dev.c:678 Rewind error on "LTO-Drive1" (/dev/nst0). ERR=No medium found.
377 3905 Device "LTO-Drive1" (/dev/nst0) open but no Bacula volume is mounted.
378 If this is not a blank tape, try unmounting and remounting the Volume.
379 - If Drive 0 is blocked, and drive 1 is set "Autoselect=no", drive 1 will
381 - Autochanger did not change volumes.
382 select * from Storage;
383 +-----------+-------------+-------------+
384 | StorageId | Name | AutoChanger |
385 +-----------+-------------+-------------+
386 | 1 | LTO-changer | 0 |
387 +-----------+-------------+-------------+
388 05-May 03:50 roxie-sd: 3302 Autochanger "loaded drive 0", result is Slot 11.
389 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Warning: Director wanted Volume "LT
390 Current Volume "LT0-002" not acceptable because:
391 1997 Volume "LT0-002" not in catalog.
392 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Error: Autochanger Volume "LT0-002"
393 Setting InChanger to zero in catalog.
394 05-May 03:50 roxie-dir: Tibs.2006-05-05_03.05.02 Error: Unable to get Media record
396 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: Error getting Volume i
397 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: Job 530 canceled.
398 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: spool.c:249 Fatal appe
399 05-May 03:49 Tibs: Tibs.2006-05-05_03.05.02 Fatal error: c:\cygwin\home\kern\bacula
408 FirstWritten: 2006-05-05 03:11:54
409 LastWritten: 2006-05-05 03:50:23
410 LabelDate: 2005-12-26 16:52:40
421 VolRetention: 31,536,000
433 Note VolStatus is blank!!!!!
440 FirstWritten: 0000-00-00 00:00:00
441 LastWritten: 0000-00-00 00:00:00
442 LabelDate: 2005-12-26 16:52:40
453 VolRetention: 31,536,000
466 Automatically selected Storage: LTO-changer
467 Enter autochanger drive[0]: 0
468 3301 Issuing autochanger "loaded drive 0" command.
469 3302 Autochanger "loaded drive 0", result: nothing loaded.
470 3301 Issuing autochanger "loaded drive 0" command.
471 3302 Autochanger "loaded drive 0", result: nothing loaded.
472 3902 Cannot mount Volume on Storage Device "LTO-Drive1" (/dev/nst0) because:
473 Couldn't rewind device "LTO-Drive1" (/dev/nst0): ERR=dev.c:678 Rewind error on "LTO-Drive1" (/dev/nst0). ERR=No medium found.
475 3905 Device "LTO-Drive1" (/dev/nst0) open but no Bacula volume is mounted.
476 If this is not a blank tape, try unmounting and remounting the Volume.
478 - http://www.dwheeler.com/essays/commercial-floss.html
479 - Add VolumeLock to prevent all but lock holder (SD) from updating
480 the Volume data (with the exception of VolumeState).
481 - The btape fill command does not seem to use the Autochanger
482 - Make Windows installer default to system disk drive.
483 - Look at using ioctl(FIOBMAP, ...) on Linux, and
484 DeviceIoControl(..., FSCTL_QUERY_ALLOCATED_RANGES, ...) on
485 Win32 for sparse files.
486 http://www.flexhex.com/docs/articles/sparse-files.phtml
487 http://www.informatik.uni-frankfurt.de/~loizides/reiserfs/fibmap.html
488 - Directive: at <event> "command"
489 - Command: pycmd "command" generates "command" event. How to
490 attach to a specific job?
491 - Integrate Christopher's St. Bernard code.
492 - run_cmd() returns int should return JobId_t
493 - get_next_jobid_from_list() returns int should return JobId_t
494 - Document export LDFLAGS=-L/usr/lib64
495 - Don't attempt to restore from "Disabled" Volumes.
496 - Network error on Win32 should set Win32 error code.
497 - What happens when you rename a Disk Volume?
498 - Job retention period in a Pool (and hence Volume). The job would
500 - Look at -D_FORTIFY_SOURCE=2
501 - Add Win32 FileSet definition somewhere
502 - Look at fixing restore status stats in SD.
503 - Look at using ioctl(FIMAP) and FIGETBSZ for sparse files.
504 http://www.informatik.uni-frankfurt.de/~loizides/reiserfs/fibmap.html
505 - Implement a mode that says when a hard read error is
506 encountered, read many times (as it currently does), and if the
507 block cannot be read, skip to the next block, and try again. If
508 that fails, skip to the next file and try again, ...
510 create table LevelType (LevelType binary(1), LevelTypeLong tinyblob);
511 insert into LevelType (LevelType,LevelTypeLong) values
515 - Show files/second in client status output.
516 - Add a recursive mark command (rmark) to restore.
517 - "Minimum Job Interval = nnn" sets minimum interval between Jobs
518 of the same level and does not permit multiple simultaneous
519 running of that Job (i.e. lets any previous invocation finish
520 before doing Interval testing).
521 - Look at simplifying File exclusions.
522 - New directive "Delete purged Volumes"
523 - new pool XXX with ScratchPoolId = MyScratchPool's PoolId and
524 let it fill itself, and RecyclePoolId = XXX's PoolId so I can
525 see if it become stable and I just have to supervise
527 - If I want to remove this pool, I set RecyclePoolId = MyScratchPool's
528 PoolId, and when it is empty remove it.
529 - Figure out how to recycle Scratch volumes back to the Scratch Pool.
531 - Allow Check Labels to be used with Bacula labels.
532 - "Resuming" a failed backup (lost line for example) by using the
533 failed backup as a sort of "base" job.
535 - Email to the user when the tape is about to need changing x
536 days before it needs changing.
537 - Command to show next tape that will be used for a job even
538 if the job is not scheduled.
539 - From: Arunav Mandal <amandal@trolltech.com>
540 1. When jobs are running and bacula for some reason crashes or if I do a
541 restart it remembers and jobs it was running before it crashed or restarted
542 as of now I loose all jobs if I restart it.
544 2. When spooling and in the midway if client is disconnected for instance a
545 laptop bacula completely discard the spool. It will be nice if it can write
546 that spool to tape so there will be some backups for that client if not all.
548 3. We have around 150 clients machines it will be nice to have a option to
549 upgrade all the client machines bacula version automatically.
551 4. Atleast one connection should be reserved for the bconsole so at heavy load
552 I should connect to the director via bconsole which at sometimes I can't
554 5. Another most important feature that is missing, say at 10am I manually
555 started backup of client abc and it was a full backup since client abc has
556 no backup history and at 10.30am bacula again automatically started backup of
557 client abc as that was in the schedule. So now we have 2 multiple Full
558 backups of the same client and if we again try to start a full backup of
559 client backup abc bacula won't complain. That should be fixed.
561 - Fix bpipe.c so that it does not modify results pointer.
562 ***FIXME*** calling sequence should be changed.
563 - For Windows disaster recovery see http://unattended.sf.net/
564 - regardless of the retention period, Bacula will not prune the
565 last Full, Diff, or Inc File data until a month after the
566 retention period for the last Full backup that was done.
567 - update volume=xxx --- add status=Full
568 - Remove old spool files on startup.
