6 - This patch will give Bacula the option to specify files in
7 FileSets which can be dropped in directories which are Included
8 which will cause that directory not the be backed up.
10 For example, my FileSet contains:
11 # List of files to be backed up
13 Name = "Remote Specified1"
18 File = "\\</etc/bacula-include"
19 IgnoreDir = .notthisone
22 File = "\\</etc/bacula-exclude"
26 And /etc/bacula-include contains:
30 But in /home, there are hundreds of directories of users and some
31 people want to indicate that they don't want to have certain
32 directories backed-up:
37 So I can put them in /etc/bacula-exclude, but that is a system
38 file and not editable for mortal users. To make it possible for
39 users to make it clear to the system that certain directories
40 don't need to be backed up, they now can create file called
43 /home/edwin/www/cache/.notthisone
44 /home/edwin/temp/.notthisone
46 so that the backup system will be clear of rubbish like stuff in
47 these two directories but still that I as administrator of the
48 system don't have to be involved in it.
49 - !!! Cannot restore two jobs a the same time that were
50 written simultaneously unless they were totally spooled.
51 - Document cleaning up the spool files:
52 db, pid, state, bsr, mail, conmsg, spool
53 - Document the multiple-drive-changer.txt script.
54 - Pruning with Admin job.
55 - Does WildFile match against full name? Doc.
56 - %d and %v only valid on Director, not for ClientRunBefore/After.
57 - During tests with the 260 char fix code, I found one problem:
58 if the system "sees" a long path once, it seems to forget it's
59 working drive (e.g. c:\), which will lead to a problem during
60 the next job (create bootstrap file will fail). Here is the
61 workaround: specify absolute working and pid directory in
62 bacula-fd.conf (e.g. c:\bacula\working instead of
64 - Document techniques for restoring large numbers of files.
65 - Document setting my.cnf to big file usage.
66 - Add example of proper index output to doc. show index from File;
67 - Correct the Include syntax in the m4.xxx files in examples/conf
68 - Document JobStatus and Termination codes.
69 - Fix the error with the "DVI file can't be opened" while
70 building the French PDF.
71 - Document more DVD stuff
79 - Document all the little details of setting up certificates for
80 the Bacula data encryption code.
81 - Document more precisely how to use master keys -- especially
82 for disaster recovery.
85 - Migration from other vendors
89 - Backup conf/exe (all daemons)
90 - Backup up system state
91 - Detect state change of system (verify)
92 - Synthetic Full, Diff, Inc (Virtual, Reconstructed)
94 - Modules for Databases, Exchange, ...
95 - Novell NSS backup http://www.novell.com/coolsolutions/tools/18952.html
96 - Compliance norms that compare restored code hash code.
97 - When glibc crash, get address with
99 - How to sync remote offices.
101 http://www.microsoft.com/technet/itshowcase/content/exchbkup.mspx
104 Extract capability (#25)
105 Continued enhancement of bweb
106 Threshold triggered migration jobs (not currently in list, but will be
108 Client triggered backups
109 Complete rework of the scheduling system (not in list)
110 Performance and usage instrumentation (not in list)
111 See email of 21Aug2007 for details.
112 - Look at: http://tech.groups.yahoo.com/group/cfg2html
113 and http://www.openeyet.nl/scc/ for managing customer changes
117 - Dangling softlinks are not restored properly. For example, take a
118 soft link such as src/testprogs/install-sh, which points to /usr/share/autoconf...
119 move the directory to another machine where the file /usr/share/autoconf does
120 not exist, back it up, then try a full restore. It fails.
121 - Check for FD compatibility -- eg .nobackup ...
122 - Re-check new dcr->reserved_volume
123 - Softlinks that point to non-existent file are not restored in restore all,
124 but are restored if the file is individually selected. BUG!
125 - Doc Duplicate Jobs.
126 - New directive "Delete purged Volumes"
128 - Prune by Job Level (Full, Differential, Incremental)
129 - Strict automatic pruning
130 - Implement unmount of USB volumes.
131 - Use "./config no-idea no-mdc2 no-rc5" on building OpenSSL for
132 Win32 to avoid patent problems.
133 - Implement multiple jobid specification for the cancel command,
134 similar to what is permitted on the update slots command.
135 - Implement Bacula plugins -- design API
136 - modify pruning to keep a fixed number of versions of a file,
138 - the cd-command should allow complete paths
139 i.e. cd /foo/bar/foo/bar
140 -> if a customer mails me the path to a certain file,
141 its faster to enter the specified directory
142 - Make tree walk routines like cd, ls, ... more user friendly
143 by handling spaces better.
147 MA = (last_MA * 3 + rate) / 4
148 rate = (bytes - last_bytes) / (runtime - last_runtime)
149 - Add a recursive mark command (rmark) to restore.
150 - "Minimum Job Interval = nnn" sets minimum interval between Jobs
151 of the same level and does not permit multiple simultaneous
152 running of that Job (i.e. lets any previous invocation finish
153 before doing Interval testing).
154 - Look at simplifying File exclusions.
156 - Auto update of slot:
157 rufus-dir: ua_run.c:456-10 JobId=10 NewJobId=10 using pool Full priority=10
158 02-Nov 12:58 rufus-dir JobId 10: Start Backup JobId 10, Job=kernsave.2007-11-02_12.58.03
159 02-Nov 12:58 rufus-dir JobId 10: Using Device "DDS-4"
160 02-Nov 12:58 rufus-sd JobId 10: Invalid slot=0 defined in catalog for Volume "Vol001" on "DDS-4" (/dev/nst0). Manual load my be required.
161 02-Nov 12:58 rufus-sd JobId 10: 3301 Issuing autochanger "loaded? drive 0" command.
162 02-Nov 12:58 rufus-sd JobId 10: 3302 Autochanger "loaded? drive 0", result is Slot 2.
163 02-Nov 12:58 rufus-sd JobId 10: Wrote label to prelabeled Volume "Vol001" on device "DDS-4" (/dev/nst0)
164 02-Nov 12:58 rufus-sd JobId 10: Alert: TapeAlert[7]: Media Life: The tape has reached the end of its useful life.
165 02-Nov 12:58 rufus-dir JobId 10: Bacula rufus-dir 2.3.6 (26Oct07): 02-Nov-2007 12:58:51
166 - Eliminate: /var is a different filesystem. Will not descend from / into /var
167 - Separate Files and Directories in catalog
168 - Create FileVersions table
169 - Look at rsysnc for incremental updates and dedupping
170 - Add MD5 or SHA1 check in SD for data validation
171 - finish implementation of fdcalled -- see ua_run.c:105
172 - Fix problem in postgresql.c in my_postgresql_query, where the
173 generation of the error message doesn't differentiate result==NULL
174 and a bad status from that result. Not only that, the result is
175 cleared on a bail_out without having generated the error message.
177 - Implement SDErrors (must return from SD)
178 - Implement USB keyboard support in rescue CD.
179 - Implement continue spooling while despooling.
180 - Remove all install temp files in Win32 PLUGINSDIR.
181 - Audit retention periods to make sure everything is 64 bit.
182 - No where in restore causes kaboom.
183 - Performance: multiple spool files for a single job.
184 - Performance: despool attributes when despooling data (problem
185 multiplexing Dir connection).
186 - Make restore use the in-use volume reservation algorithm.
187 - When Pool specifies Storage command override does not work.
188 - Implement wait_for_sysop() message display in wait_for_device(), which
189 now prints warnings too often.
190 - Ensure that each device in an Autochanger has a different
192 - Look at sg_logs -a /dev/sg0 for getting soft errors.
193 - btape "test" command with Offline on Unmount = yes
195 This test is essential to Bacula.
197 I'm going to write one record in file 0,
198 two records in file 1,
199 and three records in file 2
201 02-Feb 11:00 btape: ABORTING due to ERROR in dev.c:715
202 dev.c:714 Bad call to rewind. Device "LTO" (/dev/nst0) not open
203 02-Feb 11:00 btape: Fatal Error because: Bacula interrupted by signal 11: Segmentation violation
204 Kaboom! btape, btape got signal 11. Attempting traceback.
206 - Encryption -- email from Landon
207 > The backup encryption algorithm is currently not configurable, and is
208 > set to AES_128_CBC in src/filed/backup.c. The encryption code
209 > supports a number of different ciphers (as well as adding arbitrary
210 > new ones) -- only a small bit of code would be required to map a
211 > configuration string value to a CRYPTO_CIPHER_* value, if anyone is
212 > interested in implementing this functionality.
214 - Figure out some way to "automatically" backup conf changes.
215 - Add the OS version back to the Win32 client info.
216 - Restarted jobs have a NULL in the from field.
217 - Modify SD status command to indicate when the SD is writing
218 to a DVD (the device is not open -- see bug #732).
219 - Look at the possibility of adding "SET NAMES UTF8" for MySQL,
220 and possibly changing the blobs into varchar.
221 - Ensure that the SD re-reads the Media record if the JobFiles
222 does not match -- it may have been updated by another job.
224 - Test Volume compatibility between machine architectures
225 - Encryption documentation
226 - Wrong jobbytes with query 12 (todo)
227 - Bare-metal recovery Windows (todo)
232 - Access Mode = Read-Only, Read-Write, Unavailable, Destroyed, Offsite
234 - Maximum number of scratch volumes
236 - Next Pool (already have)
237 - Reclamation threshold
239 - Reuse delay (after all files purged from volume before it can be used)
240 - Copy Pool = xx, yyy (or multiple lines).
