9 - !!! Cannot restore two jobs a the same time that were
10 written simultaneously unless they were totally spooled.
11 - Document cleaning up the spool files:
12 db, pid, state, bsr, mail, conmsg, spool
13 - Document the multiple-drive-changer.txt script.
14 - Pruning with Admin job.
15 - Does WildFile match against full name? Doc.
16 - %d and %v only valid on Director, not for ClientRunBefore/After.
17 - During tests with the 260 char fix code, I found one problem:
18 if the system "sees" a long path once, it seems to forget it's
19 working drive (e.g. c:\), which will lead to a problem during
20 the next job (create bootstrap file will fail). Here is the
21 workaround: specify absolute working and pid directory in
22 bacula-fd.conf (e.g. c:\bacula\working instead of
24 - Document techniques for restoring large numbers of files.
25 - Document setting my.cnf to big file usage.
26 - Add example of proper index output to doc. show index from File;
27 - Correct the Include syntax in the m4.xxx files in examples/conf
28 - Document JobStatus and Termination codes.
29 - Fix the error with the "DVI file can't be opened" while
30 building the French PDF.
31 - Document more DVD stuff
39 - Document all the little details of setting up certificates for
40 the Bacula data encryption code.
41 - Document more precisely how to use master keys -- especially
42 for disaster recovery.
45 - Migration from other vendors
49 - Backup conf/exe (all daemons)
50 - Backup up system state
51 - Detect state change of system (verify)
54 - When Pool specifies Storage command override does not work.
55 - Please mount volume "xxx" on Storage device ... should also list
56 Pool and MediaType in case user needs to create a new volume.
57 - Implement wait_for_sysop() message display in wait_for_device(), which
58 now prints warnings too often.
60 - the director seg faulted when I omitted the pool directive from a
61 job resource. I was experimenting and thought it redundant that I had
62 specified Pool, Full Backup Pool. and Differential Backup Pool. but
63 apparently not. This happened when I removed the pool directive and
66 - On restore add Restore Client, Original Client.
67 01-Apr 00:42 rufus-dir: Start Backup JobId 55, Job=kernsave.2007-04-01_00.42.48
68 01-Apr 00:42 rufus-sd: Python SD JobStart: JobId=55 Client=Rufus
69 01-Apr 00:42 rufus-dir: Created new Volume "Full0001" in catalog.
70 01-Apr 00:42 rufus-dir: Using Device "File"
71 01-Apr 00:42 rufus-sd: kernsave.2007-04-01_00.42.48 Warning: Device "File" (/tmp) not configured to autolabel Volumes.
72 01-Apr 00:42 rufus-sd: kernsave.2007-04-01_00.42.48 Warning: Device "File" (/tmp) not configured to autolabel Volumes.
73 01-Apr 00:42 rufus-sd: Please mount Volume "Full0001" on Storage Device "File" (/tmp) for Job kernsave.2007-04-01_00.42.48
74 01-Apr 00:44 rufus-sd: Wrote label to prelabeled Volume "Full0001" on device "File" (/tmp)
76 - Add Where: client:/.... to restore job report.
77 - Ensure that each device in an Autochanger has a different
79 - Add Catalog = to Pool resource so that pools will exist
80 in only one catalog -- currently Pools are "global".
81 - Look at sg_logs -a /dev/sg0 for getting soft errors.
82 - btape "test" command with Offline on Unmount = yes
84 This test is essential to Bacula.
86 I'm going to write one record in file 0,
87 two records in file 1,
88 and three records in file 2
90 02-Feb 11:00 btape: ABORTING due to ERROR in dev.c:715
91 dev.c:714 Bad call to rewind. Device "LTO" (/dev/nst0) not open
92 02-Feb 11:00 btape: Fatal Error because: Bacula interrupted by signal 11: Segmentation violation
93 Kaboom! btape, btape got signal 11. Attempting traceback.
95 - Ensure that moving a purged Volume in ua_purge.c to the RecyclePool
97 - Why doesn't @"xxx abc" work in a conf file?
98 - Figure out some way to "automatically" backup conf changes.
99 - Add the OS version back to the Win32 client info.
100 - Restarted jobs have a NULL in the from field.
101 - Modify SD status command to indicate when the SD is writing
102 to a DVD (the device is not open -- see bug #732).
103 - Look at the possibility of adding "SET NAMES UTF8" for MySQL,
104 and possibly changing the blobs into varchar.
105 - Check if gnome-console works with TLS.
106 - Ensure that the SD re-reads the Media record if the JobFiles
107 does not match -- it may have been updated by another job.
108 - Look at moving the Storage directive from the Job to the
109 Pool in the default conf files.
111 - Test Volume compatibility between machine architectures
112 - Encryption documentation
113 - Wrong jobbytes with query 12 (todo)
114 - bacula-1.38.2-ssl.patch
115 - Bare-metal recovery Windows (todo)
122 - Add doc for bweb -- especially Installation
124 http://www.orangecrate.com/modules.php?name=News&file=article&sid=501
126 - FD-SD quick disconnect
127 - Despool attributes in separate thread
130 - Check why restore repeatedly sends Rechdrs between
131 each data chunk -- according to James Harper 9Jan07.
132 - Building the in memory restore tree is slow.
135 - Full at least once a month, ...
136 - Cancel Inc if Diff/Full running
137 - More intelligent re-run
138 - New/deleted file backup
140 - Incremental backup -- rsync, Stow
144 - Look at mondo/mindi
145 - Don't restore Solaris Door files:
146 #define S_IFDOOR in st_mode.
147 see: http://docs.sun.com/app/docs/doc/816-5173/6mbb8ae23?a=view#indexterm-360
148 - Make Bacula by default not backup tmpfs, procfs, sysfs, ...
149 - Fix hardlinked immutable files when linking a second file, the
150 immutable flag must be removed prior to trying to link it.
151 - Implement Python event for backing up/restoring a file.
152 - Change dbcheck to tell users to use native tools for fixing
153 broken databases, and to ensure they have the proper indexes.
154 - add udev rules for Bacula devices.
155 - If a job terminates, the DIR connection can close before the
156 Volume info is updated, leaving the File count wrong.
157 - Look at why SIGPIPE during connection can cause seg fault in
158 writing the daemon message, when Dir dropped to bacula:bacula
159 - Look at zlib 32 => 64 problems.
160 - Possibly turn on St. Bernard code.
161 - Fix bextract to restore ACLs, or better yet, use common routines.
162 - Do we migrate appendable Volumes?
163 - Remove queue.c code.
164 - Print warning message if LANG environment variable does not specify
166 - New dot commands from Arno.
167 .show device=xxx lists information from one storage device, including
168 devices (I'm not even sure that information exists in the DIR...)
169 .move eject device=xxx mostly the same as 'unmount xxx' but perhaps with
170 better machine-readable output like "Ok" or "Error busy"
171 .move eject device=xxx toslot=yyy the same as above, but with a new
172 target slot. The catalog should be updated accordingly.
173 .move transfer device=xxx fromslot=yyy toslot=zzz
176 - Article: http://www.heise.de/open/news/meldung/83231
177 - Article: http://www.golem.de/0701/49756.html
178 - Article: http://lwn.net/Articles/209809/
179 - Article: http://www.onlamp.com/pub/a/onlamp/2004/01/09/bacula.html
180 - Article: http://www.linuxdevcenter.com/pub/a/linux/2005/04/07/bacula.html
181 - Article: http://www.osreviews.net/reviews/admin/bacula
182 - Article: http://www.debianhelp.co.uk/baculaweb.htm
184 - It appears to me that you have run into some sort of race
185 condition where two threads want to use the same Volume and they
186 were both given access. Normally that is no problem. However,
187 one thread wanted the particular Volume in drive 0, but it was
188 loaded into drive 1 so it decided to unload it from drive 1 and
189 then loaded it into drive 0, while the second thread went on
190 thinking that the Volume could be used in drive 1 not realizing
191 that in between time, it was loaded in drive 0.
192 I'll look at the code to see if there is some way we can avoid
193 this kind of problem. Probably the best solution is to make the
194 first thread simply start using the Volume in drive 1 rather than
195 transferring it to drive 0.
196 - Fix re-read of last block to check if job has actually written
197 a block, and check if block was written by a different job
198 (i.e. multiple simultaneous jobs writing).