569 - Exclude SD spool/working directory.
570 - Refuse to prune last valid Full backup. Same goes for Catalog.
572 - Make a callback when Rerun failed levels is called.
573 - Give Python program access to Scheduled jobs.
574 - Add setting Volume State via Python.
575 - Python script to save with Python, not save, save with Bacula.
576 - Python script to do backup.
578 - Change the Priority, Client, Storage, JobStatus (error)
579 at the start of a job.
580 - Why is SpoolDirectory = /home/bacula/spool; not reported
581 as an error when writing a DVD?
582 - Make bootstrap file handle multiple MediaTypes (SD)
583 - Remove all old Device resource code in Dir and code to pass it
584 back in SD -- better, rework it to pass back device statistics.
585 - Check locking of resources -- be sure to lock devices where previously
586 resources were locked.
587 - The last part is left in the spool dir.
590 - In restore don't compare byte count on a raw device -- directory
591 entry does not contain bytes.
595 MA = (last_MA * 3 + rate) / 4
596 rate = (bytes - last_bytes) / (runtime - last_runtime)
597 - Max Vols limit in Pool off by one?
598 - Implement Files/Bytes,... stats for restore job.
599 - Implement Total Bytes Written, ... for restore job.
600 - Despool attributes simultaneously with data in a separate
601 thread, rejoined at end of data spooling.
602 - 7. Implement new Console commands to allow offlining/reserving drives,
603 and possibly manipulating the autochanger (much asked for).
604 - Add start/end date editing in messages (%t %T, %e?) ...
605 - Add ClientDefs similar to JobDefs.
606 - Print more info when bextract -p accepts a bad block.
607 - Fix FD JobType to be set before RunBeforeJob in FD.
608 - Look at adding full Volume and Pool information to a Volume
609 label so that bscan can get *all* the info.
610 - If the user puts "Purge Oldest Volume = yes" or "Recycle Oldest Volume = yes"
611 and there is only one volume in the pool, refuse to do it -- otherwise
612 he fills the Volume, then immediately starts reusing it.
613 - Implement copies and stripes.
614 - Add history file to console.
615 - Each file on tape creates a JobMedia record. Peter has 4 million
616 files spread over 10000 tape files and four tapes. A restore takes
617 16 hours to build the restore list.
618 - Add and option to check if the file size changed during backup.
619 - Make sure SD deletes spool files on error exit.
620 - Delete old spool files when SD starts.
621 - When labeling tapes, if you enter 000026, Bacula uses
622 the tape index rather than the Volume name 000026.
623 - Add offline tape command to Bacula console.
625 Enter MediaId or Volume name: 32
626 Enter new Volume name: DLT-20Dec04
627 Automatically selected Pool: Default
628 Connecting to Storage daemon DLTDrive at 192.168.68.104:9103 ...
629 Sending relabel command from "DLT-28Jun03" to "DLT-20Dec04" ...
630 block.c:552 Write error at 0:0 on device /dev/nst0. ERR=Bad file descriptor.
631 Error writing final EOF to tape. This tape may not be readable.
632 dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
633 askdir.c:219 NULL Volume name. This shouldn't happen!!!
634 3912 Failed to label Volume: ERR=dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
635 Label command failed for Volume DLT-20Dec04.
636 Do not forget to mount the drive!!!
637 - Bug: if a job is manually scheduled to run later, it does not appear
638 in any status report and cannot be cancelled.
640 ==== Keeping track of deleted/new files ====
641 - To mark files as deleted, run essentially a Verify to disk, and
642 when a file is found missing (MarkId != JobId), then create
643 a new File record with FileIndex == -1. This could be done
644 by the FD at the same time as the backup.
646 My "trick" for keeping track of deletions is the following.
647 Assuming the user turns on this option, after all the files
648 have been backed up, but before the job has terminated, the
649 FD will make a pass through all the files and send their
650 names to the DIR (*exactly* the same as what a Verify job
651 currently does). This will probably be done at the same
652 time the files are being sent to the SD avoiding a second
653 pass. The DIR will then compare that to what is stored in
654 the catalog. Any files in the catalog but not in what the
655 FD sent will receive a catalog File entry that indicates
656 that at that point in time the file was deleted. This
657 either transmitted to the FD or simultaneously computed in
658 the FD, so that the FD can put a record on the tape that
659 indicates that the file has been deleted at this point.
660 A delete file entry could potentially be one with a FileIndex
661 of 0 or perhaps -1 (need to check if FileIndex is used for
662 some other thing as many of the Bacula fields are "overloaded"
665 During a restore, any file initially picked up by some
666 backup (Full, ...) then subsequently having a File entry
667 marked "delete" will be removed from the tree, so will not
668 be restored. If a file with the same name is later OK it
669 will be inserted in the tree -- this already happens. All
670 will be consistent except for possible changes during the
673 Since I'm on the subject, some of you may be wondering what
674 the utility of the in memory tree is if you are going to
675 restore everything (at least it comes up from time to time
676 on the list). Well, it is still *very* useful because it
677 allows only the last item found for a particular filename
678 (full path) to be entered into the tree, and thus if a file
679 is backed up 10 times, only the last copy will be restored.
680 I recently (last Friday) restored a complete directory, and
681 the Full and all the Differential and Incremental backups
682 spanned 3 Volumes. The first Volume was not even mounted
683 because all the files had been updated and hence backed up
684 since the Full backup was made. In this case, the tree
685 saved me a *lot* of time.
687 Make sure this information is stored on the tape too so
688 that it can be restored directly from the tape.
690 All the code (with the exception of formally generating and
691 saving the delete file entries) already exists in the Verify
692 Catalog command. It explicitly recognizes added/deleted files since
693 the last InitCatalog. It is more or less a "simple" matter of
694 taking that code and adapting it slightly to work for backups.
696 Comments from Martin Simmons (I think they are all covered):
697 Ok, that should cover the basics. There are few issues though:
699 - Restore will depend on the catalog. I think it is better to include the
700 extra data in the backup as well, so it can be seen by bscan and bextract.
702 - I'm not sure if it will preserve multiple hard links to the same inode. Or
703 maybe adding or removing links will cause the data to be dumped again?
705 - I'm not sure if it will handle renamed directories. Possibly it will work
706 by dumping the whole tree under a renamed directory?
708 - It remains to be seen how the backup performance of the DIR's will be
709 affected when comparing the catalog for a large filesystem.
713 How about introducing a Type = MgmtPolicy job type? That job type would
714 be responsible for scanning the Bacula environment looking for specific
715 conditions, and submitting the appropriate jobs for implementing said
719 Name = "Migration-Policy"
721 Policy Selection Job Type = Migrate
722 Scope = "<keyword> <operator> <regexp>"
723 Threshold = "<keyword> <operator> <regexp>"
724 Job Template = <template-name>
727 Where <keyword> is any legal job keyword, <operator> is a comparison
728 operator (=,<,>,!=, logical operators AND/OR/NOT) and <regexp> is a
729 appropriate regexp. I could see an argument for Scope and Threshold
730 being SQL queries if we want to support full flexibility. The
731 Migration-Policy job would then get scheduled as frequently as a site
732 felt necessary (suggested default: every 15 minutes).