242 - Allow pool selection during restore.
244 - Average tape size from Eric
245 SELECT COALESCE(media_avg_size.volavg,0) * count(Media.MediaId) AS volmax, GROUP BY Media.MediaType, Media.PoolId, media_avg_size.volavg
246 count(Media.MediaId) AS volnum,
247 sum(Media.VolBytes) AS voltotal,
248 Media.PoolId AS PoolId,
249 Media.MediaType AS MediaType
251 LEFT JOIN (SELECT avg(Media.VolBytes) AS volavg,
252 Media.MediaType AS MediaType
254 WHERE Media.VolStatus = 'Full'
255 GROUP BY Media.MediaType
256 ) AS media_avg_size ON (Media.MediaType = media_avg_size.MediaType)
257 GROUP BY Media.MediaType, Media.PoolId, media_avg_size.volavg
261 - Add doc for bweb -- especially Installation
263 http://www.orangecrate.com/modules.php?name=News&file=article&sid=501
265 - Despool attributes in separate thread
268 - Check why restore repeatedly sends Rechdrs between
269 each data chunk -- according to James Harper 9Jan07.
272 - Full at least once a month, ...
273 - Cancel Inc if Diff/Full running
274 - More intelligent re-run
275 - New/deleted file backup
277 - Incremental backup -- rsync, Stow
281 - Try to fix bscan not working with multiple DVD volumes bug #912.
282 - Look at mondo/mindi
283 - Make Bacula by default not backup tmpfs, procfs, sysfs, ...
284 - Fix hardlinked immutable files when linking a second file, the
285 immutable flag must be removed prior to trying to link it.
286 - Implement Python event for backing up/restoring a file.
287 - Change dbcheck to tell users to use native tools for fixing
288 broken databases, and to ensure they have the proper indexes.
289 - add udev rules for Bacula devices.
290 - If a job terminates, the DIR connection can close before the
291 Volume info is updated, leaving the File count wrong.
292 - Look at why SIGPIPE during connection can cause seg fault in
293 writing the daemon message, when Dir dropped to bacula:bacula
294 - Look at zlib 32 => 64 problems.
295 - Possibly turn on St. Bernard code.
296 - Fix bextract to restore ACLs, or better yet, use common routines.
297 - Do we migrate appendable Volumes?
298 - Remove queue.c code.
299 - Print warning message if LANG environment variable does not specify
301 - New dot commands from Arno.
302 .show device=xxx lists information from one storage device, including
303 devices (I'm not even sure that information exists in the DIR...)
304 .move eject device=xxx mostly the same as 'unmount xxx' but perhaps with
305 better machine-readable output like "Ok" or "Error busy"
306 .move eject device=xxx toslot=yyy the same as above, but with a new
307 target slot. The catalog should be updated accordingly.
308 .move transfer device=xxx fromslot=yyy toslot=zzz
311 - Article: http://www.heise.de/open/news/meldung/83231
312 - Article: http://www.golem.de/0701/49756.html
313 - Article: http://lwn.net/Articles/209809/
314 - Article: http://www.onlamp.com/pub/a/onlamp/2004/01/09/bacula.html
315 - Article: http://www.linuxdevcenter.com/pub/a/linux/2005/04/07/bacula.html
316 - Article: http://www.osreviews.net/reviews/admin/bacula
317 - Article: http://www.debianhelp.co.uk/baculaweb.htm
319 - Wikis mentioning Bacula
320 http://wiki.finkproject.org/index.php/Admin:Backups
321 http://wiki.linuxquestions.org/wiki/Bacula
322 http://www.openpkg.org/product/packages/?package=bacula
323 http://www.iterating.com/products/Bacula
324 http://net-snmp.sourceforge.net/wiki/index.php/Net-snmp_extensions
325 http://www.section6.net/wiki/index.php/Using_Bacula_for_Tape_Backups
326 http://bacula.darwinports.com/
327 http://wiki.mandriva.com/en/Releases/Corporate/Server_4/Notes#Bacula
328 http://en.wikipedia.org/wiki/Bacula
331 http://www.devco.net/pubwiki/Bacula/
332 http://paramount.ind.wpi.edu/wiki/doku.php
333 http://gentoo-wiki.com/HOWTO_Backup
334 http://www.georglutz.de/wiki/Bacula
335 http://www.clarkconnect.com/wiki/index.php?title=Modules_-_LAN_Backup/Recovery
336 http://linuxwiki.de/Bacula (in German)
338 - Possibly allow SD to spool even if a tape is not mounted.
339 - Fix re-read of last block to check if job has actually written
340 a block, and check if block was written by a different job
341 (i.e. multiple simultaneous jobs writing).
342 - Figure out how to configure query.sql. Suggestion to use m4:
343 == changequote.m4 ===
344 changequote(`[',`]')dnl
345 ==== query.sql.in ===
346 :List next 20 volumes to expire
348 Pool.Name AS PoolName,
353 [ FROM_UNIXTIME(UNIX_TIMESTAMP(Media.LastWritten) Media.VolRetention) AS Expire, ])dnl
355 [ media.lastwritten + interval '1 second' * media.volretention as expire, ])dnl
359 ON Media.PoolId=Pool.PoolId
360 WHERE Media.LastWritten>0
364 Command: m4 -DmySQL changequote.m4 query.sql.in >query.sql
366 The problem is that it requires m4, which is not present on all machines
368 - Given all the problems with FIFOs, I think the solution is to do something a
369 little different, though I will look at the code and see if there is not some
370 simple solution (i.e. some bug that was introduced). What might be a better
371 solution would be to use a FIFO as a sort of "key" to tell Bacula to read and
372 write data to a program rather than the FIFO. For example, suppose you
377 Then, I could imagine if you backup and restore this file with a direct
378 reference as is currently done for fifos, instead, during backup Bacula will
381 /home/kern/my-fifo.backup
383 and read the data that my-fifo.backup writes to stdout. For restore, Bacula
386 /home/kern/my-fifo.restore
388 and send the data backed up to stdout. These programs can either be an
389 executable or a shell script and they need only read/write to stdin/stdout.
391 I think this would give a lot of flexibility to the user without making any
392 significant changes to Bacula.
397 select FilenameId from Filename where Name='';
398 # Get list of all directories referenced in a Backup.
399 select Path.Path from Path,File where File.JobId=nnn and
400 File.FilenameId=(FilenameId-from-above) and File.PathId=Path.PathId
401 order by Path.Path ASC;
403 - Look into using Dart for testing
404 http://public.kitware.com/Dart/HTML/Index.shtml
406 - Look into replacing autotools with cmake
407 http://www.cmake.org/HTML/Index.html
409 - Mount on an Autochanger with no tape in the drive causes:
410 Automatically selected Storage: LTO-changer
411 Enter autochanger drive[0]: 0
412 3301 Issuing autochanger "loaded drive 0" command.
413 3302 Autochanger "loaded drive 0", result: nothing loaded.
414 3301 Issuing autochanger "loaded drive 0" command.
415 3302 Autochanger "loaded drive 0", result: nothing loaded.
416 3902 Cannot mount Volume on Storage Device "LTO-Drive1" (/dev/nst0) because:
417 Couldn't rewind device "LTO-Drive1" (/dev/nst0): ERR=dev.c:678 Rewind error on "LTO-Drive1" (/dev/nst0). ERR=No medium found.
418 3905 Device "LTO-Drive1" (/dev/nst0) open but no Bacula volume is mounted.
419 If this is not a blank tape, try unmounting and remounting the Volume.
420 - If Drive 0 is blocked, and drive 1 is set "Autoselect=no", drive 1 will
422 - Autochanger did not change volumes.
423 select * from Storage;
424 +-----------+-------------+-------------+
425 | StorageId | Name | AutoChanger |
426 +-----------+-------------+-------------+
427 | 1 | LTO-changer | 0 |
428 +-----------+-------------+-------------+
429 05-May 03:50 roxie-sd: 3302 Autochanger "loaded drive 0", result is Slot 11.
430 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Warning: Director wanted Volume "LT
431 Current Volume "LT0-002" not acceptable because:
432 1997 Volume "LT0-002" not in catalog.
433 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Error: Autochanger Volume "LT0-002"
434 Setting InChanger to zero in catalog.
435 05-May 03:50 roxie-dir: Tibs.2006-05-05_03.05.02 Error: Unable to get Media record
437 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: Error getting Volume i
438 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: Job 530 canceled.
439 05-May 03:50 roxie-sd: Tibs.2006-05-05_03.05.02 Fatal error: spool.c:249 Fatal appe
440 05-May 03:49 Tibs: Tibs.2006-05-05_03.05.02 Fatal error: c:\cygwin\home\kern\bacula
449 FirstWritten: 2006-05-05 03:11:54
450 LastWritten: 2006-05-05 03:50:23
451 LabelDate: 2005-12-26 16:52:40
462 VolRetention: 31,536,000
474 Note VolStatus is blank!!!!!
481 FirstWritten: 0000-00-00 00:00:00
482 LastWritten: 0000-00-00 00:00:00
483 LabelDate: 2005-12-26 16:52:40
494 VolRetention: 31,536,000
507 Automatically selected Storage: LTO-changer
508 Enter autochanger drive[0]: 0
509 3301 Issuing autochanger "loaded drive 0" command.