199 - Figure out how to configure query.sql. Suggestion to use m4:
200 == changequote.m4 ===
201 changequote(`[',`]')dnl
202 ==== query.sql.in ===
203 :List next 20 volumes to expire
205 Pool.Name AS PoolName,
210 [ FROM_UNIXTIME(UNIX_TIMESTAMP(Media.LastWritten) Media.VolRetention) AS Expire, ])dnl
212 [ media.lastwritten + interval '1 second' * media.volretention as expire, ])dnl
216 ON Media.PoolId=Pool.PoolId
217 WHERE Media.LastWritten>0
221 Command: m4 -DmySQL changequote.m4 query.sql.in >query.sql
223 The problem is that it requires m4, which is not present on all machines
225 - Given all the problems with FIFOs, I think the solution is to do something a
226 little different, though I will look at the code and see if there is not some
227 simple solution (i.e. some bug that was introduced). What might be a better
228 solution would be to use a FIFO as a sort of "key" to tell Bacula to read and
229 write data to a program rather than the FIFO. For example, suppose you
234 Then, I could imagine if you backup and restore this file with a direct
235 reference as is currently done for fifos, instead, during backup Bacula will
238 /home/kern/my-fifo.backup
240 and read the data that my-fifo.backup writes to stdout. For restore, Bacula
243 /home/kern/my-fifo.restore
245 and send the data backed up to stdout. These programs can either be an
246 executable or a shell script and they need only read/write to stdin/stdout.
248 I think this would give a lot of flexibility to the user without making any
249 significant changes to Bacula.
254 select FilenameId from Filename where Name='';
255 # Get list of all directories referenced in a Backup.
256 select Path.Path from Path,File where File.JobId=nnn and
257 File.FilenameId=(FilenameId-from-above) and File.PathId=Path.PathId
258 order by Path.Path ASC;
260 - Look into using Dart for testing
261 http://public.kitware.com/Dart/HTML/Index.shtml
263 - Look into replacing autotools with cmake
264 http://www.cmake.org/HTML/Index.html
266 === Migration from David ===
267 What I'd like to see:
270 Name = "<poolname>-migrate"
274 Migration Selection Type = LowestUtil | OldestVol | PoolOccupancy |
275 Client | PoolResidence | Volume | JobName | SQLquery
276 Migration Selection Pattern = "regexp"
277 Next Pool = <override>
280 There should be no need for a Level (migration is always Full, since you
281 don't calculate differential/incremental differences for migration),
282 Storage should be determined by the volume types in the pool, and Client
283 is really a selection issue. Migration should always occur to the
284 NextPool defined in the pool definition. If no nextpool is defined, the
285 job should end with a reason of "no place to go". If Next Pool statement
286 is present, we override the check in the pool definition and use the
289 Here's how I'd define Migration Selection Types:
292 Client -- Migrate data from selected client only. Migration Selection
293 Pattern regexp provides pattern to select client names, eg ^FS00* makes
294 all client names starting with FS00 eligible for migration.
296 Jobname -- Migration all jobs matching name. Migration Selection Pattern
297 regexp provides pattern to select jobnames existing in pool.
299 Volume -- Migrate all data on specified volumes. Migration Selection
300 Pattern regexp provides selection criteria for volumes to be migrated.
301 Volumes must exist in pool to be eligible for migration.
305 LowestUtil -- Identify the volume in the pool with the least data on it
306 and empty it. No Migration Selection Pattern required.
308 OldestVol -- Identify the LRU volume with data written, and empty it. No
309 Migration Selection Pattern required.
311 PoolOccupancy -- if pool occupancy exceeds <highmig>, migrate volumes
312 (starting with most full volumes) until pool occupancy drops below
313 <lowmig>. Pool highmig and lowmig values are in pool definition, no
314 Migration Selection Pattern required.
318 SQLQuery -- Migrate all jobuids returned by the supplied SQL query.
319 Migration Selection Pattern contains SQL query to execute; should return
320 a list of 1 or more jobuids to migrate.
322 PoolResidence -- Migrate data sitting in pool for longer than
323 PoolResidence value in pool definition. Migration Selection Pattern
324 optional; if specified, override value in pool definition (value in
328 [ possibly a Python event -- kes ]
330 - Mount on an Autochanger with no tape in the drive causes:
331 Automatically selected Storage: LTO-changer
332 Enter autochanger drive[0]: 0
333 3301 Issuing autochanger "loaded drive 0" command.
334 3302 Autochanger "loaded drive 0", result: nothing loaded.
335 3301 Issuing autochanger "loaded drive 0" command.
336 3302 Autochanger "loaded drive 0", result: nothing loaded.
337 3902 Cannot mount Volume on Storage Device "LTO-Drive1" (/dev/nst0) because:
338 Couldn't rewind device "LTO-Drive1" (/dev/nst0): ERR=dev.c:678 Rewind error on "LTO-Drive1" (/dev/nst0). ERR=No medium found.
339 3905 Device "LTO-Drive1" (/dev/nst0) open but no Bacula volume is mounted.
340 If this is not a blank tape, try unmounting and remounting the Volume.
341 - If Drive 0 is blocked, and drive 1 is set "Autoselect=no", drive 1 will
343 - Autochanger did not change volumes.
344 select * from Storage;
345 +-----------+-------------+-------------+
346 | StorageId | Name | AutoChanger |
347 +-----------+-------------+-------------+
348 | 1 | LTO-changer | 0 |
349 +-----------+-------------+-------------+
350 05-May 03:50 roxie-sd: 3302 Autochanger "loaded drive 0", result is Slot 11.
351 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Warning: Director wanted Volume "LT
352 Current Volume "LT0-002" not acceptable because:
353 1997 Volume "LT0-002" not in catalog.
354 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Error: Autochanger Volume "LT0-002"
355 Setting InChanger to zero in catalog.
356 05-May 03:50 roxie-dir: Tibs.2006-05-05_03.05.02 Error: Unable to get Media record
358 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: Error getting Volume i
359 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: Job 530 canceled.
360 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: spool.c:249 Fatal appe
361 05-May 03:49 Tibs: Tibs.2006-05-05_03.05.02 Fatal error: c:\cygwin\home\kern\bacula
370 FirstWritten: 2006-05-05 03:11:54
371 LastWritten: 2006-05-05 03:50:23
372 LabelDate: 2005-12-26 16:52:40
383 VolRetention: 31,536,000
395 Note VolStatus is blank!!!!!
402 FirstWritten: 0000-00-00 00:00:00
403 LastWritten: 0000-00-00 00:00:00
404 LabelDate: 2005-12-26 16:52:40
415 VolRetention: 31,536,000
428 Automatically selected Storage: LTO-changer
429 Enter autochanger drive[0]: 0
430 3301 Issuing autochanger "loaded drive 0" command.
431 3302 Autochanger "loaded drive 0", result: nothing loaded.
432 3301 Issuing autochanger "loaded drive 0" command.
433 3302 Autochanger "loaded drive 0", result: nothing loaded.
434 3902 Cannot mount Volume on Storage Device "LTO-Drive1" (/dev/nst0) because:
435 Couldn't rewind device "LTO-Drive1" (/dev/nst0): ERR=dev.c:678 Rewind error on "LTO-Drive1" (/dev/nst0). ERR=No medium found.
437 3905 Device "LTO-Drive1" (/dev/nst0) open but no Bacula volume is mounted.
438 If this is not a blank tape, try unmounting and remounting the Volume.
440 - http://www.dwheeler.com/essays/commercial-floss.html
441 - Add VolumeLock to prevent all but lock holder (SD) from updating
442 the Volume data (with the exception of VolumeState).
443 - The btape fill command does not seem to use the Autochanger
444 - Make Windows installer default to system disk drive.
445 - Look at using ioctl(FIOBMAP, ...) on Linux, and
446 DeviceIoControl(..., FSCTL_QUERY_ALLOCATED_RANGES, ...) on
447 Win32 for sparse files.
448 http://www.flexhex.com/docs/articles/sparse-files.phtml
449 http://www.informatik.uni-frankfurt.de/~loizides/reiserfs/fibmap.html
450 - Directive: at <event> "command"
451 - Command: pycmd "command" generates "command" event. How to
452 attach to a specific job?
453 - Integrate Christopher's St. Bernard code.
454 - run_cmd() returns int should return JobId_t
455 - get_next_jobid_from_list() returns int should return JobId_t
456 - Document export LDFLAGS=-L/usr/lib64
457 - Don't attempt to restore from "Disabled" Volumes.
458 - Network error on Win32 should set Win32 error code.
459 - What happens when you rename a Disk Volume?
460 - Job retention period in a Pool (and hence Volume). The job would
462 - Look at -D_FORTIFY_SOURCE=2
463 - Add Win32 FileSet definition somewhere
464 - Look at fixing restore status stats in SD.
465 - Look at using ioctl(FIMAP) and FIGETBSZ for sparse files.