737 Name = "Migration-Policy"
739 Policy Selection Job Type = Migration
741 Threshold = "Migration Selection Type = LowestUtil"
742 Job Template = "MigrationTemplate"
745 would select all pools for examination and generate a job based on
746 MigrationTemplate to automatically select the volume with the lowest
747 usage and migrate it's contents to the nextpool defined for that pool.
749 This policy abstraction would be really handy for adjusting the behavior
750 of Bacula according to site-selectable criteria (one thing that pops
751 into mind is Amanda's ability to automatically adjust backup levels
752 depending on various criteria).
758 - Add Pool/Storage override regression test.
759 - Add delete JobId to regression.
760 - Add a regression test for dbcheck.
761 - New test to add bscan to four-concurrent-jobs regression,
762 i.e. after the four-concurrent jobs zap the
763 database as is done in the bscan-test, then use bscan to
764 restore the database, do a restore and compare with the
766 - Add restore of specific JobId to regression (item 3
767 on the restore prompt)
768 - Add IPv6 to regression
769 - Add database test to regression. Test each function like delete,
772 - AntiVir can slow down backups on Win32 systems.
773 - Win32 systems with FAT32 can be much slower than NTFS for
774 more than 1000 files per directory.
778 - A HOLD command to stop all jobs from starting.
779 - A PAUSE command to pause all running jobs ==> release the
781 - Media Type = LTO,LTO-2,LTO-3
782 Media Type Read = LTO,LTO2,LTO3
783 Media Type Write = LTO2, LTO3
785 === From Carsten Menke <bootsy52@gmx.net>
787 Following is a list of what I think in the situations where I'm faced with,
788 could be a usefull enhancement to bacula, which I'm certain other users will
789 benefit from as well.
791 1. NextJob/NextJobs Directive within a Job Resource in the form of
792 NextJobs = job1,job2.
795 I currently solved the problem with running multiple jobs each after each
796 by setting the Max Wait Time for a job to 8 hours, and give
797 the jobs different Priorities. However, there scenarios where
798 1 Job is directly depending on another job, so if the former job fails,
799 the job after it needn't to be run
800 while maybe other jobs should run despite of that
803 A Backup Job and a Verify job, if the backup job fails there is no need to run
804 the verify job, as the backup job already failed. However, one may like
805 to backup the Catalog to disk despite of that the main backup job failed.
808 I see that this is related to the Event Handlers which are on the ToDo
809 list, also it is maybe a good idea to check for the return value and
810 execute different actions based on the return value
813 3. offline capability to bconsole
816 Currently I use a script which I execute within the last Job via the
817 RunAfterJob Directive, to release and eject the tape.
818 So I have to call bconsole "release=Storage-Name" and afterwards
819 mt -f /dev/nst0 eject to get the tape out.
821 If I have multiple Storage Devices, than these may not be /dev/nst0 and
822 I have to modify the script or call it with parameters etc.
823 This would actually not be needed, as everything is already defined
824 in bacula-sd.conf and if I can invoke bconsole with the
825 storage name via $1 in the script than I'm done and information is
828 4. %s for Storage Name added to the chars being substituted in "RunAfterJob"
832 For the reason mentioned in 3. to have the ability to call a
833 script with /scripts/foobar %s and in the script use $1
834 to pass the Storage Name to bconsole
836 5. Setting Volume State within a Job Resource
839 Instead of using "Maximum Volume Jobs" in the Pool Resource,
840 I would have the possibilty to define
841 in a Job Resource that after this certain job is run, the Volume State
842 should be set to "Volume State = Used", this give more flexibility (IMHO).
844 6. Localization of Bacula Messages
847 Unfortunatley many,many people I work with don't speak english very well.
848 So if at least the Reporting messages would be localized then they
849 would understand that they have to change the tape,etc. etc.
851 I volunteer to do the german translations, and if I can convince my wife also
852 french and Morre (western african language).
854 7. OK, this is evil, probably bound to security risks and maybe not possible
855 due to the design of bacula.
857 Implementation of Backtics ( `command` ) for shell comand execution to
858 the "Label Format" Directive.
862 Currently I have defined BACULA_DAY_OF_WEEK="day1|day2..." resulting in
863 Label Format = "HolyBackup-${BACULA_DAY_OF_WEEK[${WeekDay}]}". If I could
864 use backticks than I could use "Label Format = HolyBackup-`date +%A` to have
865 the localized name for the day of the week appended to the
866 format string. Then I have the tape labeled automatically with weekday
867 name in the correct language.
869 - Yes, that is surely the case. I probably should turn those into Warning
870 errors. In addition, you just made me think that it might not be bad to
871 add an option to check the file size after backing up the file and
872 report if it changes. This would be done as an option because it would
875 Kern, good idea. If you do do that, mention in the output: file
876 shrunk, or file expanded, just to make it obvious to the user
877 (without having to the refer to file size), just how the file size
880 Would this option be for all file, or just one file? Or a fileset?
881 - Make output from status use html table tags for nicely
882 presenting in a browser.
883 - Can one write tapes faster with 8192 byte block sizes?
884 - Document security problems with the same password for everyone in
885 rpm and Win32 releases.
886 - Browse generations of files.
887 - I've seen an error when my catalog's File table fills up. I
888 then have to recreate the File table with a larger maximum row
889 size. Relevant information is at
890 http://dev.mysql.com/doc/mysql/en/Full_table.html ; I think the
891 "Installing and Configuring MySQL" chapter should talk a bit
892 about this potential problem, and recommend a solution.
893 - For Solaris must use POSIX awk.
894 - Want speed of writing to tape while despooling.
895 - Supported autochanger:
903 Wangtek 6525ES (SCSI-1 QIC drive, 525MB), under Linux 2.4.something,
904 bacula 1.36.0/1 works with blocksize 16k INSIDE bacula-sd.conf.
905 - Add regex from http://www.pcre.org to Bacula for Win32.
906 - Use only shell tools no make in CDROM package.
907 - Include within include does it work?
908 - Implement a Pool of type Cleaning?
909 - Implement VolReadTime and VolWriteTime in SD
910 - Modify Backing up Your Database to include a bootstrap file.
911 - Think about making certain database errors fatal.
912 - Look at correcting the time jump in the scheduler for daylight
913 savings time changes.
914 - Add a "real" timer to network connections.
915 - Promote to Full = Time period
916 - Check dates entered by user for correctness (month/day/... ranges)
917 - Compress restore Volume listing by date and first file.
918 - Look at patches/bacula_db.b2z postgresql that loops during restore.
920 - Perhaps add read/write programs and/or plugins to FileSets.
921 - How to handle backing up portables ...
922 - Add some sort of guaranteed Interval for upgrading jobs.
923 - Can we write the state file after every job terminates? On Win32
924 the system crashes and the state file is not updated.
927 Documentation to do: (any release a little bit at a time)
928 - Doc to do unmount before removing magazine.
929 - Alternative to static linking "ldd prog" save all binaries listed,
930 restore them and point LD_LIBRARY_PATH to them.
931 - Document add "</dev/null >/dev/null 2>&1" to the bacula-fd command line
932 - Document query file format.