510 3302 Autochanger "loaded drive 0", result: nothing loaded.
511 3301 Issuing autochanger "loaded drive 0" command.
512 3302 Autochanger "loaded drive 0", result: nothing loaded.
513 3902 Cannot mount Volume on Storage Device "LTO-Drive1" (/dev/nst0) because:
514 Couldn't rewind device "LTO-Drive1" (/dev/nst0): ERR=dev.c:678 Rewind error on "LTO-Drive1" (/dev/nst0). ERR=No medium found.
516 3905 Device "LTO-Drive1" (/dev/nst0) open but no Bacula volume is mounted.
517 If this is not a blank tape, try unmounting and remounting the Volume.
519 - http://www.dwheeler.com/essays/commercial-floss.html
520 - Add VolumeLock to prevent all but lock holder (SD) from updating
521 the Volume data (with the exception of VolumeState).
522 - The btape fill command does not seem to use the Autochanger
523 - Make Windows installer default to system disk drive.
524 - Look at using ioctl(FIOBMAP, ...) on Linux, and
525 DeviceIoControl(..., FSCTL_QUERY_ALLOCATED_RANGES, ...) on
526 Win32 for sparse files.
527 http://www.flexhex.com/docs/articles/sparse-files.phtml
528 http://www.informatik.uni-frankfurt.de/~loizides/reiserfs/fibmap.html
529 - Directive: at <event> "command"
530 - Command: pycmd "command" generates "command" event. How to
531 attach to a specific job?
532 - Integrate Christopher's St. Bernard code.
533 - run_cmd() returns int should return JobId_t
534 - get_next_jobid_from_list() returns int should return JobId_t
535 - Document export LDFLAGS=-L/usr/lib64
536 - Don't attempt to restore from "Disabled" Volumes.
537 - Network error on Win32 should set Win32 error code.
538 - What happens when you rename a Disk Volume?
539 - Job retention period in a Pool (and hence Volume). The job would
541 - Look at -D_FORTIFY_SOURCE=2
542 - Add Win32 FileSet definition somewhere
543 - Look at fixing restore status stats in SD.
544 - Look at using ioctl(FIMAP) and FIGETBSZ for sparse files.
545 http://www.informatik.uni-frankfurt.de/~loizides/reiserfs/fibmap.html
546 - Implement a mode that says when a hard read error is
547 encountered, read many times (as it currently does), and if the
548 block cannot be read, skip to the next block, and try again. If
549 that fails, skip to the next file and try again, ...
551 create table LevelType (LevelType binary(1), LevelTypeLong tinyblob);
552 insert into LevelType (LevelType,LevelTypeLong) values
556 - Show files/second in client status output.
557 - new pool XXX with ScratchPoolId = MyScratchPool's PoolId and
558 let it fill itself, and RecyclePoolId = XXX's PoolId so I can
559 see if it become stable and I just have to supervise
561 - If I want to remove this pool, I set RecyclePoolId = MyScratchPool's
562 PoolId, and when it is empty remove it.
564 - Allow Check Labels to be used with Bacula labels.
565 - "Resuming" a failed backup (lost line for example) by using the
566 failed backup as a sort of "base" job.
568 - Email to the user when the tape is about to need changing x
569 days before it needs changing.
570 - Command to show next tape that will be used for a job even
571 if the job is not scheduled.
572 - From: Arunav Mandal <amandal@trolltech.com>
573 1. When jobs are running and bacula for some reason crashes or if I do a
574 restart it remembers and jobs it was running before it crashed or restarted
575 as of now I loose all jobs if I restart it.
577 2. When spooling and in the midway if client is disconnected for instance a
578 laptop bacula completely discard the spool. It will be nice if it can write
579 that spool to tape so there will be some backups for that client if not all.
581 3. We have around 150 clients machines it will be nice to have a option to
582 upgrade all the client machines bacula version automatically.
584 4. Atleast one connection should be reserved for the bconsole so at heavy load
585 I should connect to the director via bconsole which at sometimes I can't
587 5. Another most important feature that is missing, say at 10am I manually
588 started backup of client abc and it was a full backup since client abc has
589 no backup history and at 10.30am bacula again automatically started backup of
590 client abc as that was in the schedule. So now we have 2 multiple Full
591 backups of the same client and if we again try to start a full backup of
592 client backup abc bacula won't complain. That should be fixed.
594 - For Windows disaster recovery see http://unattended.sf.net/
595 - regardless of the retention period, Bacula will not prune the
596 last Full, Diff, or Inc File data until a month after the
597 retention period for the last Full backup that was done.
598 - update volume=xxx --- add status=Full
599 - Remove old spool files on startup.
600 - Exclude SD spool/working directory.
601 - Refuse to prune last valid Full backup. Same goes for Catalog.
603 - Make a callback when Rerun failed levels is called.
604 - Give Python program access to Scheduled jobs.
605 - Add setting Volume State via Python.
606 - Python script to save with Python, not save, save with Bacula.
607 - Python script to do backup.
609 - Change the Priority, Client, Storage, JobStatus (error)
610 at the start of a job.
611 - Why is SpoolDirectory = /home/bacula/spool; not reported
612 as an error when writing a DVD?
613 - Make bootstrap file handle multiple MediaTypes (SD)
614 - Remove all old Device resource code in Dir and code to pass it
615 back in SD -- better, rework it to pass back device statistics.
616 - Check locking of resources -- be sure to lock devices where previously
617 resources were locked.
618 - The last part is left in the spool dir.
621 - In restore don't compare byte count on a raw device -- directory
622 entry does not contain bytes.
625 - Max Vols limit in Pool off by one?
626 - Implement Files/Bytes,... stats for restore job.
627 - Implement Total Bytes Written, ... for restore job.
628 - Despool attributes simultaneously with data in a separate
629 thread, rejoined at end of data spooling.
630 - 7. Implement new Console commands to allow offlining/reserving drives,
631 and possibly manipulating the autochanger (much asked for).
632 - Add start/end date editing in messages (%t %T, %e?) ...
633 - Add ClientDefs similar to JobDefs.
634 - Print more info when bextract -p accepts a bad block.
635 - Fix FD JobType to be set before RunBeforeJob in FD.
636 - Look at adding full Volume and Pool information to a Volume
637 label so that bscan can get *all* the info.
638 - If the user puts "Purge Oldest Volume = yes" or "Recycle Oldest Volume = yes"
639 and there is only one volume in the pool, refuse to do it -- otherwise
640 he fills the Volume, then immediately starts reusing it.
641 - Implement copies and stripes.
642 - Add history file to console.
643 - Each file on tape creates a JobMedia record. Peter has 4 million
644 files spread over 10000 tape files and four tapes. A restore takes
645 16 hours to build the restore list.
646 - Add and option to check if the file size changed during backup.
647 - Make sure SD deletes spool files on error exit.
648 - Delete old spool files when SD starts.
649 - When labeling tapes, if you enter 000026, Bacula uses
650 the tape index rather than the Volume name 000026.
651 - Add offline tape command to Bacula console.
653 Enter MediaId or Volume name: 32
654 Enter new Volume name: DLT-20Dec04
655 Automatically selected Pool: Default
656 Connecting to Storage daemon DLTDrive at 192.168.68.104:9103 ...
657 Sending relabel command from "DLT-28Jun03" to "DLT-20Dec04" ...
658 block.c:552 Write error at 0:0 on device /dev/nst0. ERR=Bad file descriptor.
659 Error writing final EOF to tape. This tape may not be readable.
660 dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
661 askdir.c:219 NULL Volume name. This shouldn't happen!!!
662 3912 Failed to label Volume: ERR=dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
663 Label command failed for Volume DLT-20Dec04.
664 Do not forget to mount the drive!!!
665 - Bug: if a job is manually scheduled to run later, it does not appear
666 in any status report and cannot be cancelled.
668 ==== Keeping track of deleted/new files ====
669 - To mark files as deleted, run essentially a Verify to disk, and
670 when a file is found missing (MarkId != JobId), then create
671 a new File record with FileIndex == -1. This could be done
672 by the FD at the same time as the backup.
674 My "trick" for keeping track of deletions is the following.
675 Assuming the user turns on this option, after all the files
676 have been backed up, but before the job has terminated, the
677 FD will make a pass through all the files and send their
678 names to the DIR (*exactly* the same as what a Verify job
679 currently does). This will probably be done at the same
680 time the files are being sent to the SD avoiding a second
681 pass. The DIR will then compare that to what is stored in
682 the catalog. Any files in the catalog but not in what the
683 FD sent will receive a catalog File entry that indicates
684 that at that point in time the file was deleted. This
685 either transmitted to the FD or simultaneously computed in
686 the FD, so that the FD can put a record on the tape that
687 indicates that the file has been deleted at this point.
688 A delete file entry could potentially be one with a FileIndex
689 of 0 or perhaps -1 (need to check if FileIndex is used for
690 some other thing as many of the Bacula fields are "overloaded"
693 During a restore, any file initially picked up by some
694 backup (Full, ...) then subsequently having a File entry
695 marked "delete" will be removed from the tree, so will not
696 be restored. If a file with the same name is later OK it
697 will be inserted in the tree -- this already happens. All
698 will be consistent except for possible changes during the
701 Since I'm on the subject, some of you may be wondering what
702 the utility of the in memory tree is if you are going to
703 restore everything (at least it comes up from time to time
704 on the list). Well, it is still *very* useful because it
705 allows only the last item found for a particular filename
706 (full path) to be entered into the tree, and thus if a file
707 is backed up 10 times, only the last copy will be restored.