466 http://www.informatik.uni-frankfurt.de/~loizides/reiserfs/fibmap.html
467 - Implement a mode that says when a hard read error is
468 encountered, read many times (as it currently does), and if the
469 block cannot be read, skip to the next block, and try again. If
470 that fails, skip to the next file and try again, ...
472 create table LevelType (LevelType binary(1), LevelTypeLong tinyblob);
473 insert into LevelType (LevelType,LevelTypeLong) values
477 - Show files/second in client status output.
478 - Add a recursive mark command (rmark) to restore.
479 - "Minimum Job Interval = nnn" sets minimum interval between Jobs
480 of the same level and does not permit multiple simultaneous
481 running of that Job (i.e. lets any previous invocation finish
482 before doing Interval testing).
483 - Look at simplifying File exclusions.
484 - New directive "Delete purged Volumes"
485 - new pool XXX with ScratchPoolId = MyScratchPool's PoolId and
486 let it fill itself, and RecyclePoolId = XXX's PoolId so I can
487 see if it become stable and I just have to supervise
489 - If I want to remove this pool, I set RecyclePoolId = MyScratchPool's
490 PoolId, and when it is empty remove it.
491 - Figure out how to recycle Scratch volumes back to the Scratch Pool.
493 - Allow Check Labels to be used with Bacula labels.
494 - "Resuming" a failed backup (lost line for example) by using the
495 failed backup as a sort of "base" job.
497 - Email to the user when the tape is about to need changing x
498 days before it needs changing.
499 - Command to show next tape that will be used for a job even
500 if the job is not scheduled.
501 - From: Arunav Mandal <amandal@trolltech.com>
502 1. When jobs are running and bacula for some reason crashes or if I do a
503 restart it remembers and jobs it was running before it crashed or restarted
504 as of now I loose all jobs if I restart it.
506 2. When spooling and in the midway if client is disconnected for instance a
507 laptop bacula completely discard the spool. It will be nice if it can write
508 that spool to tape so there will be some backups for that client if not all.
510 3. We have around 150 clients machines it will be nice to have a option to
511 upgrade all the client machines bacula version automatically.
513 4. Atleast one connection should be reserved for the bconsole so at heavy load
514 I should connect to the director via bconsole which at sometimes I can't
516 5. Another most important feature that is missing, say at 10am I manually
517 started backup of client abc and it was a full backup since client abc has
518 no backup history and at 10.30am bacula again automatically started backup of
519 client abc as that was in the schedule. So now we have 2 multiple Full
520 backups of the same client and if we again try to start a full backup of
521 client backup abc bacula won't complain. That should be fixed.
523 - Fix bpipe.c so that it does not modify results pointer.
524 ***FIXME*** calling sequence should be changed.
525 - For Windows disaster recovery see http://unattended.sf.net/
526 - regardless of the retention period, Bacula will not prune the
527 last Full, Diff, or Inc File data until a month after the
528 retention period for the last Full backup that was done.
529 - update volume=xxx --- add status=Full
530 - Remove old spool files on startup.
531 - Exclude SD spool/working directory.
532 - Refuse to prune last valid Full backup. Same goes for Catalog.
534 - Make a callback when Rerun failed levels is called.
535 - Give Python program access to Scheduled jobs.
536 - Add setting Volume State via Python.
537 - Python script to save with Python, not save, save with Bacula.
538 - Python script to do backup.
540 - Change the Priority, Client, Storage, JobStatus (error)
541 at the start of a job.
542 - Why is SpoolDirectory = /home/bacula/spool; not reported
543 as an error when writing a DVD?
544 - Make bootstrap file handle multiple MediaTypes (SD)
545 - Remove all old Device resource code in Dir and code to pass it
546 back in SD -- better, rework it to pass back device statistics.
547 - Check locking of resources -- be sure to lock devices where previously
548 resources were locked.
549 - The last part is left in the spool dir.
552 - In restore don't compare byte count on a raw device -- directory
553 entry does not contain bytes.
557 MA = (last_MA * 3 + rate) / 4
558 rate = (bytes - last_bytes) / (runtime - last_runtime)
559 - Max Vols limit in Pool off by one?
560 - Implement Files/Bytes,... stats for restore job.
561 - Implement Total Bytes Written, ... for restore job.
562 - Despool attributes simultaneously with data in a separate
563 thread, rejoined at end of data spooling.
564 - 7. Implement new Console commands to allow offlining/reserving drives,
565 and possibly manipulating the autochanger (much asked for).
566 - Add start/end date editing in messages (%t %T, %e?) ...
567 - Add ClientDefs similar to JobDefs.
568 - Print more info when bextract -p accepts a bad block.
569 - Fix FD JobType to be set before RunBeforeJob in FD.
570 - Look at adding full Volume and Pool information to a Volume
571 label so that bscan can get *all* the info.
572 - If the user puts "Purge Oldest Volume = yes" or "Recycle Oldest Volume = yes"
573 and there is only one volume in the pool, refuse to do it -- otherwise
574 he fills the Volume, then immediately starts reusing it.
575 - Implement copies and stripes.
576 - Add history file to console.
577 - Each file on tape creates a JobMedia record. Peter has 4 million
578 files spread over 10000 tape files and four tapes. A restore takes
579 16 hours to build the restore list.
580 - Add and option to check if the file size changed during backup.
581 - Make sure SD deletes spool files on error exit.
582 - Delete old spool files when SD starts.
583 - When labeling tapes, if you enter 000026, Bacula uses
584 the tape index rather than the Volume name 000026.
585 - Add offline tape command to Bacula console.
587 Enter MediaId or Volume name: 32
588 Enter new Volume name: DLT-20Dec04
589 Automatically selected Pool: Default
590 Connecting to Storage daemon DLTDrive at 192.168.68.104:9103 ...
591 Sending relabel command from "DLT-28Jun03" to "DLT-20Dec04" ...
592 block.c:552 Write error at 0:0 on device /dev/nst0. ERR=Bad file descriptor.
593 Error writing final EOF to tape. This tape may not be readable.
594 dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
595 askdir.c:219 NULL Volume name. This shouldn't happen!!!
596 3912 Failed to label Volume: ERR=dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
597 Label command failed for Volume DLT-20Dec04.
598 Do not forget to mount the drive!!!
599 - Bug: if a job is manually scheduled to run later, it does not appear
600 in any status report and cannot be cancelled.
602 ==== Keeping track of deleted/new files ====
603 - To mark files as deleted, run essentially a Verify to disk, and
604 when a file is found missing (MarkId != JobId), then create
605 a new File record with FileIndex == -1. This could be done
606 by the FD at the same time as the backup.
608 My "trick" for keeping track of deletions is the following.
609 Assuming the user turns on this option, after all the files
610 have been backed up, but before the job has terminated, the
611 FD will make a pass through all the files and send their
612 names to the DIR (*exactly* the same as what a Verify job
613 currently does). This will probably be done at the same
614 time the files are being sent to the SD avoiding a second
615 pass. The DIR will then compare that to what is stored in
616 the catalog. Any files in the catalog but not in what the
617 FD sent will receive a catalog File entry that indicates
618 that at that point in time the file was deleted. This
619 either transmitted to the FD or simultaneously computed in
620 the FD, so that the FD can put a record on the tape that
621 indicates that the file has been deleted at this point.
622 A delete file entry could potentially be one with a FileIndex
623 of 0 or perhaps -1 (need to check if FileIndex is used for
624 some other thing as many of the Bacula fields are "overloaded"
627 During a restore, any file initially picked up by some
628 backup (Full, ...) then subsequently having a File entry
629 marked "delete" will be removed from the tree, so will not
630 be restored. If a file with the same name is later OK it
631 will be inserted in the tree -- this already happens. All
632 will be consistent except for possible changes during the
635 Since I'm on the subject, some of you may be wondering what
636 the utility of the in memory tree is if you are going to
637 restore everything (at least it comes up from time to time
638 on the list). Well, it is still *very* useful because it
639 allows only the last item found for a particular filename
640 (full path) to be entered into the tree, and thus if a file
641 is backed up 10 times, only the last copy will be restored.
642 I recently (last Friday) restored a complete directory, and
643 the Full and all the Differential and Incremental backups
644 spanned 3 Volumes. The first Volume was not even mounted
645 because all the files had been updated and hence backed up
646 since the Full backup was made. In this case, the tree
647 saved me a *lot* of time.
649 Make sure this information is stored on the tape too so
650 that it can be restored directly from the tape.
652 All the code (with the exception of formally generating and
653 saving the delete file entries) already exists in the Verify
654 Catalog command. It explicitly recognizes added/deleted files since
655 the last InitCatalog. It is more or less a "simple" matter of
656 taking that code and adapting it slightly to work for backups.