933 - Add more documentation for bsr files.
934 - Document problems with Verify and pruning.
935 - Document how to use multiple databases.
936 - VXA drives have a "cleaning required"
937 indicator, but Exabyte recommends preventive cleaning after every 75
940 In this context, it should be noted that Exabyte has a command-line
941 vxatool utility available for free download. (The current version is
942 vxatool-3.72.) It can get diagnostic info, read, write and erase tapes,
943 test the drive, unload tapes, change drive settings, flash new firmware,
945 Of particular interest in this context is that vxatool <device> -i will
946 report, among other details, the time since last cleaning in tape motion
947 minutes. This information can be retrieved (and settings changed, for
948 that matter) through the generic-SCSI device even when Bacula has the
949 regular tape device locked. (Needless to say, I don't recommend
950 changing tape settings while a job is running.)
951 - Lookup HP cleaning recommendations.
952 - Lookup HP tape replacement recommendations (see trouble shooting autochanger)
953 - Document doing table repair
956 ===================================
957 - Add macro expansions in JobDefs.
958 Run Before Job = "SomeFile %{Level} %{Client}"
959 Write Bootstrap="/some/dir/%{JobName}_%{Client}.bsr"
960 - Use non-blocking network I/O but if no data is available, use
962 - Use gather write() for network I/O.
963 - Autorestart on crash.
964 - Add bandwidth limiting.
965 - Add acks every once and a while from the SD to keep
966 the line from timing out.
967 - When an error in input occurs and conio beeps, you can back
968 up through the prompt.
969 - Detect fixed tape block mode during positioning by looking at
970 block numbers in btape "test". Possibly adjust in Bacula.
971 - Fix list volumes to output volume retention in some other
972 units, perhaps via a directive.
973 - Allow Simultaneous Priorities = yes => run up to Max concurrent jobs even
974 with multiple priorities.
975 - If you use restore replace=never, the directory attributes for
976 non-existent directories will not be restored properly.
978 - see lzma401.zip in others directory for new compression
980 - Allow the user to select JobType for manual pruning/purging.
981 - bscan does not put first of two volumes back with all info in
983 - Implement the FreeBSD nodump flag in chflags.
984 - Figure out how to make named console messages go only to that
985 console and to the non-restricted console (new console class?).
986 - Make restricted console prompt for password if *ask* is set or
987 perhaps if password is undefined.
988 - Implement "from ISO-date/time every x hours/days/weeks/months" in
991 ==== from Marc Schoechlin
992 - the help-command should be more verbose
993 (it should explain the paramters of the different
995 -> its time-comsuming to consult the manual anytime
996 you need a special parameter
997 -> maybe its more easy to maintain this, if the
998 descriptions of that commands are outsourced to
1000 - the cd-command should allow complete paths
1001 i.e. cd /foo/bar/foo/bar
1002 -> if a customer mails me the path to a certain file,
1003 its faster to enter the specified directory
1004 - if the password is not configured in bconsole.conf
1005 you should be asked for it.
1006 -> sometimes you like to do restore on a customer-machine
1007 which shouldnt know the password for bacula.
1008 -> adding the password to the file favours admins
1009 to forget to remove the password after usage
1011 the protection of that file is less important
1012 - long-listed-output of commands should be scrollable
1013 like the unix more/less-command does
1014 -> if someone runs 200 and more machines, the lists could
1015 be a little long and complex
1016 - command-output should be shown column by column
1017 to reduce scrolling and to increase clarity
1019 - lsmark should list the selected files with full
1021 - wildcards for selecting and file and directories would be nice
1022 - any actions should be interuptable with STRG+C
1023 - command-expansion would be pretty cool
1025 - When the replace Never option is set, new directory permissions
1026 are not restored. See bug 213. To fix this requires creating a
1027 list of newly restored directories so that those directory
1028 permissions *can* be restored.
1029 - Add prune all command
1030 - Document fact that purge can destroy a part of a restore by purging
1031 one volume while others remain valid -- perhaps mark Jobs.
1032 - Add multiple-media-types.txt
1033 - look at mxt-changer.html
1034 - Make ? do a help command (no return needed).
1035 - Implement restore directory.
1036 - Document streams and how to implement them.
1037 - Try not to re-backup a file if a new hard link is added.
1038 - Add feature to backup hard links only, but not the data.
1039 - Fix stream handling to be simpler.
1040 - Add Priority and Bootstrap to Run a Job.
1041 - Eliminate Restore "Run Restore Job" prompt by allowing new "run command
1043 - Remove View FileSet button from Run a Job dialog.
1044 - Handle prompt for restore job at end of Restore command.
1045 - Add display of total selected files to Restore window.
1046 - Add tree pane to left of window.
1047 - Add progress meter.
1048 - Max wait time or max run time causes seg fault -- see runtime-bug.txt
1049 - Add message to user to check for fixed block size when the forward
1050 space test fails in btape.
1051 - When unmarking a directory check if all files below are unmarked and
1052 then remove the + flag -- in the restore tree.
1053 - Possibly implement: Action = Unmount Device="TapeDrive1" in Admin jobs.
1054 - Setup lrrd graphs: (http://www.linpro.no/projects/lrrd/) Mike Acar.
1055 - Revisit the question of multiple Volumes (disk) on a single device.
1056 - Add a block copy option to bcopy.
1057 - Finish work on Gnome restore GUI.
1058 - Fix "llist jobid=xx" where no fileset or client exists.
1059 - For each job type (Admin, Restore, ...) require only the really necessary
1060 fields.- Pass Director resource name as an option to the Console.
1061 - Add a "batch" mode to the Console (no unsolicited queries, ...).
1062 - Add a .list all files in the restore tree (probably also a list all files)
1063 Do both a long and short form.
1064 - Allow browsing the catalog to see all versions of a file (with
1065 stat data on each file).
1066 - Restore attributes of directory if replace=never set but directory
1068 - Use SHA1 on authentication if possible.
1069 - See comtest-xxx.zip for Windows code to talk to USB.
1070 - Add John's appended files:
1071 Appended = { /files/server/logs/http/*log }
1072 and such files would be treated as follows.On a FULL backup, they would
1073 be backed up like any other file.On an INCREMENTAL backup, where a
1074 previous INCREMENTAL or FULL was already in thecatalogue and the length
1075 of the file wasgreater than the length of the last backup, only thedata
1076 added since the last backup will be dumped.On an INCREMENTAL backup, if
1077 the length of the file is less than thelength of the file with the same
1078 name last backed up, the completefile is dumped.On Windows systems, with
1079 creation date of files, we can be evensmarter about this and not count
1080 entirely upon the length.On a restore, the full and all incrementals
1081 since it will beapplied in sequence to restore the file.
1082 - Check new HAVE_WIN32 open bits.
1083 - Check if the tape has moved before writing.
1084 - Handling removable disks -- see below:
1085 - Keep track of tape use time, and report when cleaning is necessary.
1086 - Add FromClient and ToClient keywords on restore command (or
1087 BackupClient RestoreClient).
1088 - Implement a JobSet, which groups any number of jobs. If the
1089 JobSet is started, all the jobs are started together.
1090 Allow Pool, Level, and Schedule overrides.
1091 - Enhance cancel to timeout BSOCK packets after a specific delay.