708 I recently (last Friday) restored a complete directory, and
709 the Full and all the Differential and Incremental backups
710 spanned 3 Volumes. The first Volume was not even mounted
711 because all the files had been updated and hence backed up
712 since the Full backup was made. In this case, the tree
713 saved me a *lot* of time.
715 Make sure this information is stored on the tape too so
716 that it can be restored directly from the tape.
718 All the code (with the exception of formally generating and
719 saving the delete file entries) already exists in the Verify
720 Catalog command. It explicitly recognizes added/deleted files since
721 the last InitCatalog. It is more or less a "simple" matter of
722 taking that code and adapting it slightly to work for backups.
724 Comments from Martin Simmons (I think they are all covered):
725 Ok, that should cover the basics. There are few issues though:
727 - Restore will depend on the catalog. I think it is better to include the
728 extra data in the backup as well, so it can be seen by bscan and bextract.
730 - I'm not sure if it will preserve multiple hard links to the same inode. Or
731 maybe adding or removing links will cause the data to be dumped again?
733 - I'm not sure if it will handle renamed directories. Possibly it will work
734 by dumping the whole tree under a renamed directory?
736 - It remains to be seen how the backup performance of the DIR's will be
737 affected when comparing the catalog for a large filesystem.
739 1. Use the current Director in-memory tree code (very fast), but currently in
740 memory. It probably could be paged.
742 2. Use some DB such as Berkeley DB or SQLite. SQLite is already compiled and
743 built for Win32, and it is something we could compile into the program.
745 3. Implement our own custom DB code.
747 Note, by appropriate use of Directives in the Director, we can dynamically
748 decide if the work is done in the Director or in the FD, and we can even
749 allow the user to choose.
751 === most recent accurate file backup/restore ===
752 Here is a sketch (i.e. more details must be filled in later) that I recently
753 made of an algorithm for doing Accurate Backup.
755 1. Dir informs FD that it is doing an Accurate backup and lookup done by
758 2. FD passes through the file system doing a normal backup based on normal
759 conditions, recording the names of all files and their attributes, and
760 indicating which files were backed up. This is very similar to what Verify
763 3. The Director receives the two lists of files at the end of the FD backup.
764 One, files backed up, and one files not backed up. It then looks up all the
765 files not backed up (using Verify style code).
767 4. The Dir sends the FD a list of:
768 a. Additional files to backup (based on user specified criteria, name, size
769 inode date, hash, ...).
772 5. Dir deletes list of file not backed up.
774 6. FD backs up additional files generates a list of those backed up and sends
775 it to the Director, which adds it to the list of files backed up. The list
776 is now complete and current.
778 7. The FD generates delete records for all the files that were deleted and
781 8. The Dir deletes the previous CurrentBackup list, and then does a
782 transaction insert of the new list that it has.
784 9. The rest works as before ...
788 Two new tables needed.
789 1. CurrentBackupId table that contains Client, JobName, FileSet, and a unique
790 BackupId. This is created during a Full save, and the BackupId can be set to
791 the JobId of the Full save. It will remain the same until another Full
792 backup is done. That is when new records are added during a Differential or
793 Incremental, they must use the same BackupId.
795 2. CurrentBackup table that contains essentially a File record (less a number
796 of fields, but with a few extra fields) -- e.g. a flag that the File was
797 backed up by a Full save (this permits doing a Differential). The unique
798 BackupId allows us to look up the CurrentBackup for a particular Client,
799 Jobname, FileSet using that unique BackupId as the key, so this table must be
800 indexed by the BackupId.
802 Note any time a file is saved by the FD other than during a Full save, the
803 Full save flag is cleared. When doing a Differential backup, if a file has
804 the Full save flag set, it is skipped, otherwise it is backed up. For an
805 Incremental backup, we check to see if the file has changed since the last
806 time we backed it up.
808 Deleted files should have FileIndex == 0
812 How about introducing a Type = MgmtPolicy job type? That job type would
813 be responsible for scanning the Bacula environment looking for specific
814 conditions, and submitting the appropriate jobs for implementing said
818 Name = "Migration-Policy"
820 Policy Selection Job Type = Migrate
821 Scope = "<keyword> <operator> <regexp>"
822 Threshold = "<keyword> <operator> <regexp>"
823 Job Template = <template-name>
826 Where <keyword> is any legal job keyword, <operator> is a comparison
827 operator (=,<,>,!=, logical operators AND/OR/NOT) and <regexp> is a
828 appropriate regexp. I could see an argument for Scope and Threshold
829 being SQL queries if we want to support full flexibility. The
830 Migration-Policy job would then get scheduled as frequently as a site
831 felt necessary (suggested default: every 15 minutes).
836 Name = "Migration-Policy"
838 Policy Selection Job Type = Migration
840 Threshold = "Migration Selection Type = LowestUtil"
841 Job Template = "MigrationTemplate"
844 would select all pools for examination and generate a job based on
845 MigrationTemplate to automatically select the volume with the lowest
846 usage and migrate it's contents to the nextpool defined for that pool.
848 This policy abstraction would be really handy for adjusting the behavior
849 of Bacula according to site-selectable criteria (one thing that pops
850 into mind is Amanda's ability to automatically adjust backup levels
851 depending on various criteria).
857 - Add Pool/Storage override regression test.
858 - Add delete JobId to regression.
859 - Add a regression test for dbcheck.
860 - New test to add bscan to four-concurrent-jobs regression,
861 i.e. after the four-concurrent jobs zap the
862 database as is done in the bscan-test, then use bscan to
863 restore the database, do a restore and compare with the
865 - Add restore of specific JobId to regression (item 3
866 on the restore prompt)
867 - Add IPv6 to regression
868 - Add database test to regression. Test each function like delete,
871 - AntiVir can slow down backups on Win32 systems.
872 - Win32 systems with FAT32 can be much slower than NTFS for
873 more than 1000 files per directory.
877 - A HOLD command to stop all jobs from starting.
878 - A PAUSE command to pause all running jobs ==> release the
880 - Media Type = LTO,LTO-2,LTO-3
881 Media Type Read = LTO,LTO2,LTO3
882 Media Type Write = LTO2, LTO3
884 === From Carsten Menke <bootsy52@gmx.net>
886 Following is a list of what I think in the situations where I'm faced with,
887 could be a usefull enhancement to bacula, which I'm certain other users will
888 benefit from as well.
890 1. NextJob/NextJobs Directive within a Job Resource in the form of
891 NextJobs = job1,job2.
894 I currently solved the problem with running multiple jobs each after each
895 by setting the Max Wait Time for a job to 8 hours, and give
896 the jobs different Priorities. However, there scenarios where
897 1 Job is directly depending on another job, so if the former job fails,
898 the job after it needn't to be run
899 while maybe other jobs should run despite of that
902 A Backup Job and a Verify job, if the backup job fails there is no need to run
903 the verify job, as the backup job already failed. However, one may like
904 to backup the Catalog to disk despite of that the main backup job failed.
907 I see that this is related to the Event Handlers which are on the ToDo
908 list, also it is maybe a good idea to check for the return value and
909 execute different actions based on the return value
912 3. offline capability to bconsole
915 Currently I use a script which I execute within the last Job via the
916 RunAfterJob Directive, to release and eject the tape.
917 So I have to call bconsole "release=Storage-Name" and afterwards
918 mt -f /dev/nst0 eject to get the tape out.
920 If I have multiple Storage Devices, than these may not be /dev/nst0 and
921 I have to modify the script or call it with parameters etc.
922 This would actually not be needed, as everything is already defined
923 in bacula-sd.conf and if I can invoke bconsole with the
924 storage name via $1 in the script than I'm done and information is
927 4. %s for Storage Name added to the chars being substituted in "RunAfterJob"
931 For the reason mentioned in 3. to have the ability to call a
932 script with /scripts/foobar %s and in the script use $1
933 to pass the Storage Name to bconsole
935 5. Setting Volume State within a Job Resource
938 Instead of using "Maximum Volume Jobs" in the Pool Resource,
939 I would have the possibilty to define
940 in a Job Resource that after this certain job is run, the Volume State
941 should be set to "Volume State = Used", this give more flexibility (IMHO).
943 6. Localization of Bacula Messages
946 Unfortunatley many,many people I work with don't speak english very well.
947 So if at least the Reporting messages would be localized then they
948 would understand that they have to change the tape,etc. etc.
950 I volunteer to do the german translations, and if I can convince my wife also
951 french and Morre (western african language).
953 7. OK, this is evil, probably bound to security risks and maybe not possible
954 due to the design of bacula.
956 Implementation of Backtics ( `command` ) for shell comand execution to
957 the "Label Format" Directive.
961 Currently I have defined BACULA_DAY_OF_WEEK="day1|day2..." resulting in
962 Label Format = "HolyBackup-${BACULA_DAY_OF_WEEK[${WeekDay}]}". If I could
963 use backticks than I could use "Label Format = HolyBackup-`date +%A` to have
964 the localized name for the day of the week appended to the
965 format string. Then I have the tape labeled automatically with weekday
966 name in the correct language.