658 Comments from Martin Simmons (I think they are all covered):
659 Ok, that should cover the basics. There are few issues though:
661 - Restore will depend on the catalog. I think it is better to include the
662 extra data in the backup as well, so it can be seen by bscan and bextract.
664 - I'm not sure if it will preserve multiple hard links to the same inode. Or
665 maybe adding or removing links will cause the data to be dumped again?
667 - I'm not sure if it will handle renamed directories. Possibly it will work
668 by dumping the whole tree under a renamed directory?
670 - It remains to be seen how the backup performance of the DIR's will be
671 affected when comparing the catalog for a large filesystem.
675 How about introducing a Type = MgmtPolicy job type? That job type would
676 be responsible for scanning the Bacula environment looking for specific
677 conditions, and submitting the appropriate jobs for implementing said
681 Name = "Migration-Policy"
683 Policy Selection Job Type = Migrate
684 Scope = "<keyword> <operator> <regexp>"
685 Threshold = "<keyword> <operator> <regexp>"
686 Job Template = <template-name>
689 Where <keyword> is any legal job keyword, <operator> is a comparison
690 operator (=,<,>,!=, logical operators AND/OR/NOT) and <regexp> is a
691 appropriate regexp. I could see an argument for Scope and Threshold
692 being SQL queries if we want to support full flexibility. The
693 Migration-Policy job would then get scheduled as frequently as a site
694 felt necessary (suggested default: every 15 minutes).
699 Name = "Migration-Policy"
701 Policy Selection Job Type = Migration
703 Threshold = "Migration Selection Type = LowestUtil"
704 Job Template = "MigrationTemplate"
707 would select all pools for examination and generate a job based on
708 MigrationTemplate to automatically select the volume with the lowest
709 usage and migrate it's contents to the nextpool defined for that pool.
711 This policy abstraction would be really handy for adjusting the behavior
712 of Bacula according to site-selectable criteria (one thing that pops
713 into mind is Amanda's ability to automatically adjust backup levels
714 depending on various criteria).
720 - Add Pool/Storage override regression test.
721 - Add delete JobId to regression.
722 - Add a regression test for dbcheck.
723 - New test to add bscan to four-concurrent-jobs regression,
724 i.e. after the four-concurrent jobs zap the
725 database as is done in the bscan-test, then use bscan to
726 restore the database, do a restore and compare with the
728 - Add restore of specific JobId to regression (item 3
729 on the restore prompt)
730 - Add IPv6 to regression
731 - Add database test to regression. Test each function like delete,
734 - AntiVir can slow down backups on Win32 systems.
735 - Win32 systems with FAT32 can be much slower than NTFS for
736 more than 1000 files per directory.
740 - A HOLD command to stop all jobs from starting.
741 - A PAUSE command to pause all running jobs ==> release the
743 - Media Type = LTO,LTO-2,LTO-3
744 Media Type Read = LTO,LTO2,LTO3
745 Media Type Write = LTO2, LTO3
747 === From Carsten Menke <bootsy52@gmx.net>
749 Following is a list of what I think in the situations where I'm faced with,
750 could be a usefull enhancement to bacula, which I'm certain other users will
751 benefit from as well.
753 1. NextJob/NextJobs Directive within a Job Resource in the form of
754 NextJobs = job1,job2.
757 I currently solved the problem with running multiple jobs each after each
758 by setting the Max Wait Time for a job to 8 hours, and give
759 the jobs different Priorities. However, there scenarios where
760 1 Job is directly depending on another job, so if the former job fails,
761 the job after it needn't to be run
762 while maybe other jobs should run despite of that
765 A Backup Job and a Verify job, if the backup job fails there is no need to run
766 the verify job, as the backup job already failed. However, one may like
767 to backup the Catalog to disk despite of that the main backup job failed.
770 I see that this is related to the Event Handlers which are on the ToDo
771 list, also it is maybe a good idea to check for the return value and
772 execute different actions based on the return value
775 3. offline capability to bconsole
778 Currently I use a script which I execute within the last Job via the
779 RunAfterJob Directive, to release and eject the tape.
780 So I have to call bconsole "release=Storage-Name" and afterwards
781 mt -f /dev/nst0 eject to get the tape out.
783 If I have multiple Storage Devices, than these may not be /dev/nst0 and
784 I have to modify the script or call it with parameters etc.
785 This would actually not be needed, as everything is already defined
786 in bacula-sd.conf and if I can invoke bconsole with the
787 storage name via $1 in the script than I'm done and information is
790 4. %s for Storage Name added to the chars being substituted in "RunAfterJob"
794 For the reason mentioned in 3. to have the ability to call a
795 script with /scripts/foobar %s and in the script use $1
796 to pass the Storage Name to bconsole
798 5. Setting Volume State within a Job Resource
801 Instead of using "Maximum Volume Jobs" in the Pool Resource,
802 I would have the possibilty to define
803 in a Job Resource that after this certain job is run, the Volume State
804 should be set to "Volume State = Used", this give more flexibility (IMHO).
806 6. Localization of Bacula Messages
809 Unfortunatley many,many people I work with don't speak english very well.
810 So if at least the Reporting messages would be localized then they
811 would understand that they have to change the tape,etc. etc.
813 I volunteer to do the german translations, and if I can convince my wife also
814 french and Morre (western african language).
816 7. OK, this is evil, probably bound to security risks and maybe not possible
817 due to the design of bacula.
819 Implementation of Backtics ( `command` ) for shell comand execution to
820 the "Label Format" Directive.
824 Currently I have defined BACULA_DAY_OF_WEEK="day1|day2..." resulting in
825 Label Format = "HolyBackup-${BACULA_DAY_OF_WEEK[${WeekDay}]}". If I could
826 use backticks than I could use "Label Format = HolyBackup-`date +%A` to have
827 the localized name for the day of the week appended to the
828 format string. Then I have the tape labeled automatically with weekday
829 name in the correct language.
831 - Yes, that is surely the case. I probably should turn those into Warning
832 errors. In addition, you just made me think that it might not be bad to
833 add an option to check the file size after backing up the file and
834 report if it changes. This would be done as an option because it would
837 Kern, good idea. If you do do that, mention in the output: file
838 shrunk, or file expanded, just to make it obvious to the user
839 (without having to the refer to file size), just how the file size
842 Would this option be for all file, or just one file? Or a fileset?
843 - Make output from status use html table tags for nicely
844 presenting in a browser.
845 - Can one write tapes faster with 8192 byte block sizes?
846 - Document security problems with the same password for everyone in
847 rpm and Win32 releases.
848 - Browse generations of files.
849 - I've seen an error when my catalog's File table fills up. I
850 then have to recreate the File table with a larger maximum row
851 size. Relevant information is at
852 http://dev.mysql.com/doc/mysql/en/Full_table.html ; I think the
853 "Installing and Configuring MySQL" chapter should talk a bit
854 about this potential problem, and recommend a solution.
855 - For Solaris must use POSIX awk.
856 - Want speed of writing to tape while despooling.
857 - Supported autochanger:
865 Wangtek 6525ES (SCSI-1 QIC drive, 525MB), under Linux 2.4.something,
866 bacula 1.36.0/1 works with blocksize 16k INSIDE bacula-sd.conf.
867 - Add regex from http://www.pcre.org to Bacula for Win32.
868 - Use only shell tools no make in CDROM package.
869 - Include within include does it work?
870 - Implement a Pool of type Cleaning?
871 - Implement VolReadTime and VolWriteTime in SD
872 - Modify Backing up Your Database to include a bootstrap file.
873 - Think about making certain database errors fatal.
874 - Look at correcting the time jump in the scheduler for daylight
875 savings time changes.
876 - Add a "real" timer to network connections.
877 - Promote to Full = Time period
878 - Check dates entered by user for correctness (month/day/... ranges)
879 - Compress restore Volume listing by date and first file.
880 - Look at patches/bacula_db.b2z postgresql that loops during restore.
882 - Perhaps add read/write programs and/or plugins to FileSets.
883 - How to handle backing up portables ...
884 - Add some sort of guaranteed Interval for upgrading jobs.
885 - Can we write the state file after every job terminates? On Win32
886 the system crashes and the state file is not updated.
889 Documentation to do: (any release a little bit at a time)
890 - Doc to do unmount before removing magazine.
891 - Alternative to static linking "ldd prog" save all binaries listed,
892 restore them and point LD_LIBRARY_PATH to them.
893 - Document add "</dev/null >/dev/null 2>&1" to the bacula-fd command line
894 - Document query file format.
895 - Add more documentation for bsr files.
896 - Document problems with Verify and pruning.
897 - Document how to use multiple databases.