1092 - Do scheduling by UTC using gmtime_r() in run_conf, scheduler, and
1093 ua_status.!!! Thanks to Alan Brown for this tip.
1094 - Look at updating Volume Jobs so that Max Volume Jobs = 1 will work
1095 correctly for multiple simultaneous jobs.
1096 - Correct code so that FileSet MD5 is calculated for < and | filename
1098 - Implement the Media record flag that indicates that the Volume does disk
1100 - Implement VolAddr, which is used when Volume is addressed like a disk,
1101 and form it from VolFile and VolBlock.
1102 - Make multiple restore jobs for multiple media types specifying
1103 the proper storage type.
1104 - Fix fast block rejection (stored/read_record.c:118). It passes a null
1105 pointer (rec) to try_repositioning().
1106 - Look at extracting Win data from BackupRead.
1107 - Implement RestoreJobRetention? Maybe better "JobRetention" in a Job,
1108 which would take precidence over the Catalog "JobRetention".
1109 - Implement Label Format in Add and Label console commands.
1110 - Possibly up network buffers to 65K. Put on variable.
1111 - Put email tape request delays on one or more variables. User wants
1112 to cancel the job after a certain time interval. Maximum Mount Wait?
1113 - Job, Client, Device, Pool, or Volume?
1114 Is it possible to make this a directive which is *optional* in multiple
1115 resources, like Level? If so, I think I'd make it an optional directive
1116 in Job, Client, and Pool, with precedence such that Job overrides Client
1117 which in turn overrides Pool.
1119 - New Storage specifications:
1120 - Want to write to multiple storage devices simultaneously
1121 - Want to write to multiple storage devices sequentially (in one job)
1122 - Want to read/write simultaneously
1123 - Key is MediaType -- it must match
1125 Passed to SD as a sort of BSR record called Storage Specification
1129 MediaType -> Next MediaType
1131 Device -> Next Device
1133 Allow multiple Storage specifications
1141 Allow Multiple Pool specifications (note, Pool currently
1143 Allow Multiple MediaType specifications in Dir conf
1144 Allow Multiple Device specifications in Dir conf
1145 Perhaps keep this in a single SSR
1146 Tie a Volume to a specific device by using a MediaType that
1147 is contained in only one device.
1148 In SD allow Device to have Multiple MediaTypes
1150 - Ideas from Jerry Scharf:
1151 First let's point out some big pluses that bacula has for this
1153 more importantly it's active. Thank you so much for that
1154 even more important, it's not flaky
1155 it has an open access catalog, opening many possibilities
1156 it's pushing toward heterogeneous systems capability
1158 Macintosh file client
1159 macs are an interesting niche, but I fear a server is a rathole
1160 working bare iron recovery for windows
1161 the option for inc/diff backups not reset on fileset revision
1162 a) use both change and inode update time against base time
1163 b) do the full catalog check (expensive but accurate)
1164 sizing guide (how much system is needed to back up N systems/files)
1165 consultants on using bacula in building a disaster recovery system
1166 an integration guide
1167 or how to get at fancy things that one could do with bacula
1168 logwatch code for bacula logs (or similar)
1169 linux distro inclusion of bacula (brings good and bad, but necessary)
1170 win2k/XP server capability (icky but you asked)
1171 support for Oracle database ??
1173 - Look at adding SQL server and Exchange support for Windows.
1174 - Make dev->file and dev->block_num signed integers so that -1 can
1175 be an invalid value which happens with BSR.
1176 - Create VolAddr for disk files in place of VolFile and VolBlock. This
1177 is needed to properly specify ranges.
1178 - Add progress of files/bytes to SD and FD.
1179 - Print warning message if FileId > 4 billion
1180 - do a "messages" before the first prompt in Console
1181 - Client does not show busy during Estimate command.
1182 - Implement Console mtx commands.
1183 - Implement a Mount Command and an Unmount Command where
1184 the users could specify a system command to be performed
1185 to do the mount, after which Bacula could attempt to
1186 read the device. This is for Removeable media such as a CDROM.
1187 - Most likely, this mount command would be invoked explicitly
1188 by the user using the current Console "mount" and "unmount"
1189 commands -- the Storage Daemon would do the right thing
1190 depending on the exact nature of the device.
1191 - As with tape drives, when Bacula wanted a new removable
1192 disk mounted, it would unmount the old one, and send a message
1193 to the user, who would then use "mount" as described above
1194 once he had actually inserted the disk.
1195 - Implement dump/print label to UA
1196 - Spool to disk only when the tape is full, then when a tape is hung move
1198 - bextract is sending everything to the log file ****FIXME****
1199 - Allow multiple Storage specifications (or multiple names on
1200 a single Storage specification) in the Job record. Thus a job
1201 can be backed up to a number of storage devices.
1202 - Implement some way for the File daemon to contact the Director
1203 to start a job or pass its DHCP obtained IP number.
1204 - Implement a query tape prompt/replace feature for a console
1205 - Copy console @ code to gnome2-console
1206 - Make tree walk routines like cd, ls, ... more user friendly
1207 by handling spaces better.
1208 - Make sure that Bacula rechecks the tape after the 20 min wait.
1209 - Set IO_NOWAIT on Bacula TCP/IP packets.
1210 - Try doing a raw partition backup and restore by mounting a
1212 - From Lars Kellers:
1213 Yes, it would allow to highly automatic the request for new tapes. If a
1214 tape is empty, bacula reads the barcodes (native or simulated), and if
1215 an unused tape is found, it runs the label command with all the
1216 necessary parameters.
1218 By the way can bacula automatically "move" an empty/purged volume say
1219 in the "short" pool to the "long" pool if this pool runs out of volume
1221 - What to do about "list files job=xxx".
1222 - Look at how fuser works and /proc/PID/fd that is how Nic found the
1223 file descriptor leak in Bacula.
1224 - Implement WrapCounters in Counters.
1225 - Add heartbeat from FD to SD if hb interval expires.
1226 - Can we dynamically change FileSets?
1227 - If pool specified to label command and Label Format is specified,
1228 automatically generate the Volume name.
1229 - Why can't SQL do the filename sort for restore?
1230 - Add ExhautiveRestoreSearch
1231 - Look at the possibility of loading only the necessary
1232 data into the restore tree (i.e. do it one directory at a
1233 time as the user walks through the tree).
1234 - Possibly use the hash code if the user selects all for a restore command.
1235 - Fix "restore all" to bypass building the tree.
1236 - Prohibit backing up archive device (findlib/find_one.c:128)
1237 - Implement Release Device in the Job resource to unmount a drive.
1238 - Implement Acquire Device in the Job resource to mount a drive,
1239 be sure this works with admin jobs so that the user can get
1240 prompted to insert the correct tape. Possibly some way to say to
1241 run the job but don't save the files.
1242 - Make things like list where a file is saved case independent for
1244 - Use autochanger to handle multiple devices.
1245 - Implement a Recycle command
1246 - Start working on Base jobs.
1247 - Implement UnsavedFiles DB record.