968 - Make output from status use html table tags for nicely
969 presenting in a browser.
970 - Can one write tapes faster with 8192 byte block sizes?
971 - Document security problems with the same password for everyone in
972 rpm and Win32 releases.
973 - Browse generations of files.
974 - I've seen an error when my catalog's File table fills up. I
975 then have to recreate the File table with a larger maximum row
976 size. Relevant information is at
977 http://dev.mysql.com/doc/mysql/en/Full_table.html ; I think the
978 "Installing and Configuring MySQL" chapter should talk a bit
979 about this potential problem, and recommend a solution.
980 - For Solaris must use POSIX awk.
981 - Want speed of writing to tape while despooling.
982 - Supported autochanger:
990 Wangtek 6525ES (SCSI-1 QIC drive, 525MB), under Linux 2.4.something,
991 bacula 1.36.0/1 works with blocksize 16k INSIDE bacula-sd.conf.
992 - Add regex from http://www.pcre.org to Bacula for Win32.
993 - Use only shell tools no make in CDROM package.
994 - Include within include does it work?
995 - Implement a Pool of type Cleaning?
996 - Implement VolReadTime and VolWriteTime in SD
997 - Modify Backing up Your Database to include a bootstrap file.
998 - Think about making certain database errors fatal.
999 - Look at correcting the time jump in the scheduler for daylight
1000 savings time changes.
1001 - Add a "real" timer to network connections.
1002 - Promote to Full = Time period
1003 - Check dates entered by user for correctness (month/day/... ranges)
1004 - Compress restore Volume listing by date and first file.
1005 - Look at patches/bacula_db.b2z postgresql that loops during restore.
1007 - Perhaps add read/write programs and/or plugins to FileSets.
1008 - How to handle backing up portables ...
1009 - Add some sort of guaranteed Interval for upgrading jobs.
1010 - Can we write the state file after every job terminates? On Win32
1011 the system crashes and the state file is not updated.
1014 Documentation to do: (any release a little bit at a time)
1015 - Doc to do unmount before removing magazine.
1016 - Alternative to static linking "ldd prog" save all binaries listed,
1017 restore them and point LD_LIBRARY_PATH to them.
1018 - Document add "</dev/null >/dev/null 2>&1" to the bacula-fd command line
1019 - Document query file format.
1020 - Add more documentation for bsr files.
1021 - Document problems with Verify and pruning.
1022 - Document how to use multiple databases.
1023 - VXA drives have a "cleaning required"
1024 indicator, but Exabyte recommends preventive cleaning after every 75
1027 In this context, it should be noted that Exabyte has a command-line
1028 vxatool utility available for free download. (The current version is
1029 vxatool-3.72.) It can get diagnostic info, read, write and erase tapes,
1030 test the drive, unload tapes, change drive settings, flash new firmware,
1032 Of particular interest in this context is that vxatool <device> -i will
1033 report, among other details, the time since last cleaning in tape motion
1034 minutes. This information can be retrieved (and settings changed, for
1035 that matter) through the generic-SCSI device even when Bacula has the
1036 regular tape device locked. (Needless to say, I don't recommend
1037 changing tape settings while a job is running.)
1038 - Lookup HP cleaning recommendations.
1039 - Lookup HP tape replacement recommendations (see trouble shooting autochanger)
1040 - Document doing table repair
1043 ===================================
1044 - Add macro expansions in JobDefs.
1045 Run Before Job = "SomeFile %{Level} %{Client}"
1046 Write Bootstrap="/some/dir/%{JobName}_%{Client}.bsr"
1047 - Use non-blocking network I/O but if no data is available, use
1049 - Use gather write() for network I/O.
1050 - Autorestart on crash.
1051 - Add bandwidth limiting.
1052 - Add acks every once and a while from the SD to keep
1053 the line from timing out.
1054 - When an error in input occurs and conio beeps, you can back
1055 up through the prompt.
1056 - Detect fixed tape block mode during positioning by looking at
1057 block numbers in btape "test". Possibly adjust in Bacula.
1058 - Fix list volumes to output volume retention in some other
1059 units, perhaps via a directive.
1060 - Allow Simultaneous Priorities = yes => run up to Max concurrent jobs even
1061 with multiple priorities.
1062 - If you use restore replace=never, the directory attributes for
1063 non-existent directories will not be restored properly.
1065 - see lzma401.zip in others directory for new compression
1067 - Allow the user to select JobType for manual pruning/purging.
1068 - bscan does not put first of two volumes back with all info in
1070 - Implement the FreeBSD nodump flag in chflags.
1071 - Figure out how to make named console messages go only to that
1072 console and to the non-restricted console (new console class?).
1073 - Make restricted console prompt for password if *ask* is set or
1074 perhaps if password is undefined.
1075 - Implement "from ISO-date/time every x hours/days/weeks/months" in
1078 ==== from Marc Schoechlin
1079 - the help-command should be more verbose
1080 (it should explain the paramters of the different
1082 -> its time-comsuming to consult the manual anytime
1083 you need a special parameter
1084 -> maybe its more easy to maintain this, if the
1085 descriptions of that commands are outsourced to
1087 - if the password is not configured in bconsole.conf
1088 you should be asked for it.
1089 -> sometimes you like to do restore on a customer-machine
1090 which shouldnt know the password for bacula.
1091 -> adding the password to the file favours admins
1092 to forget to remove the password after usage
1094 the protection of that file is less important
1095 - long-listed-output of commands should be scrollable
1096 like the unix more/less-command does
1097 -> if someone runs 200 and more machines, the lists could
1098 be a little long and complex
1099 - command-output should be shown column by column
1100 to reduce scrolling and to increase clarity
1102 - lsmark should list the selected files with full
1104 - wildcards for selecting and file and directories would be nice
1105 - any actions should be interuptable with STRG+C
1106 - command-expansion would be pretty cool
1108 - When the replace Never option is set, new directory permissions
1109 are not restored. See bug 213. To fix this requires creating a
1110 list of newly restored directories so that those directory
1111 permissions *can* be restored.
1112 - Add prune all command
1113 - Document fact that purge can destroy a part of a restore by purging
1114 one volume while others remain valid -- perhaps mark Jobs.
1115 - Add multiple-media-types.txt
1116 - look at mxt-changer.html
1117 - Make ? do a help command (no return needed).
1118 - Implement restore directory.
1119 - Document streams and how to implement them.
1120 - Try not to re-backup a file if a new hard link is added.
1121 - Add feature to backup hard links only, but not the data.
1122 - Fix stream handling to be simpler.
1123 - Add Priority and Bootstrap to Run a Job.
1124 - Eliminate Restore "Run Restore Job" prompt by allowing new "run command
1126 - Remove View FileSet button from Run a Job dialog.
1127 - Handle prompt for restore job at end of Restore command.
1128 - Add display of total selected files to Restore window.
1129 - Add tree pane to left of window.
1130 - Add progress meter.
1131 - Max wait time or max run time causes seg fault -- see runtime-bug.txt
1132 - Add message to user to check for fixed block size when the forward
1133 space test fails in btape.
1134 - When unmarking a directory check if all files below are unmarked and
1135 then remove the + flag -- in the restore tree.
1136 - Possibly implement: Action = Unmount Device="TapeDrive1" in Admin jobs.
1137 - Setup lrrd graphs: (http://www.linpro.no/projects/lrrd/) Mike Acar.
1138 - Revisit the question of multiple Volumes (disk) on a single device.
1139 - Add a block copy option to bcopy.
1140 - Finish work on Gnome restore GUI.
1141 - Fix "llist jobid=xx" where no fileset or client exists.
1142 - For each job type (Admin, Restore, ...) require only the really necessary
1143 fields.- Pass Director resource name as an option to the Console.
1144 - Add a "batch" mode to the Console (no unsolicited queries, ...).
1145 - Add a .list all files in the restore tree (probably also a list all files)
1146 Do both a long and short form.
1147 - Allow browsing the catalog to see all versions of a file (with
1148 stat data on each file).
1149 - Restore attributes of directory if replace=never set but directory
1151 - Use SHA1 on authentication if possible.
1152 - See comtest-xxx.zip for Windows code to talk to USB.
1153 - Add John's appended files:
1154 Appended = { /files/server/logs/http/*log }
1155 and such files would be treated as follows.On a FULL backup, they would
1156 be backed up like any other file.On an INCREMENTAL backup, where a
1157 previous INCREMENTAL or FULL was already in thecatalogue and the length
1158 of the file wasgreater than the length of the last backup, only thedata
1159 added since the last backup will be dumped.On an INCREMENTAL backup, if
1160 the length of the file is less than thelength of the file with the same
1161 name last backed up, the completefile is dumped.On Windows systems, with
1162 creation date of files, we can be evensmarter about this and not count
1163 entirely upon the length.On a restore, the full and all incrementals
1164 since it will beapplied in sequence to restore the file.
1165 - Check new HAVE_WIN32 open bits.
1166 - Check if the tape has moved before writing.
1167 - Handling removable disks -- see below:
1168 - Keep track of tape use time, and report when cleaning is necessary.