898 - VXA drives have a "cleaning required"
899 indicator, but Exabyte recommends preventive cleaning after every 75
902 In this context, it should be noted that Exabyte has a command-line
903 vxatool utility available for free download. (The current version is
904 vxatool-3.72.) It can get diagnostic info, read, write and erase tapes,
905 test the drive, unload tapes, change drive settings, flash new firmware,
907 Of particular interest in this context is that vxatool <device> -i will
908 report, among other details, the time since last cleaning in tape motion
909 minutes. This information can be retrieved (and settings changed, for
910 that matter) through the generic-SCSI device even when Bacula has the
911 regular tape device locked. (Needless to say, I don't recommend
912 changing tape settings while a job is running.)
913 - Lookup HP cleaning recommendations.
914 - Lookup HP tape replacement recommendations (see trouble shooting autochanger)
915 - Document doing table repair
918 ===================================
919 - Add macro expansions in JobDefs.
920 Run Before Job = "SomeFile %{Level} %{Client}"
921 Write Bootstrap="/some/dir/%{JobName}_%{Client}.bsr"
922 - Use non-blocking network I/O but if no data is available, use
924 - Use gather write() for network I/O.
925 - Autorestart on crash.
926 - Add bandwidth limiting.
927 - Add acks every once and a while from the SD to keep
928 the line from timing out.
929 - When an error in input occurs and conio beeps, you can back
930 up through the prompt.
931 - Detect fixed tape block mode during positioning by looking at
932 block numbers in btape "test". Possibly adjust in Bacula.
933 - Fix list volumes to output volume retention in some other
934 units, perhaps via a directive.
935 - Allow Simultaneous Priorities = yes => run up to Max concurrent jobs even
936 with multiple priorities.
937 - If you use restore replace=never, the directory attributes for
938 non-existent directories will not be restored properly.
940 - see lzma401.zip in others directory for new compression
942 - Allow the user to select JobType for manual pruning/purging.
943 - bscan does not put first of two volumes back with all info in
945 - Implement the FreeBSD nodump flag in chflags.
946 - Figure out how to make named console messages go only to that
947 console and to the non-restricted console (new console class?).
948 - Make restricted console prompt for password if *ask* is set or
949 perhaps if password is undefined.
950 - Implement "from ISO-date/time every x hours/days/weeks/months" in
953 ==== from Marc Schoechlin
954 - the help-command should be more verbose
955 (it should explain the paramters of the different
957 -> its time-comsuming to consult the manual anytime
958 you need a special parameter
959 -> maybe its more easy to maintain this, if the
960 descriptions of that commands are outsourced to
962 - the cd-command should allow complete paths
963 i.e. cd /foo/bar/foo/bar
964 -> if a customer mails me the path to a certain file,
965 its faster to enter the specified directory
966 - if the password is not configured in bconsole.conf
967 you should be asked for it.
968 -> sometimes you like to do restore on a customer-machine
969 which shouldnt know the password for bacula.
970 -> adding the password to the file favours admins
971 to forget to remove the password after usage
973 the protection of that file is less important
974 - long-listed-output of commands should be scrollable
975 like the unix more/less-command does
976 -> if someone runs 200 and more machines, the lists could
977 be a little long and complex
978 - command-output should be shown column by column
979 to reduce scrolling and to increase clarity
981 - lsmark should list the selected files with full
983 - wildcards for selecting and file and directories would be nice
984 - any actions should be interuptable with STRG+C
985 - command-expansion would be pretty cool
987 - When the replace Never option is set, new directory permissions
988 are not restored. See bug 213. To fix this requires creating a
989 list of newly restored directories so that those directory
990 permissions *can* be restored.
991 - Add prune all command
992 - Document fact that purge can destroy a part of a restore by purging
993 one volume while others remain valid -- perhaps mark Jobs.
994 - Add multiple-media-types.txt
995 - look at mxt-changer.html
996 - Make ? do a help command (no return needed).
997 - Implement restore directory.
998 - Document streams and how to implement them.
999 - Try not to re-backup a file if a new hard link is added.
1000 - Add feature to backup hard links only, but not the data.
1001 - Fix stream handling to be simpler.
1002 - Add Priority and Bootstrap to Run a Job.
1003 - Eliminate Restore "Run Restore Job" prompt by allowing new "run command
1005 - Remove View FileSet button from Run a Job dialog.
1006 - Handle prompt for restore job at end of Restore command.
1007 - Add display of total selected files to Restore window.
1008 - Add tree pane to left of window.
1009 - Add progress meter.
1010 - Max wait time or max run time causes seg fault -- see runtime-bug.txt
1011 - Add message to user to check for fixed block size when the forward
1012 space test fails in btape.
1013 - When unmarking a directory check if all files below are unmarked and
1014 then remove the + flag -- in the restore tree.
1015 - Possibly implement: Action = Unmount Device="TapeDrive1" in Admin jobs.
1016 - Setup lrrd graphs: (http://www.linpro.no/projects/lrrd/) Mike Acar.
1017 - Revisit the question of multiple Volumes (disk) on a single device.
1018 - Add a block copy option to bcopy.
1019 - Finish work on Gnome restore GUI.
1020 - Fix "llist jobid=xx" where no fileset or client exists.
1021 - For each job type (Admin, Restore, ...) require only the really necessary
1022 fields.- Pass Director resource name as an option to the Console.
1023 - Add a "batch" mode to the Console (no unsolicited queries, ...).
1024 - Add a .list all files in the restore tree (probably also a list all files)
1025 Do both a long and short form.
1026 - Allow browsing the catalog to see all versions of a file (with
1027 stat data on each file).
1028 - Restore attributes of directory if replace=never set but directory
1030 - Use SHA1 on authentication if possible.
1031 - See comtest-xxx.zip for Windows code to talk to USB.
1032 - Add John's appended files:
1033 Appended = { /files/server/logs/http/*log }
1034 and such files would be treated as follows.On a FULL backup, they would
1035 be backed up like any other file.On an INCREMENTAL backup, where a
1036 previous INCREMENTAL or FULL was already in thecatalogue and the length
1037 of the file wasgreater than the length of the last backup, only thedata
1038 added since the last backup will be dumped.On an INCREMENTAL backup, if
1039 the length of the file is less than thelength of the file with the same
1040 name last backed up, the completefile is dumped.On Windows systems, with
1041 creation date of files, we can be evensmarter about this and not count
1042 entirely upon the length.On a restore, the full and all incrementals
1043 since it will beapplied in sequence to restore the file.
1044 - Check new HAVE_WIN32 open bits.
1045 - Check if the tape has moved before writing.
1046 - Handling removable disks -- see below:
1047 - Keep track of tape use time, and report when cleaning is necessary.
1048 - Add FromClient and ToClient keywords on restore command (or
1049 BackupClient RestoreClient).
1050 - Implement a JobSet, which groups any number of jobs. If the
1051 JobSet is started, all the jobs are started together.
1052 Allow Pool, Level, and Schedule overrides.
1053 - Enhance cancel to timeout BSOCK packets after a specific delay.
1054 - Do scheduling by UTC using gmtime_r() in run_conf, scheduler, and
1055 ua_status.!!! Thanks to Alan Brown for this tip.
1056 - Look at updating Volume Jobs so that Max Volume Jobs = 1 will work
1057 correctly for multiple simultaneous jobs.
1058 - Correct code so that FileSet MD5 is calculated for < and | filename
1060 - Implement the Media record flag that indicates that the Volume does disk
1062 - Implement VolAddr, which is used when Volume is addressed like a disk,
1063 and form it from VolFile and VolBlock.
1064 - Make multiple restore jobs for multiple media types specifying
1065 the proper storage type.
1066 - Fix fast block rejection (stored/read_record.c:118). It passes a null
1067 pointer (rec) to try_repositioning().
1068 - Look at extracting Win data from BackupRead.
1069 - Implement RestoreJobRetention? Maybe better "JobRetention" in a Job,
1070 which would take precidence over the Catalog "JobRetention".
1071 - Implement Label Format in Add and Label console commands.
1072 - Possibly up network buffers to 65K. Put on variable.
1073 - Put email tape request delays on one or more variables. User wants
1074 to cancel the job after a certain time interval. Maximum Mount Wait?
1075 - Job, Client, Device, Pool, or Volume?
1076 Is it possible to make this a directive which is *optional* in multiple
1077 resources, like Level? If so, I think I'd make it an optional directive
1078 in Job, Client, and Pool, with precedence such that Job overrides Client
1079 which in turn overrides Pool.