1248 - From Phil Stracchino:
1249 It would probably be a per-client option, and would be called
1250 something like, say, "Automatically purge obsoleted jobs". What it
1251 would do is, when you successfully complete a Differential backup of a
1252 client, it would automatically purge all Incremental backups for that
1253 client that are rendered redundant by that Differential. Likewise,
1254 when a Full backup on a client completed, it would automatically purge
1255 all Differential and Incremental jobs obsoleted by that Full backup.
1256 This would let people minimize the number of tapes they're keeping on
1257 hand without having to master the art of retention times.
1258 - When doing a Backup send all attributes back to the Director, who
1259 would then figure out what files have been deleted.
1260 - Currently in mount.c:236 the SD simply creates a Volume. It should have
1261 explicit permission to do so. It should also mark the tape in error
1262 if there is an error.
1263 - Cancel waiting for Client connect in SD if FD goes away.
1265 - Implement timeout in response() when it should come quickly.
1266 - Implement a Slot priority (loaded/not loaded).
1267 - Implement "vacation" Incremental only saves.
1268 - Implement create "FileSet"?
1269 - Add prefixlinks to where or not where absolute links to FD.
1270 - Issue message to mount a new tape before the rewind.
1271 - Simplified client job initiation for portables.
1272 - If SD cannot open a drive, make it periodically retry.
1273 - Add more of the config info to the tape label.
1275 - Refine SD waiting output:
1276 Device is being positioned
1277 > Device is being positioned for append
1278 > Device is being positioned to file x
1280 - Figure out some way to estimate output size and to avoid splitting
1281 a backup across two Volumes -- this could be useful for writing CDROMs
1282 where you really prefer not to have it split -- not serious.
1283 - Have SD compute MD5 or SHA1 and compare to what FD computes.
1284 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
1285 actual data on the Volume and compare it.
1286 - Implement Bacula plugins -- design API
1287 - Make bcopy read through bad tape records.
1288 - Program files (i.e. execute a program to read/write files).
1289 Pass read date of last backup, size of file last time.
1290 - Add Signature type to File DB record.
1291 - CD into subdirectory when open()ing files for backup to
1292 speed up things. Test with testfind().
1293 - Priority job to go to top of list.
1294 - Why are save/restore of device different sizes (sparse?) Yup! Fix it.
1295 - Implement some way for the Console to dynamically create a job.
1296 - Solaris -I on tar for include list
1297 - Need a verbose mode in restore, perhaps to bsr.
1298 - bscan without -v is too quiet -- perhaps show jobs.
1299 - Add code to reject whole blocks if not wanted on restore.
1300 - Check if we can increase Bacula FD priorty in Win2000
1301 - Make sure the MaxVolFiles is fully implemented in SD
1302 - Check if both CatalogFiles and UseCatalog are set to SD.
1303 - Possibly add email to Watchdog if drive is unmounted too
1304 long and a job is waiting on the drive.
1305 - After unmount, if restore job started, ask to mount.
1306 - Add UA rc and history files.
1307 - put termcap (used by console) in ./configure and
1308 allow -with-termcap-dir.
1309 - Fix Autoprune for Volumes to respect need for full save.
1310 - Compare tape to Client files (attributes, or attributes and data)
1311 - Make all database Ids 64 bit.
1312 - Allow console commands to detach or run in background.
1313 - Add SD message variables to control operator wait time
1314 - Maximum Operator Wait
1315 - Minimum Message Interval
1316 - Maximum Message Interval
1317 - Send Operator message when cannot read tape label.
1318 - Verify level=Volume (scan only), level=Data (compare of data to file).
1319 Verify level=Catalog, level=InitCatalog
1321 - Add keyword search to show command in Console.
1322 - Events : tape has more than xxx bytes.
1323 - Complete code in Bacula Resources -- this will permit
1324 reading a new config file at any time.
1325 - Handle ctl-c in Console
1326 - Implement script driven addition of File daemon to config files.
1327 - Think about how to make Bacula work better with File (non-tape) archives.
1328 - Write Unix emulator for Windows.
1329 - Put memory utilization in Status output of each daemon
1330 if full status requested or if some level of debug on.
1331 - Make database type selectable by .conf files i.e. at runtime
1332 - Set flag for uname -a. Add to Volume label.
1333 - Restore files modified after date
1334 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
1335 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
1336 - Timout a job or terminate if link goes down, or reopen link and query.
1337 - Concept of precious tapes (cannot be reused).
1338 - Make bcopy copy with a single tape drive.
1339 - Permit changing ownership during restore.
1342 > My suggestion: Add a feature on the systray menu-icon menu to request
1343 > an immediate backup now. This would be useful for laptop users who may
1344 > not be on the network when the regular scheduled backup is run.
1346 > My wife's suggestion: Add a setting to the win32 client to allow it to
1347 > shut down the machine after backup is complete (after, of course,
1348 > displaying a "System will shut down in one minute, click here to cancel"
1349 > warning dialog). This would be useful for sites that want user
1350 > woorkstations to be shut down overnight to save power.
1353 - Autolabel should be specified by DIR instead of SD.
1355 - Add media capacity
1356 - AutoScan (check checksum of tape)
1357 - Format command = "format /dev/nst0"
1361 - Seek resolution (usually corresponds to buffer size)
1362 - EODErrorCode=ENOSPC or code
1363 - Partial Read error code
1364 - Partial write error code
1365 - Nonformatted read error
1366 - Nonformatted write error
1367 - WriteProtected error
1371 - IgnoreCloseErrors=yes
1381 - FD sends unsaved file list to Director at end of job (see
1383 - File daemon should build list of files skipped, and then
1384 at end of save retry and report any errors.
1385 - Write a Storage daemon that uses pipes and
1386 standard Unix programs to write to the tape.
1388 - Need something that monitors the JCR queue and
1389 times out jobs by asking the deamons where they are.
1390 - Enhance Jmsg code to permit buffering and saving to disk.
1391 - device driver = "xxxx" for drives.
1392 - Verify from Volume
1393 - Ensure that /dev/null works
1394 - Need report class for messages. Perhaps
1395 report resource where report=group of messages
1396 - enhance scan_attrib and rename scan_jobtype, and
1397 fill in code for "since" option
1398 - Director needs a time after which the report status is sent
1399 anyway -- or better yet, a retry time for the job.
1400 - Don't reschedule a job if previous incarnation is still running.
1401 - Some way to automatically backup everything is needed????
1402 - Need a structure for pending actions:
1404 - termination status (part of buffered msgs?)
1406 Read, Write, Clean, Delete
1407 - Login to Bacula; Bacula users with different permissions:
1408 owner, group, user, quotas
1409 - Store info on each file system type (probably in the job header on tape.
1410 This could be the output of df; or perhaps some sort of /etc/mtab record.
1412 ========= ideas ===============
1413 From: "Jerry K. Schieffer" <jerry@skylinetechnology.com>
1414 To: <kern@sibbald.com>
1415 Subject: RE: [Bacula-users] future large programming jobs
1416 Date: Thu, 26 Feb 2004 11:34:54 -0600
1418 I noticed the subject thread and thought I would offer the following
1419 merely as sources of ideas, i.e. something to think about, not even as
1420 strong as a request. In my former life (before retiring) I often
1421 dealt with backups and storage management issues/products as a
1422 developer and as a consultant. I am currently migrating my personal
1423 network from amanda to bacula specifically because of the ability to
1424 cross media boundaries during storing backups.