1169 - Add FromClient and ToClient keywords on restore command (or
1170 BackupClient RestoreClient).
1171 - Implement a JobSet, which groups any number of jobs. If the
1172 JobSet is started, all the jobs are started together.
1173 Allow Pool, Level, and Schedule overrides.
1174 - Enhance cancel to timeout BSOCK packets after a specific delay.
1175 - Do scheduling by UTC using gmtime_r() in run_conf, scheduler, and
1176 ua_status.!!! Thanks to Alan Brown for this tip.
1177 - Look at updating Volume Jobs so that Max Volume Jobs = 1 will work
1178 correctly for multiple simultaneous jobs.
1179 - Correct code so that FileSet MD5 is calculated for < and | filename
1181 - Implement the Media record flag that indicates that the Volume does disk
1183 - Implement VolAddr, which is used when Volume is addressed like a disk,
1184 and form it from VolFile and VolBlock.
1185 - Make multiple restore jobs for multiple media types specifying
1186 the proper storage type.
1187 - Fix fast block rejection (stored/read_record.c:118). It passes a null
1188 pointer (rec) to try_repositioning().
1189 - Look at extracting Win data from BackupRead.
1190 - Implement RestoreJobRetention? Maybe better "JobRetention" in a Job,
1191 which would take precidence over the Catalog "JobRetention".
1192 - Implement Label Format in Add and Label console commands.
1193 - Possibly up network buffers to 65K. Put on variable.
1194 - Put email tape request delays on one or more variables. User wants
1195 to cancel the job after a certain time interval. Maximum Mount Wait?
1196 - Job, Client, Device, Pool, or Volume?
1197 Is it possible to make this a directive which is *optional* in multiple
1198 resources, like Level? If so, I think I'd make it an optional directive
1199 in Job, Client, and Pool, with precedence such that Job overrides Client
1200 which in turn overrides Pool.
1202 - New Storage specifications:
1203 - Want to write to multiple storage devices simultaneously
1204 - Want to write to multiple storage devices sequentially (in one job)
1205 - Want to read/write simultaneously
1206 - Key is MediaType -- it must match
1208 Passed to SD as a sort of BSR record called Storage Specification
1212 MediaType -> Next MediaType
1214 Device -> Next Device
1216 Allow multiple Storage specifications
1224 Allow Multiple Pool specifications (note, Pool currently
1226 Allow Multiple MediaType specifications in Dir conf
1227 Allow Multiple Device specifications in Dir conf
1228 Perhaps keep this in a single SSR
1229 Tie a Volume to a specific device by using a MediaType that
1230 is contained in only one device.
1231 In SD allow Device to have Multiple MediaTypes
1233 - Ideas from Jerry Scharf:
1234 First let's point out some big pluses that bacula has for this
1236 more importantly it's active. Thank you so much for that
1237 even more important, it's not flaky
1238 it has an open access catalog, opening many possibilities
1239 it's pushing toward heterogeneous systems capability
1241 Macintosh file client
1242 macs are an interesting niche, but I fear a server is a rathole
1243 working bare iron recovery for windows
1244 the option for inc/diff backups not reset on fileset revision
1245 a) use both change and inode update time against base time
1246 b) do the full catalog check (expensive but accurate)
1247 sizing guide (how much system is needed to back up N systems/files)
1248 consultants on using bacula in building a disaster recovery system
1249 an integration guide
1250 or how to get at fancy things that one could do with bacula
1251 logwatch code for bacula logs (or similar)
1252 linux distro inclusion of bacula (brings good and bad, but necessary)
1253 win2k/XP server capability (icky but you asked)
1254 support for Oracle database ??
1256 - Look at adding SQL server and Exchange support for Windows.
1257 - Make dev->file and dev->block_num signed integers so that -1 can
1258 be an invalid value which happens with BSR.
1259 - Create VolAddr for disk files in place of VolFile and VolBlock. This
1260 is needed to properly specify ranges.
1261 - Add progress of files/bytes to SD and FD.
1262 - Print warning message if FileId > 4 billion
1263 - do a "messages" before the first prompt in Console
1264 - Client does not show busy during Estimate command.
1265 - Implement Console mtx commands.
1266 - Implement a Mount Command and an Unmount Command where
1267 the users could specify a system command to be performed
1268 to do the mount, after which Bacula could attempt to
1269 read the device. This is for Removeable media such as a CDROM.
1270 - Most likely, this mount command would be invoked explicitly
1271 by the user using the current Console "mount" and "unmount"
1272 commands -- the Storage Daemon would do the right thing
1273 depending on the exact nature of the device.
1274 - As with tape drives, when Bacula wanted a new removable
1275 disk mounted, it would unmount the old one, and send a message
1276 to the user, who would then use "mount" as described above
1277 once he had actually inserted the disk.
1278 - Implement dump/print label to UA
1279 - Spool to disk only when the tape is full, then when a tape is hung move
1281 - bextract is sending everything to the log file ****FIXME****
1282 - Allow multiple Storage specifications (or multiple names on
1283 a single Storage specification) in the Job record. Thus a job
1284 can be backed up to a number of storage devices.
1285 - Implement some way for the File daemon to contact the Director
1286 to start a job or pass its DHCP obtained IP number.
1287 - Implement a query tape prompt/replace feature for a console
1288 - Copy console @ code to gnome2-console
1289 - Make sure that Bacula rechecks the tape after the 20 min wait.
1290 - Set IO_NOWAIT on Bacula TCP/IP packets.
1291 - Try doing a raw partition backup and restore by mounting a
1293 - From Lars Kellers:
1294 Yes, it would allow to highly automatic the request for new tapes. If a
1295 tape is empty, bacula reads the barcodes (native or simulated), and if
1296 an unused tape is found, it runs the label command with all the
1297 necessary parameters.
1299 By the way can bacula automatically "move" an empty/purged volume say
1300 in the "short" pool to the "long" pool if this pool runs out of volume
1302 - What to do about "list files job=xxx".
1303 - Look at how fuser works and /proc/PID/fd that is how Nic found the
1304 file descriptor leak in Bacula.
1305 - Implement WrapCounters in Counters.
1306 - Add heartbeat from FD to SD if hb interval expires.
1307 - Can we dynamically change FileSets?
1308 - If pool specified to label command and Label Format is specified,
1309 automatically generate the Volume name.
1310 - Why can't SQL do the filename sort for restore?
1311 - Add ExhautiveRestoreSearch
1312 - Look at the possibility of loading only the necessary
1313 data into the restore tree (i.e. do it one directory at a
1314 time as the user walks through the tree).
1315 - Possibly use the hash code if the user selects all for a restore command.
1316 - Fix "restore all" to bypass building the tree.
1317 - Prohibit backing up archive device (findlib/find_one.c:128)
1318 - Implement Release Device in the Job resource to unmount a drive.
1319 - Implement Acquire Device in the Job resource to mount a drive,
1320 be sure this works with admin jobs so that the user can get
1321 prompted to insert the correct tape. Possibly some way to say to
1322 run the job but don't save the files.
1323 - Make things like list where a file is saved case independent for
1325 - Use autochanger to handle multiple devices.
1326 - Implement a Recycle command
1327 - Start working on Base jobs.
1328 - Implement UnsavedFiles DB record.
1329 - From Phil Stracchino:
1330 It would probably be a per-client option, and would be called
1331 something like, say, "Automatically purge obsoleted jobs". What it
1332 would do is, when you successfully complete a Differential backup of a
1333 client, it would automatically purge all Incremental backups for that
1334 client that are rendered redundant by that Differential. Likewise,
1335 when a Full backup on a client completed, it would automatically purge
1336 all Differential and Incremental jobs obsoleted by that Full backup.
1337 This would let people minimize the number of tapes they're keeping on
1338 hand without having to master the art of retention times.
1339 - When doing a Backup send all attributes back to the Director, who
1340 would then figure out what files have been deleted.
1341 - Currently in mount.c:236 the SD simply creates a Volume. It should have
1342 explicit permission to do so. It should also mark the tape in error
1343 if there is an error.
1344 - Cancel waiting for Client connect in SD if FD goes away.
1346 - Implement timeout in response() when it should come quickly.
1347 - Implement a Slot priority (loaded/not loaded).
1348 - Implement "vacation" Incremental only saves.
1349 - Implement create "FileSet"?
1350 - Add prefixlinks to where or not where absolute links to FD.
1351 - Issue message to mount a new tape before the rewind.
1352 - Simplified client job initiation for portables.
1353 - If SD cannot open a drive, make it periodically retry.
1354 - Add more of the config info to the tape label.
1356 - Refine SD waiting output:
1357 Device is being positioned
1358 > Device is being positioned for append
1359 > Device is being positioned to file x
1361 - Figure out some way to estimate output size and to avoid splitting
1362 a backup across two Volumes -- this could be useful for writing CDROMs
1363 where you really prefer not to have it split -- not serious.
1364 - Have SD compute MD5 or SHA1 and compare to what FD computes.
1365 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
1366 actual data on the Volume and compare it.
1367 - Make bcopy read through bad tape records.
1368 - Program files (i.e. execute a program to read/write files).
1369 Pass read date of last backup, size of file last time.
1370 - Add Signature type to File DB record.
1371 - CD into subdirectory when open()ing files for backup to
1372 speed up things. Test with testfind().
1373 - Priority job to go to top of list.
1374 - Why are save/restore of device different sizes (sparse?) Yup! Fix it.