1081 - New Storage specifications:
1082 - Want to write to multiple storage devices simultaneously
1083 - Want to write to multiple storage devices sequentially (in one job)
1084 - Want to read/write simultaneously
1085 - Key is MediaType -- it must match
1087 Passed to SD as a sort of BSR record called Storage Specification
1091 MediaType -> Next MediaType
1093 Device -> Next Device
1095 Allow multiple Storage specifications
1103 Allow Multiple Pool specifications (note, Pool currently
1105 Allow Multiple MediaType specifications in Dir conf
1106 Allow Multiple Device specifications in Dir conf
1107 Perhaps keep this in a single SSR
1108 Tie a Volume to a specific device by using a MediaType that
1109 is contained in only one device.
1110 In SD allow Device to have Multiple MediaTypes
1112 - Ideas from Jerry Scharf:
1113 First let's point out some big pluses that bacula has for this
1115 more importantly it's active. Thank you so much for that
1116 even more important, it's not flaky
1117 it has an open access catalog, opening many possibilities
1118 it's pushing toward heterogeneous systems capability
1120 Macintosh file client
1121 macs are an interesting niche, but I fear a server is a rathole
1122 working bare iron recovery for windows
1123 the option for inc/diff backups not reset on fileset revision
1124 a) use both change and inode update time against base time
1125 b) do the full catalog check (expensive but accurate)
1126 sizing guide (how much system is needed to back up N systems/files)
1127 consultants on using bacula in building a disaster recovery system
1128 an integration guide
1129 or how to get at fancy things that one could do with bacula
1130 logwatch code for bacula logs (or similar)
1131 linux distro inclusion of bacula (brings good and bad, but necessary)
1132 win2k/XP server capability (icky but you asked)
1133 support for Oracle database ??
1135 - Look at adding SQL server and Exchange support for Windows.
1136 - Make dev->file and dev->block_num signed integers so that -1 can
1137 be an invalid value which happens with BSR.
1138 - Create VolAddr for disk files in place of VolFile and VolBlock. This
1139 is needed to properly specify ranges.
1140 - Add progress of files/bytes to SD and FD.
1141 - Print warning message if FileId > 4 billion
1142 - do a "messages" before the first prompt in Console
1143 - Client does not show busy during Estimate command.
1144 - Implement Console mtx commands.
1145 - Implement a Mount Command and an Unmount Command where
1146 the users could specify a system command to be performed
1147 to do the mount, after which Bacula could attempt to
1148 read the device. This is for Removeable media such as a CDROM.
1149 - Most likely, this mount command would be invoked explicitly
1150 by the user using the current Console "mount" and "unmount"
1151 commands -- the Storage Daemon would do the right thing
1152 depending on the exact nature of the device.
1153 - As with tape drives, when Bacula wanted a new removable
1154 disk mounted, it would unmount the old one, and send a message
1155 to the user, who would then use "mount" as described above
1156 once he had actually inserted the disk.
1157 - Implement dump/print label to UA
1158 - Spool to disk only when the tape is full, then when a tape is hung move
1160 - bextract is sending everything to the log file ****FIXME****
1161 - Allow multiple Storage specifications (or multiple names on
1162 a single Storage specification) in the Job record. Thus a job
1163 can be backed up to a number of storage devices.
1164 - Implement some way for the File daemon to contact the Director
1165 to start a job or pass its DHCP obtained IP number.
1166 - Implement a query tape prompt/replace feature for a console
1167 - Copy console @ code to gnome2-console
1168 - Make tree walk routines like cd, ls, ... more user friendly
1169 by handling spaces better.
1170 - Make sure that Bacula rechecks the tape after the 20 min wait.
1171 - Set IO_NOWAIT on Bacula TCP/IP packets.
1172 - Try doing a raw partition backup and restore by mounting a
1174 - From Lars Kellers:
1175 Yes, it would allow to highly automatic the request for new tapes. If a
1176 tape is empty, bacula reads the barcodes (native or simulated), and if
1177 an unused tape is found, it runs the label command with all the
1178 necessary parameters.
1180 By the way can bacula automatically "move" an empty/purged volume say
1181 in the "short" pool to the "long" pool if this pool runs out of volume
1183 - What to do about "list files job=xxx".
1184 - Look at how fuser works and /proc/PID/fd that is how Nic found the
1185 file descriptor leak in Bacula.
1186 - Implement WrapCounters in Counters.
1187 - Add heartbeat from FD to SD if hb interval expires.
1188 - Can we dynamically change FileSets?
1189 - If pool specified to label command and Label Format is specified,
1190 automatically generate the Volume name.
1191 - Why can't SQL do the filename sort for restore?
1192 - Add ExhautiveRestoreSearch
1193 - Look at the possibility of loading only the necessary
1194 data into the restore tree (i.e. do it one directory at a
1195 time as the user walks through the tree).
1196 - Possibly use the hash code if the user selects all for a restore command.
1197 - Fix "restore all" to bypass building the tree.
1198 - Prohibit backing up archive device (findlib/find_one.c:128)
1199 - Implement Release Device in the Job resource to unmount a drive.
1200 - Implement Acquire Device in the Job resource to mount a drive,
1201 be sure this works with admin jobs so that the user can get
1202 prompted to insert the correct tape. Possibly some way to say to
1203 run the job but don't save the files.
1204 - Make things like list where a file is saved case independent for
1206 - Use autochanger to handle multiple devices.
1207 - Implement a Recycle command
1208 - Start working on Base jobs.
1209 - Implement UnsavedFiles DB record.
1210 - From Phil Stracchino:
1211 It would probably be a per-client option, and would be called
1212 something like, say, "Automatically purge obsoleted jobs". What it
1213 would do is, when you successfully complete a Differential backup of a
1214 client, it would automatically purge all Incremental backups for that
1215 client that are rendered redundant by that Differential. Likewise,
1216 when a Full backup on a client completed, it would automatically purge
1217 all Differential and Incremental jobs obsoleted by that Full backup.
1218 This would let people minimize the number of tapes they're keeping on
1219 hand without having to master the art of retention times.
1220 - When doing a Backup send all attributes back to the Director, who
1221 would then figure out what files have been deleted.
1222 - Currently in mount.c:236 the SD simply creates a Volume. It should have
1223 explicit permission to do so. It should also mark the tape in error
1224 if there is an error.
1225 - Cancel waiting for Client connect in SD if FD goes away.
1227 - Implement timeout in response() when it should come quickly.
1228 - Implement a Slot priority (loaded/not loaded).
1229 - Implement "vacation" Incremental only saves.
1230 - Implement create "FileSet"?
1231 - Add prefixlinks to where or not where absolute links to FD.
1232 - Issue message to mount a new tape before the rewind.
1233 - Simplified client job initiation for portables.
1234 - If SD cannot open a drive, make it periodically retry.
1235 - Add more of the config info to the tape label.
1237 - Refine SD waiting output:
1238 Device is being positioned
1239 > Device is being positioned for append
1240 > Device is being positioned to file x
1242 - Figure out some way to estimate output size and to avoid splitting
1243 a backup across two Volumes -- this could be useful for writing CDROMs
1244 where you really prefer not to have it split -- not serious.
1245 - Have SD compute MD5 or SHA1 and compare to what FD computes.
1246 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
1247 actual data on the Volume and compare it.
1248 - Implement Bacula plugins -- design API
1249 - Make bcopy read through bad tape records.
1250 - Program files (i.e. execute a program to read/write files).
1251 Pass read date of last backup, size of file last time.
1252 - Add Signature type to File DB record.
1253 - CD into subdirectory when open()ing files for backup to
1254 speed up things. Test with testfind().
1255 - Priority job to go to top of list.
1256 - Why are save/restore of device different sizes (sparse?) Yup! Fix it.
1257 - Implement some way for the Console to dynamically create a job.
1258 - Solaris -I on tar for include list
1259 - Need a verbose mode in restore, perhaps to bsr.
1260 - bscan without -v is too quiet -- perhaps show jobs.
1261 - Add code to reject whole blocks if not wanted on restore.
1262 - Check if we can increase Bacula FD priorty in Win2000
1263 - Make sure the MaxVolFiles is fully implemented in SD
1264 - Check if both CatalogFiles and UseCatalog are set to SD.
1265 - Possibly add email to Watchdog if drive is unmounted too
1266 long and a job is waiting on the drive.
1267 - After unmount, if restore job started, ask to mount.
1268 - Add UA rc and history files.
1269 - put termcap (used by console) in ./configure and
1270 allow -with-termcap-dir.
1271 - Fix Autoprune for Volumes to respect need for full save.
1272 - Compare tape to Client files (attributes, or attributes and data)
1273 - Make all database Ids 64 bit.
1274 - Allow console commands to detach or run in background.
1275 - Add SD message variables to control operator wait time
1276 - Maximum Operator Wait
1277 - Minimum Message Interval
1278 - Maximum Message Interval
1279 - Send Operator message when cannot read tape label.