1425 Are you familiar with the commercial product called ADSM (I think IBM
1426 now sells it under the Tivoli label)? It has a couple of interesting
1427 ideas that may apply to the following topics.
1429 1. Migration: Consider that when you need to restore a system, there
1430 may be pressure to hurry. If all the information for a single client
1431 can eventually end up on the same media (and in chronological order),
1432 the restore is facillitated by not having to search past information
1433 from other clients. ADSM has the concept of "client affinity" that
1434 may be associated with it's storage pools. It seems to me that this
1435 concept (as an optional feature) might fit in your architecture for
1438 ADSM also has the concept of defining one or more storage pools as
1439 "copy pools" (almost mirrors, but only in the sense of contents).
1440 These pools provide the ability to have duplicte data stored both
1441 onsite and offsite. The copy process can be scheduled to be handled
1442 by their storage manager during periods when there is no backup
1443 activity. Again, the migration process might be a place to consider
1444 implementing something like this.
1447 > It strikes me that it would be very nice to be able to do things
1449 > have the Job(s) backing up the machines run, and once they have all
1450 > completed, start a migration job to copy the data from disks Volumes
1452 > a tape library and then to offsite storage. Maybe this can already
1454 > done with some careful scheduling and Job prioritzation; the events
1455 > mechanism described below would probably make it very easy.
1457 This is the goal. In the first step (before events), you simply
1459 the Migration to tape later.
1461 2. Base jobs: In ADSM, each copy of each stored file is tracked in
1462 the database. Once a file (unique by path and metadata such as dates,
1463 size, ownership, etc.) is in a copy pool, no more copies are made. In
1464 other words, when you start ADSM, it begins like your concept of a
1465 base job. After that it is in the "incremental" mode. You can
1466 configure the number of "generations" of files to be retained, plus a
1467 retention date after which even old generations are purged. The
1468 database tracks the contents of media and projects the percentage of
1469 each volume that is valid. When the valid content of a volume drops
1470 below a configured percentage, the valid data are migrated to another
1471 volume and the old volume is marked as empty. Note, this requires
1472 ADSM to have an idea of the contents of a client, i.e. marking the
1473 database when an existing file was deleted, but this would solve your
1474 issue of restoring a client without restoring deleted files.
1476 This is pretty far from what bacula now does, but if you are going to
1477 rip things up for Base jobs,.....
1478 Also, the benefits of this are huge for very large shops, especially
1479 with media robots, but are a pain for shops with manual media
1483 > Base jobs sound pretty useful, but I'm not dying for them.
1485 Nobody is dying for them, but when you see what it does, you will die
1488 3. Restoring deleted files: Since I think my comments in (2) above
1489 have low probability of implementation, I'll also suggest that you
1490 could approach the issue of deleted files by a mechanism of having the
1491 fd report to the dir, a list of all files on the client for every
1492 backup job. The dir could note in the database entry for each file
1493 the date that the file was seen. Then if a restore as of date X takes
1494 place, only files that exist from before X until after X would be
1495 restored. Probably the major cost here is the extra date container in
1496 each row of the files table.
1498 Thanks for "listening". I hope some of this helps. If you want to
1499 contact me, please send me an email - I read some but not all of the
1500 mailing list traffic and might miss a reply there.
1502 Please accept my compliments for bacula. It is doing a great job for
1503 me!! I sympathize with you in the need to wrestle with excelence in
1504 execution vs. excelence in feature inclusion.
1509 ==============================
1512 - Design at hierarchial storage for Bacula. Migration and Clone.
1513 - Implement FSM (File System Modules).
1514 - Audit M_ error codes to ensure they are correct and consistent.
1515 - Add variable break characters to lex analyzer.
1516 Either a bit mask or a string of chars so that
1517 the caller can change the break characters.
1518 - Make a single T_BREAK to replace T_COMMA, etc.
1519 - Ensure that File daemon and Storage daemon can
1520 continue a save if the Director goes down (this
1521 is NOT currently the case). Must detect socket error,
1522 buffer messages for later.
1523 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
1524 - Add ability to backup to two Storage devices (two SD sessions) at
1525 the same time -- e.g. onsite, offsite.
1526 - Add the ability to consolidate old backup sets (basically do a restore
1527 to tape and appropriately update the catalog). Compress Volume sets.
1528 Might need to spool via file is only one drive is available.
1529 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
1530 someway to do so with every volume that has less than x% valid
1534 Migration: Move a backup from one Volume to another
1535 Clone: Copy a backup -- two Volumes
1538 ======================================================
1540 It is somewhat like a Full save becomes an incremental since
1541 the Base job (or jobs) plus other non-base files.
1543 - A Base backup is same as Full backup, just different type.
1544 - New BaseFiles table that contains:
1546 BaseJobId - Base JobId referenced for this FileId (needed ???)
1547 JobId - JobId currently running
1548 FileId - File not backed up, exists in Base Job
1549 FileIndex - FileIndex from Base Job.
1550 i.e. for each base file that exists but is not saved because
1551 it has not changed, the File daemon sends the JobId, BaseId,
1552 FileId, FileIndex back to the Director who creates the DB entry.
1553 - To initiate a Base save, the Director sends the FD
1554 the FileId, and full filename for each file in the Base.
1555 - When the FD finds a Base file, he requests the Director to
1556 send him the full File entry (stat packet plus MD5), or
1557 conversely, the FD sends it to the Director and the Director
1558 says yes or no. This can be quite rapid if the FileId is kept
1559 by the FD for each Base Filename.
1560 - It is probably better to have the comparison done by the FD
1561 despite the fact that the File entry must be sent across the
1563 - An alternative would be to send the FD the whole File entry
1564 from the start. The disadvantage is that it requires a lot of
1565 space. The advantage is that it requires less communications
1567 - The Job record must be updated to indicate that one or more
1569 - At end of Job, FD returns:
1570 1. Count of base files/bytes not written to tape (i.e. matches)
1571 2. Count of base file that were saved i.e. had changed.
1572 - No tape record would be written for a Base file that matches, in the
1573 same way that no tape record is written for Incremental jobs where
1574 the file is not saved because it is unchanged.
1575 - On a restore, all the Base file records must explicitly be
1576 found from the BaseFile tape. I.e. for each Full save that is marked
1577 to have one or more Base Jobs, search the BaseFile for all occurrences
1579 - An optimization might be to make the BaseFile have:
1585 This would avoid the need to explicitly fetch each File record for
1586 the Base job. The Base Job record will be fetched to get the
1587 VolSessionId and VolSessionTime.
1588 =========================================================
1591 ==========================================================
1593 For each Incremental job that is run, there may be files that
1594 were found but not saved because they were locked (this applies
1595 only to Windows). Such a system could send back to the Director
1596 a list of Unsaved files.