1375 - Implement some way for the Console to dynamically create a job.
1376 - Solaris -I on tar for include list
1377 - Need a verbose mode in restore, perhaps to bsr.
1378 - bscan without -v is too quiet -- perhaps show jobs.
1379 - Add code to reject whole blocks if not wanted on restore.
1380 - Check if we can increase Bacula FD priorty in Win2000
1381 - Make sure the MaxVolFiles is fully implemented in SD
1382 - Check if both CatalogFiles and UseCatalog are set to SD.
1383 - Possibly add email to Watchdog if drive is unmounted too
1384 long and a job is waiting on the drive.
1385 - After unmount, if restore job started, ask to mount.
1386 - Add UA rc and history files.
1387 - put termcap (used by console) in ./configure and
1388 allow -with-termcap-dir.
1389 - Fix Autoprune for Volumes to respect need for full save.
1390 - Compare tape to Client files (attributes, or attributes and data)
1391 - Make all database Ids 64 bit.
1392 - Allow console commands to detach or run in background.
1393 - Add SD message variables to control operator wait time
1394 - Maximum Operator Wait
1395 - Minimum Message Interval
1396 - Maximum Message Interval
1397 - Send Operator message when cannot read tape label.
1398 - Verify level=Volume (scan only), level=Data (compare of data to file).
1399 Verify level=Catalog, level=InitCatalog
1401 - Add keyword search to show command in Console.
1402 - Events : tape has more than xxx bytes.
1403 - Complete code in Bacula Resources -- this will permit
1404 reading a new config file at any time.
1405 - Handle ctl-c in Console
1406 - Implement script driven addition of File daemon to config files.
1407 - Think about how to make Bacula work better with File (non-tape) archives.
1408 - Write Unix emulator for Windows.
1409 - Put memory utilization in Status output of each daemon
1410 if full status requested or if some level of debug on.
1411 - Make database type selectable by .conf files i.e. at runtime
1412 - Set flag for uname -a. Add to Volume label.
1413 - Restore files modified after date
1414 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
1415 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
1416 - Timout a job or terminate if link goes down, or reopen link and query.
1417 - Concept of precious tapes (cannot be reused).
1418 - Make bcopy copy with a single tape drive.
1419 - Permit changing ownership during restore.
1422 > My suggestion: Add a feature on the systray menu-icon menu to request
1423 > an immediate backup now. This would be useful for laptop users who may
1424 > not be on the network when the regular scheduled backup is run.
1426 > My wife's suggestion: Add a setting to the win32 client to allow it to
1427 > shut down the machine after backup is complete (after, of course,
1428 > displaying a "System will shut down in one minute, click here to cancel"
1429 > warning dialog). This would be useful for sites that want user
1430 > woorkstations to be shut down overnight to save power.
1433 - Autolabel should be specified by DIR instead of SD.
1435 - Add media capacity
1436 - AutoScan (check checksum of tape)
1437 - Format command = "format /dev/nst0"
1441 - Seek resolution (usually corresponds to buffer size)
1442 - EODErrorCode=ENOSPC or code
1443 - Partial Read error code
1444 - Partial write error code
1445 - Nonformatted read error
1446 - Nonformatted write error
1447 - WriteProtected error
1451 - IgnoreCloseErrors=yes
1461 - FD sends unsaved file list to Director at end of job (see
1463 - File daemon should build list of files skipped, and then
1464 at end of save retry and report any errors.
1465 - Write a Storage daemon that uses pipes and
1466 standard Unix programs to write to the tape.
1468 - Need something that monitors the JCR queue and
1469 times out jobs by asking the deamons where they are.
1470 - Enhance Jmsg code to permit buffering and saving to disk.
1471 - device driver = "xxxx" for drives.
1472 - Verify from Volume
1473 - Ensure that /dev/null works
1474 - Need report class for messages. Perhaps
1475 report resource where report=group of messages
1476 - enhance scan_attrib and rename scan_jobtype, and
1477 fill in code for "since" option
1478 - Director needs a time after which the report status is sent
1479 anyway -- or better yet, a retry time for the job.
1480 - Don't reschedule a job if previous incarnation is still running.
1481 - Some way to automatically backup everything is needed????
1482 - Need a structure for pending actions:
1484 - termination status (part of buffered msgs?)
1486 Read, Write, Clean, Delete
1487 - Login to Bacula; Bacula users with different permissions:
1488 owner, group, user, quotas
1489 - Store info on each file system type (probably in the job header on tape.
1490 This could be the output of df; or perhaps some sort of /etc/mtab record.
1492 ========= ideas ===============
1493 From: "Jerry K. Schieffer" <jerry@skylinetechnology.com>
1494 To: <kern@sibbald.com>
1495 Subject: RE: [Bacula-users] future large programming jobs
1496 Date: Thu, 26 Feb 2004 11:34:54 -0600
1498 I noticed the subject thread and thought I would offer the following
1499 merely as sources of ideas, i.e. something to think about, not even as
1500 strong as a request. In my former life (before retiring) I often
1501 dealt with backups and storage management issues/products as a
1502 developer and as a consultant. I am currently migrating my personal
1503 network from amanda to bacula specifically because of the ability to
1504 cross media boundaries during storing backups.
1505 Are you familiar with the commercial product called ADSM (I think IBM
1506 now sells it under the Tivoli label)? It has a couple of interesting
1507 ideas that may apply to the following topics.
1509 1. Migration: Consider that when you need to restore a system, there
1510 may be pressure to hurry. If all the information for a single client
1511 can eventually end up on the same media (and in chronological order),
1512 the restore is facillitated by not having to search past information
1513 from other clients. ADSM has the concept of "client affinity" that
1514 may be associated with it's storage pools. It seems to me that this
1515 concept (as an optional feature) might fit in your architecture for
1518 ADSM also has the concept of defining one or more storage pools as
1519 "copy pools" (almost mirrors, but only in the sense of contents).
1520 These pools provide the ability to have duplicte data stored both
1521 onsite and offsite. The copy process can be scheduled to be handled
1522 by their storage manager during periods when there is no backup
1523 activity. Again, the migration process might be a place to consider
1524 implementing something like this.
1527 > It strikes me that it would be very nice to be able to do things
1529 > have the Job(s) backing up the machines run, and once they have all
1530 > completed, start a migration job to copy the data from disks Volumes
1532 > a tape library and then to offsite storage. Maybe this can already
1534 > done with some careful scheduling and Job prioritzation; the events
1535 > mechanism described below would probably make it very easy.
1537 This is the goal. In the first step (before events), you simply
1539 the Migration to tape later.
1541 2. Base jobs: In ADSM, each copy of each stored file is tracked in
1542 the database. Once a file (unique by path and metadata such as dates,
1543 size, ownership, etc.) is in a copy pool, no more copies are made. In
1544 other words, when you start ADSM, it begins like your concept of a
1545 base job. After that it is in the "incremental" mode. You can
1546 configure the number of "generations" of files to be retained, plus a
1547 retention date after which even old generations are purged. The
1548 database tracks the contents of media and projects the percentage of
1549 each volume that is valid. When the valid content of a volume drops
1550 below a configured percentage, the valid data are migrated to another
1551 volume and the old volume is marked as empty. Note, this requires
1552 ADSM to have an idea of the contents of a client, i.e. marking the
1553 database when an existing file was deleted, but this would solve your
1554 issue of restoring a client without restoring deleted files.
1556 This is pretty far from what bacula now does, but if you are going to
1557 rip things up for Base jobs,.....
1558 Also, the benefits of this are huge for very large shops, especially
1559 with media robots, but are a pain for shops with manual media
1563 > Base jobs sound pretty useful, but I'm not dying for them.
1565 Nobody is dying for them, but when you see what it does, you will die
1568 3. Restoring deleted files: Since I think my comments in (2) above
1569 have low probability of implementation, I'll also suggest that you
1570 could approach the issue of deleted files by a mechanism of having the
1571 fd report to the dir, a list of all files on the client for every
1572 backup job. The dir could note in the database entry for each file
1573 the date that the file was seen. Then if a restore as of date X takes
1574 place, only files that exist from before X until after X would be
1575 restored. Probably the major cost here is the extra date container in
1576 each row of the files table.
1578 Thanks for "listening". I hope some of this helps. If you want to
1579 contact me, please send me an email - I read some but not all of the
1580 mailing list traffic and might miss a reply there.
1582 Please accept my compliments for bacula. It is doing a great job for
1583 me!! I sympathize with you in the need to wrestle with excelence in
1584 execution vs. excelence in feature inclusion.
1589 ==============================
1592 - Design at hierarchial storage for Bacula. Migration and Clone.
1593 - Implement FSM (File System Modules).
1594 - Audit M_ error codes to ensure they are correct and consistent.
1595 - Add variable break characters to lex analyzer.
1596 Either a bit mask or a string of chars so that
1597 the caller can change the break characters.
1598 - Make a single T_BREAK to replace T_COMMA, etc.
1599 - Ensure that File daemon and Storage daemon can
1600 continue a save if the Director goes down (this
1601 is NOT currently the case). Must detect socket error,
1602 buffer messages for later.
1603 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
1604 - Add ability to backup to two Storage devices (two SD sessions) at
1605 the same time -- e.g. onsite, offsite.