1280 - Verify level=Volume (scan only), level=Data (compare of data to file).
1281 Verify level=Catalog, level=InitCatalog
1283 - Add keyword search to show command in Console.
1284 - Events : tape has more than xxx bytes.
1285 - Complete code in Bacula Resources -- this will permit
1286 reading a new config file at any time.
1287 - Handle ctl-c in Console
1288 - Implement script driven addition of File daemon to config files.
1289 - Think about how to make Bacula work better with File (non-tape) archives.
1290 - Write Unix emulator for Windows.
1291 - Put memory utilization in Status output of each daemon
1292 if full status requested or if some level of debug on.
1293 - Make database type selectable by .conf files i.e. at runtime
1294 - Set flag for uname -a. Add to Volume label.
1295 - Restore files modified after date
1296 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
1297 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
1298 - Timout a job or terminate if link goes down, or reopen link and query.
1299 - Concept of precious tapes (cannot be reused).
1300 - Make bcopy copy with a single tape drive.
1301 - Permit changing ownership during restore.
1304 > My suggestion: Add a feature on the systray menu-icon menu to request
1305 > an immediate backup now. This would be useful for laptop users who may
1306 > not be on the network when the regular scheduled backup is run.
1308 > My wife's suggestion: Add a setting to the win32 client to allow it to
1309 > shut down the machine after backup is complete (after, of course,
1310 > displaying a "System will shut down in one minute, click here to cancel"
1311 > warning dialog). This would be useful for sites that want user
1312 > woorkstations to be shut down overnight to save power.
1315 - Autolabel should be specified by DIR instead of SD.
1317 - Add media capacity
1318 - AutoScan (check checksum of tape)
1319 - Format command = "format /dev/nst0"
1323 - Seek resolution (usually corresponds to buffer size)
1324 - EODErrorCode=ENOSPC or code
1325 - Partial Read error code
1326 - Partial write error code
1327 - Nonformatted read error
1328 - Nonformatted write error
1329 - WriteProtected error
1333 - IgnoreCloseErrors=yes
1343 - FD sends unsaved file list to Director at end of job (see
1345 - File daemon should build list of files skipped, and then
1346 at end of save retry and report any errors.
1347 - Write a Storage daemon that uses pipes and
1348 standard Unix programs to write to the tape.
1350 - Need something that monitors the JCR queue and
1351 times out jobs by asking the deamons where they are.
1352 - Enhance Jmsg code to permit buffering and saving to disk.
1353 - device driver = "xxxx" for drives.
1354 - Verify from Volume
1355 - Ensure that /dev/null works
1356 - Need report class for messages. Perhaps
1357 report resource where report=group of messages
1358 - enhance scan_attrib and rename scan_jobtype, and
1359 fill in code for "since" option
1360 - Director needs a time after which the report status is sent
1361 anyway -- or better yet, a retry time for the job.
1362 - Don't reschedule a job if previous incarnation is still running.
1363 - Some way to automatically backup everything is needed????
1364 - Need a structure for pending actions:
1366 - termination status (part of buffered msgs?)
1368 Read, Write, Clean, Delete
1369 - Login to Bacula; Bacula users with different permissions:
1370 owner, group, user, quotas
1371 - Store info on each file system type (probably in the job header on tape.
1372 This could be the output of df; or perhaps some sort of /etc/mtab record.
1374 ========= ideas ===============
1375 From: "Jerry K. Schieffer" <jerry@skylinetechnology.com>
1376 To: <kern@sibbald.com>
1377 Subject: RE: [Bacula-users] future large programming jobs
1378 Date: Thu, 26 Feb 2004 11:34:54 -0600
1380 I noticed the subject thread and thought I would offer the following
1381 merely as sources of ideas, i.e. something to think about, not even as
1382 strong as a request. In my former life (before retiring) I often
1383 dealt with backups and storage management issues/products as a
1384 developer and as a consultant. I am currently migrating my personal
1385 network from amanda to bacula specifically because of the ability to
1386 cross media boundaries during storing backups.
1387 Are you familiar with the commercial product called ADSM (I think IBM
1388 now sells it under the Tivoli label)? It has a couple of interesting
1389 ideas that may apply to the following topics.
1391 1. Migration: Consider that when you need to restore a system, there
1392 may be pressure to hurry. If all the information for a single client
1393 can eventually end up on the same media (and in chronological order),
1394 the restore is facillitated by not having to search past information
1395 from other clients. ADSM has the concept of "client affinity" that
1396 may be associated with it's storage pools. It seems to me that this
1397 concept (as an optional feature) might fit in your architecture for
1400 ADSM also has the concept of defining one or more storage pools as
1401 "copy pools" (almost mirrors, but only in the sense of contents).
1402 These pools provide the ability to have duplicte data stored both
1403 onsite and offsite. The copy process can be scheduled to be handled
1404 by their storage manager during periods when there is no backup
1405 activity. Again, the migration process might be a place to consider
1406 implementing something like this.
1409 > It strikes me that it would be very nice to be able to do things
1411 > have the Job(s) backing up the machines run, and once they have all
1412 > completed, start a migration job to copy the data from disks Volumes
1414 > a tape library and then to offsite storage. Maybe this can already
1416 > done with some careful scheduling and Job prioritzation; the events
1417 > mechanism described below would probably make it very easy.
1419 This is the goal. In the first step (before events), you simply
1421 the Migration to tape later.
1423 2. Base jobs: In ADSM, each copy of each stored file is tracked in
1424 the database. Once a file (unique by path and metadata such as dates,
1425 size, ownership, etc.) is in a copy pool, no more copies are made. In
1426 other words, when you start ADSM, it begins like your concept of a
1427 base job. After that it is in the "incremental" mode. You can
1428 configure the number of "generations" of files to be retained, plus a
1429 retention date after which even old generations are purged. The
1430 database tracks the contents of media and projects the percentage of
1431 each volume that is valid. When the valid content of a volume drops
1432 below a configured percentage, the valid data are migrated to another
1433 volume and the old volume is marked as empty. Note, this requires
1434 ADSM to have an idea of the contents of a client, i.e. marking the
1435 database when an existing file was deleted, but this would solve your
1436 issue of restoring a client without restoring deleted files.
1438 This is pretty far from what bacula now does, but if you are going to
1439 rip things up for Base jobs,.....
1440 Also, the benefits of this are huge for very large shops, especially
1441 with media robots, but are a pain for shops with manual media
1445 > Base jobs sound pretty useful, but I'm not dying for them.
1447 Nobody is dying for them, but when you see what it does, you will die
1450 3. Restoring deleted files: Since I think my comments in (2) above
1451 have low probability of implementation, I'll also suggest that you
1452 could approach the issue of deleted files by a mechanism of having the
1453 fd report to the dir, a list of all files on the client for every
1454 backup job. The dir could note in the database entry for each file
1455 the date that the file was seen. Then if a restore as of date X takes
1456 place, only files that exist from before X until after X would be
1457 restored. Probably the major cost here is the extra date container in
1458 each row of the files table.
1460 Thanks for "listening". I hope some of this helps. If you want to
1461 contact me, please send me an email - I read some but not all of the
1462 mailing list traffic and might miss a reply there.
1464 Please accept my compliments for bacula. It is doing a great job for
1465 me!! I sympathize with you in the need to wrestle with excelence in
1466 execution vs. excelence in feature inclusion.
1471 ==============================
1474 - Design at hierarchial storage for Bacula. Migration and Clone.
1475 - Implement FSM (File System Modules).
1476 - Audit M_ error codes to ensure they are correct and consistent.
1477 - Add variable break characters to lex analyzer.
1478 Either a bit mask or a string of chars so that
1479 the caller can change the break characters.
1480 - Make a single T_BREAK to replace T_COMMA, etc.
1481 - Ensure that File daemon and Storage daemon can
1482 continue a save if the Director goes down (this
1483 is NOT currently the case). Must detect socket error,
1484 buffer messages for later.
1485 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
1486 - Add ability to backup to two Storage devices (two SD sessions) at
1487 the same time -- e.g. onsite, offsite.
1488 - Add the ability to consolidate old backup sets (basically do a restore
1489 to tape and appropriately update the catalog). Compress Volume sets.
1490 Might need to spool via file is only one drive is available.
1491 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
1492 someway to do so with every volume that has less than x% valid
1496 Migration: Move a backup from one Volume to another
1497 Clone: Copy a backup -- two Volumes
1500 ======================================================
1502 It is somewhat like a Full save becomes an incremental since
1503 the Base job (or jobs) plus other non-base files.
1505 - A Base backup is same as Full backup, just different type.