1598 - New UnSavedFiles table that contains:
1602 - Then in the next Incremental job, the list of Unsaved Files will be
1603 feed to the FD, who will ensure that they are explicitly chosen even
1604 if standard date/time check would not have selected them.
1605 =============================================================
1609 Multiple drive autochanger data: see Alan Brown
1610 mtx -f xxx unloadStorage Element 1 is Already Full(drive 0 was empty)
1611 Unloading Data Transfer Element into Storage Element 1...source Element
1612 Address 480 is Empty
1614 (drive 0 was empty and so was slot 1)
1615 > mtx -f xxx load 15 0
1616 no response, just returns to the command prompt when complete.
1617 > mtx -f xxx status Storage Changer /dev/changer:2 Drives, 60 Slots ( 2 Import/Export )
1618 Data Transfer Element 0:Full (Storage Element 15 Loaded):VolumeTag = HX001
1619 Data Transfer Element 1:Empty
1620 Storage Element 1:Empty
1621 Storage Element 2:Full :VolumeTag=HX002
1622 Storage Element 3:Full :VolumeTag=HX003
1623 Storage Element 4:Full :VolumeTag=HX004
1624 Storage Element 5:Full :VolumeTag=HX005
1625 Storage Element 6:Full :VolumeTag=HX006
1626 Storage Element 7:Full :VolumeTag=HX007
1627 Storage Element 8:Full :VolumeTag=HX008
1628 Storage Element 9:Full :VolumeTag=HX009
1629 Storage Element 10:Full :VolumeTag=HX010
1630 Storage Element 11:Empty
1631 Storage Element 12:Empty
1632 Storage Element 13:Empty
1633 Storage Element 14:Empty
1634 Storage Element 15:Empty
1635 Storage Element 16:Empty....
1636 Storage Element 28:Empty
1637 Storage Element 29:Full :VolumeTag=CLNU01L1
1638 Storage Element 30:Empty....
1639 Storage Element 57:Empty
1640 Storage Element 58:Full :VolumeTag=NEX261L2
1641 Storage Element 59 IMPORT/EXPORT:Empty
1642 Storage Element 60 IMPORT/EXPORT:Empty
1644 Unloading Data Transfer Element into Storage Element 15...done
1646 (just to verify it remembers where it came from, however it can be
1647 overrriden with mtx unload {slotnumber} to go to any storage slot.)
1649 There needs to be a table of drive # to devices somewhere - If there are
1650 multiple changers or drives there may not be a 1:1 correspondance between
1651 changer drive number and system device name - and depending on the way the
1652 drives are hooked up to scsi busses, they may not be linearly numbered
1653 from an offset point either.something like
1655 Autochanger drives = 2
1656 Autochanger drive 0 = /dev/nst1
1657 Autochanger drive 1 = /dev/nst2
1658 IMHO, it would be _safest_ to use explicit mtx unload commands at all
1659 times, not just for multidrive changers. For a 1 drive changer, that's
1665 MTX's manpage (1.2.15):
1666 unload [<slotnum>] [ <drivenum> ]
1667 Unloads media from drive <drivenum> into slot
1668 <slotnum>. If <drivenum> is omitted, defaults to
1669 drive 0 (as do all commands). If <slotnum> is
1670 omitted, defaults to the slot that the drive was
1671 loaded from. Note that there's currently no way
1672 to say 'unload drive 1's media to the slot it
1673 came from', other than to explicitly use that
1674 slot number as the destination.AB
1680 undef# camcontrol devlist
1681 <WANGTEK 51000 SCSI M74H 12B3> at scbus0 target 2 lun 0 (pass0,sa0)
1682 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 0 (pass1,sa1)
1683 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 1 (pass2)
1685 tapeinfo -f /dev/sg0 with a bad tape in drive 1:
1686 [kern@rufus mtx-1.2.17kes]$ ./tapeinfo -f /dev/sg0
1687 Product Type: Tape Drive
1689 Product ID: 'C5713A '
1691 Attached Changer: No
1692 TapeAlert[3]: Hard Error: Uncorrectable read/write error.
1693 TapeAlert[20]: Clean Now: The tape drive neads cleaning NOW.
1700 Medium Type: Not Loaded
1703 DataCompEnabled: yes
1704 DataCompCapable: yes
1705 DataDeCompEnabled: yes
1712 Handling removable disks
1714 From: Karl Cunningham <karlc@keckec.com>
1716 My backups are only to hard disk these days, in removable bays. This is my
1717 idea of how a backup to hard disk would work more smoothly. Some of these
1718 things Bacula does already, but I mention them for completeness. If others
1719 have better ways to do this, I'd like to hear about it.
1721 1. Accommodate several disks, rotated similar to how tapes are. Identified
1722 by partition volume ID or perhaps by the name of a subdirectory.
1723 2. Abort & notify the admin if the wrong disk is in the bay.
1724 3. Write backups to different subdirectories for each machine to be backed
1726 4. Volumes (files) get created as needed in the proper subdirectory, one
1728 5. When a disk is recycled, remove or zero all old backup files. This is
1729 important as the disk being recycled may be close to full. This may be
1730 better done manually since the backup files for many machines may be
1731 scattered in many subdirectories.
1736 - Why the heck doesn't bacula drop root priviledges before connecting to
1738 - Look at using posix_fadvise(2) for backups -- see bug #751.
1739 Possibly add the code at findlib/bfile.c:795
1740 /* TCP socket options */
1741 #define TCP_KEEPIDLE 4 /* Start keeplives after this period */
1742 - Fix bnet_connect() code to set a timer and to use time to
1744 - Implement 4th argument to make_catalog_backup that passes hostname.
1745 - Test FIFO backup/restore -- make regression
1746 - Please mount volume "xxx" on Storage device ... should also list
1747 Pool and MediaType in case user needs to create a new volume.
1748 - On restore add Restore Client, Original Client.
1749 01-Apr 00:42 rufus-dir: Start Backup JobId 55, Job=kernsave.2007-04-01_00.42.48
1750 01-Apr 00:42 rufus-sd: Python SD JobStart: JobId=55 Client=Rufus
1751 01-Apr 00:42 rufus-dir: Created new Volume "Full0001" in catalog.
1752 01-Apr 00:42 rufus-dir: Using Device "File"
1753 01-Apr 00:42 rufus-sd: kernsave.2007-04-01_00.42.48 Warning: Device "File" (/tmp) not configured to autolabel Volumes.
1754 01-Apr 00:42 rufus-sd: kernsave.2007-04-01_00.42.48 Warning: Device "File" (/tmp) not configured to autolabel Volumes.
1755 01-Apr 00:42 rufus-sd: Please mount Volume "Full0001" on Storage Device "File" (/tmp) for Job kernsave.2007-04-01_00.42.48
1756 01-Apr 00:44 rufus-sd: Wrote label to prelabeled Volume "Full0001" on device "File" (/tmp)
1757 - Check if gnome-console works with TLS.
1758 - the director seg faulted when I omitted the pool directive from a
1759 job resource. I was experimenting and thought it redundant that I had
1760 specified Pool, Full Backup Pool. and Differential Backup Pool. but
1761 apparently not. This happened when I removed the pool directive and
1762 started the director.
1763 - Add Where: client:/.... to restore job report.