1506 - New BaseFiles table that contains:
1508 BaseJobId - Base JobId referenced for this FileId (needed ???)
1509 JobId - JobId currently running
1510 FileId - File not backed up, exists in Base Job
1511 FileIndex - FileIndex from Base Job.
1512 i.e. for each base file that exists but is not saved because
1513 it has not changed, the File daemon sends the JobId, BaseId,
1514 FileId, FileIndex back to the Director who creates the DB entry.
1515 - To initiate a Base save, the Director sends the FD
1516 the FileId, and full filename for each file in the Base.
1517 - When the FD finds a Base file, he requests the Director to
1518 send him the full File entry (stat packet plus MD5), or
1519 conversely, the FD sends it to the Director and the Director
1520 says yes or no. This can be quite rapid if the FileId is kept
1521 by the FD for each Base Filename.
1522 - It is probably better to have the comparison done by the FD
1523 despite the fact that the File entry must be sent across the
1525 - An alternative would be to send the FD the whole File entry
1526 from the start. The disadvantage is that it requires a lot of
1527 space. The advantage is that it requires less communications
1529 - The Job record must be updated to indicate that one or more
1531 - At end of Job, FD returns:
1532 1. Count of base files/bytes not written to tape (i.e. matches)
1533 2. Count of base file that were saved i.e. had changed.
1534 - No tape record would be written for a Base file that matches, in the
1535 same way that no tape record is written for Incremental jobs where
1536 the file is not saved because it is unchanged.
1537 - On a restore, all the Base file records must explicitly be
1538 found from the BaseFile tape. I.e. for each Full save that is marked
1539 to have one or more Base Jobs, search the BaseFile for all occurrences
1541 - An optimization might be to make the BaseFile have:
1547 This would avoid the need to explicitly fetch each File record for
1548 the Base job. The Base Job record will be fetched to get the
1549 VolSessionId and VolSessionTime.
1550 =========================================================
1553 ==========================================================
1555 For each Incremental job that is run, there may be files that
1556 were found but not saved because they were locked (this applies
1557 only to Windows). Such a system could send back to the Director
1558 a list of Unsaved files.
1560 - New UnSavedFiles table that contains:
1564 - Then in the next Incremental job, the list of Unsaved Files will be
1565 feed to the FD, who will ensure that they are explicitly chosen even
1566 if standard date/time check would not have selected them.
1567 =============================================================
1571 Multiple drive autochanger data: see Alan Brown
1572 mtx -f xxx unloadStorage Element 1 is Already Full(drive 0 was empty)
1573 Unloading Data Transfer Element into Storage Element 1...source Element
1574 Address 480 is Empty
1576 (drive 0 was empty and so was slot 1)
1577 > mtx -f xxx load 15 0
1578 no response, just returns to the command prompt when complete.
1579 > mtx -f xxx status Storage Changer /dev/changer:2 Drives, 60 Slots ( 2 Import/Export )
1580 Data Transfer Element 0:Full (Storage Element 15 Loaded):VolumeTag = HX001
1581 Data Transfer Element 1:Empty
1582 Storage Element 1:Empty
1583 Storage Element 2:Full :VolumeTag=HX002
1584 Storage Element 3:Full :VolumeTag=HX003
1585 Storage Element 4:Full :VolumeTag=HX004
1586 Storage Element 5:Full :VolumeTag=HX005
1587 Storage Element 6:Full :VolumeTag=HX006
1588 Storage Element 7:Full :VolumeTag=HX007
1589 Storage Element 8:Full :VolumeTag=HX008
1590 Storage Element 9:Full :VolumeTag=HX009
1591 Storage Element 10:Full :VolumeTag=HX010
1592 Storage Element 11:Empty
1593 Storage Element 12:Empty
1594 Storage Element 13:Empty
1595 Storage Element 14:Empty
1596 Storage Element 15:Empty
1597 Storage Element 16:Empty....
1598 Storage Element 28:Empty
1599 Storage Element 29:Full :VolumeTag=CLNU01L1
1600 Storage Element 30:Empty....
1601 Storage Element 57:Empty
1602 Storage Element 58:Full :VolumeTag=NEX261L2
1603 Storage Element 59 IMPORT/EXPORT:Empty
1604 Storage Element 60 IMPORT/EXPORT:Empty
1606 Unloading Data Transfer Element into Storage Element 15...done
1608 (just to verify it remembers where it came from, however it can be
1609 overrriden with mtx unload {slotnumber} to go to any storage slot.)
1611 There needs to be a table of drive # to devices somewhere - If there are
1612 multiple changers or drives there may not be a 1:1 correspondance between
1613 changer drive number and system device name - and depending on the way the
1614 drives are hooked up to scsi busses, they may not be linearly numbered
1615 from an offset point either.something like
1617 Autochanger drives = 2
1618 Autochanger drive 0 = /dev/nst1
1619 Autochanger drive 1 = /dev/nst2
1620 IMHO, it would be _safest_ to use explicit mtx unload commands at all
1621 times, not just for multidrive changers. For a 1 drive changer, that's
1627 MTX's manpage (1.2.15):
1628 unload [<slotnum>] [ <drivenum> ]
1629 Unloads media from drive <drivenum> into slot
1630 <slotnum>. If <drivenum> is omitted, defaults to
1631 drive 0 (as do all commands). If <slotnum> is
1632 omitted, defaults to the slot that the drive was
1633 loaded from. Note that there's currently no way
1634 to say 'unload drive 1's media to the slot it
1635 came from', other than to explicitly use that
1636 slot number as the destination.AB
1642 undef# camcontrol devlist
1643 <WANGTEK 51000 SCSI M74H 12B3> at scbus0 target 2 lun 0 (pass0,sa0)
1644 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 0 (pass1,sa1)
1645 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 1 (pass2)
1647 tapeinfo -f /dev/sg0 with a bad tape in drive 1:
1648 [kern@rufus mtx-1.2.17kes]$ ./tapeinfo -f /dev/sg0
1649 Product Type: Tape Drive
1651 Product ID: 'C5713A '
1653 Attached Changer: No
1654 TapeAlert[3]: Hard Error: Uncorrectable read/write error.
1655 TapeAlert[20]: Clean Now: The tape drive neads cleaning NOW.
1662 Medium Type: Not Loaded
1665 DataCompEnabled: yes
1666 DataCompCapable: yes
1667 DataDeCompEnabled: yes
1674 Handling removable disks
1676 From: Karl Cunningham <karlc@keckec.com>
1678 My backups are only to hard disk these days, in removable bays. This is my
1679 idea of how a backup to hard disk would work more smoothly. Some of these
1680 things Bacula does already, but I mention them for completeness. If others
1681 have better ways to do this, I'd like to hear about it.
1683 1. Accommodate several disks, rotated similar to how tapes are. Identified
1684 by partition volume ID or perhaps by the name of a subdirectory.
1685 2. Abort & notify the admin if the wrong disk is in the bay.
1686 3. Write backups to different subdirectories for each machine to be backed
1688 4. Volumes (files) get created as needed in the proper subdirectory, one
1690 5. When a disk is recycled, remove or zero all old backup files. This is
1691 important as the disk being recycled may be close to full. This may be
1692 better done manually since the backup files for many machines may be
1693 scattered in many subdirectories.
1698 - Why the heck doesn't bacula drop root priviledges before connecting to
1700 - Look at using posix_fadvise(2) for backups -- see bug #751.
1701 Possibly add the code at findlib/bfile.c:795
1702 /* TCP socket options */
1703 #define TCP_NODELAY 1 /* Turn off Nagle's algorithm. */
1704 #define TCP_MAXSEG 2 /* Limit MSS */
1705 #define TCP_CORK 3 /* Never send partially complete segments */
1706 #define TCP_KEEPIDLE 4 /* Start keeplives after this period */
1707 #define TCP_KEEPINTVL 5 /* Interval between keepalives */
1708 #define TCP_KEEPCNT 6 /* Number of keepalives before death */
1709 #define TCP_SYNCNT 7 /* Number of SYN retransmits */
1710 #define TCP_LINGER2 8 /* Life time of orphaned FIN-WAIT-2 state */
1711 #define TCP_DEFER_ACCEPT 9 /* Wake up listener only when data arrive */
1712 #define TCP_WINDOW_CLAMP 10 /* Bound advertised window */
1713 #define TCP_INFO 11 /* Information about this connection. */
1714 #define TCP_QUICKACK 12 /* Block/reenable quick acks */
1715 #define TCP_CONGESTION 13 /* Congestion control algorithm */
1716 - Fix bnet_connect() code to set a timer and to use time to
1718 - Implement 4th argument to make_catalog_backup that passes hostname.
1719 - Test FIFO backup/restore -- make regression