1606 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
1607 someway to do so with every volume that has less than x% valid
1611 Migration: Move a backup from one Volume to another
1612 Clone: Copy a backup -- two Volumes
1615 ======================================================
1617 It is somewhat like a Full save becomes an incremental since
1618 the Base job (or jobs) plus other non-base files.
1620 - A Base backup is same as Full backup, just different type.
1621 - New BaseFiles table that contains:
1623 BaseJobId - Base JobId referenced for this FileId (needed ???)
1624 JobId - JobId currently running
1625 FileId - File not backed up, exists in Base Job
1626 FileIndex - FileIndex from Base Job.
1627 i.e. for each base file that exists but is not saved because
1628 it has not changed, the File daemon sends the JobId, BaseId,
1629 FileId, FileIndex back to the Director who creates the DB entry.
1630 - To initiate a Base save, the Director sends the FD
1631 the FileId, and full filename for each file in the Base.
1632 - When the FD finds a Base file, he requests the Director to
1633 send him the full File entry (stat packet plus MD5), or
1634 conversely, the FD sends it to the Director and the Director
1635 says yes or no. This can be quite rapid if the FileId is kept
1636 by the FD for each Base Filename.
1637 - It is probably better to have the comparison done by the FD
1638 despite the fact that the File entry must be sent across the
1640 - An alternative would be to send the FD the whole File entry
1641 from the start. The disadvantage is that it requires a lot of
1642 space. The advantage is that it requires less communications
1644 - The Job record must be updated to indicate that one or more
1646 - At end of Job, FD returns:
1647 1. Count of base files/bytes not written to tape (i.e. matches)
1648 2. Count of base file that were saved i.e. had changed.
1649 - No tape record would be written for a Base file that matches, in the
1650 same way that no tape record is written for Incremental jobs where
1651 the file is not saved because it is unchanged.
1652 - On a restore, all the Base file records must explicitly be
1653 found from the BaseFile tape. I.e. for each Full save that is marked
1654 to have one or more Base Jobs, search the BaseFile for all occurrences
1656 - An optimization might be to make the BaseFile have:
1662 This would avoid the need to explicitly fetch each File record for
1663 the Base job. The Base Job record will be fetched to get the
1664 VolSessionId and VolSessionTime.
1665 =========================================================
1670 Multiple drive autochanger data: see Alan Brown
1671 mtx -f xxx unloadStorage Element 1 is Already Full(drive 0 was empty)
1672 Unloading Data Transfer Element into Storage Element 1...source Element
1673 Address 480 is Empty
1675 (drive 0 was empty and so was slot 1)
1676 > mtx -f xxx load 15 0
1677 no response, just returns to the command prompt when complete.
1678 > mtx -f xxx status Storage Changer /dev/changer:2 Drives, 60 Slots ( 2 Import/Export )
1679 Data Transfer Element 0:Full (Storage Element 15 Loaded):VolumeTag = HX001
1680 Data Transfer Element 1:Empty
1681 Storage Element 1:Empty
1682 Storage Element 2:Full :VolumeTag=HX002
1683 Storage Element 3:Full :VolumeTag=HX003
1684 Storage Element 4:Full :VolumeTag=HX004
1685 Storage Element 5:Full :VolumeTag=HX005
1686 Storage Element 6:Full :VolumeTag=HX006
1687 Storage Element 7:Full :VolumeTag=HX007
1688 Storage Element 8:Full :VolumeTag=HX008
1689 Storage Element 9:Full :VolumeTag=HX009
1690 Storage Element 10:Full :VolumeTag=HX010
1691 Storage Element 11:Empty
1692 Storage Element 12:Empty
1693 Storage Element 13:Empty
1694 Storage Element 14:Empty
1695 Storage Element 15:Empty
1696 Storage Element 16:Empty....
1697 Storage Element 28:Empty
1698 Storage Element 29:Full :VolumeTag=CLNU01L1
1699 Storage Element 30:Empty....
1700 Storage Element 57:Empty
1701 Storage Element 58:Full :VolumeTag=NEX261L2
1702 Storage Element 59 IMPORT/EXPORT:Empty
1703 Storage Element 60 IMPORT/EXPORT:Empty
1705 Unloading Data Transfer Element into Storage Element 15...done
1707 (just to verify it remembers where it came from, however it can be
1708 overrriden with mtx unload {slotnumber} to go to any storage slot.)
1710 There needs to be a table of drive # to devices somewhere - If there are
1711 multiple changers or drives there may not be a 1:1 correspondance between
1712 changer drive number and system device name - and depending on the way the
1713 drives are hooked up to scsi busses, they may not be linearly numbered
1714 from an offset point either.something like
1716 Autochanger drives = 2
1717 Autochanger drive 0 = /dev/nst1
1718 Autochanger drive 1 = /dev/nst2
1719 IMHO, it would be _safest_ to use explicit mtx unload commands at all
1720 times, not just for multidrive changers. For a 1 drive changer, that's
1726 MTX's manpage (1.2.15):
1727 unload [<slotnum>] [ <drivenum> ]
1728 Unloads media from drive <drivenum> into slot
1729 <slotnum>. If <drivenum> is omitted, defaults to
1730 drive 0 (as do all commands). If <slotnum> is
1731 omitted, defaults to the slot that the drive was
1732 loaded from. Note that there's currently no way
1733 to say 'unload drive 1's media to the slot it
1734 came from', other than to explicitly use that
1735 slot number as the destination.AB
1741 undef# camcontrol devlist
1742 <WANGTEK 51000 SCSI M74H 12B3> at scbus0 target 2 lun 0 (pass0,sa0)
1743 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 0 (pass1,sa1)
1744 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 1 (pass2)
1746 tapeinfo -f /dev/sg0 with a bad tape in drive 1:
1747 [kern@rufus mtx-1.2.17kes]$ ./tapeinfo -f /dev/sg0
1748 Product Type: Tape Drive
1750 Product ID: 'C5713A '
1752 Attached Changer: No
1753 TapeAlert[3]: Hard Error: Uncorrectable read/write error.
1754 TapeAlert[20]: Clean Now: The tape drive neads cleaning NOW.
1761 Medium Type: Not Loaded
1764 DataCompEnabled: yes
1765 DataCompCapable: yes
1766 DataDeCompEnabled: yes
1773 Handling removable disks
1775 From: Karl Cunningham <karlc@keckec.com>
1777 My backups are only to hard disk these days, in removable bays. This is my
1778 idea of how a backup to hard disk would work more smoothly. Some of these
1779 things Bacula does already, but I mention them for completeness. If others
1780 have better ways to do this, I'd like to hear about it.
1782 1. Accommodate several disks, rotated similar to how tapes are. Identified
1783 by partition volume ID or perhaps by the name of a subdirectory.
1784 2. Abort & notify the admin if the wrong disk is in the bay.
1785 3. Write backups to different subdirectories for each machine to be backed
1787 4. Volumes (files) get created as needed in the proper subdirectory, one
1789 5. When a disk is recycled, remove or zero all old backup files. This is
1790 important as the disk being recycled may be close to full. This may be
1791 better done manually since the backup files for many machines may be
1792 scattered in many subdirectories.
1797 - Why the heck doesn't bacula drop root priviledges before connecting to
1799 - Look at using posix_fadvise(2) for backups -- see bug #751.
1800 Possibly add the code at findlib/bfile.c:795
1801 /* TCP socket options */
1802 #define TCP_KEEPIDLE 4 /* Start keeplives after this period */
1803 - Fix bnet_connect() code to set a timer and to use time to
1805 - Implement 4th argument to make_catalog_backup that passes hostname.
1806 - Test FIFO backup/restore -- make regression
1807 - Please mount volume "xxx" on Storage device ... should also list
1808 Pool and MediaType in case user needs to create a new volume.
1809 - On restore add Restore Client, Original Client.
1810 01-Apr 00:42 rufus-dir: Start Backup JobId 55, Job=kernsave.2007-04-01_00.42.48
1811 01-Apr 00:42 rufus-sd: Python SD JobStart: JobId=55 Client=Rufus
1812 01-Apr 00:42 rufus-dir: Created new Volume "Full0001" in catalog.
1813 01-Apr 00:42 rufus-dir: Using Device "File"
1814 01-Apr 00:42 rufus-sd: kernsave.2007-04-01_00.42.48 Warning: Device "File" (/tmp) not configured to autolabel Volumes.
1815 01-Apr 00:42 rufus-sd: kernsave.2007-04-01_00.42.48 Warning: Device "File" (/tmp) not configured to autolabel Volumes.
1816 01-Apr 00:42 rufus-sd: Please mount Volume "Full0001" on Storage Device "File" (/tmp) for Job kernsave.2007-04-01_00.42.48
1817 01-Apr 00:44 rufus-sd: Wrote label to prelabeled Volume "Full0001" on device "File" (/tmp)
1818 - Check if gnome-console works with TLS.
1819 - the director seg faulted when I omitted the pool directive from a
1820 job resource. I was experimenting and thought it redundant that I had
1821 specified Pool, Full Backup Pool. and Differential Backup Pool. but
1822 apparently not. This happened when I removed the pool directive and
1823 started the director.
1824 - Add Where: client:/.... to restore job report.
1825 - Ensure that moving a purged Volume in ua_purge.c to the RecyclePool
1826 does the right thing.
1827 - FD-SD quick disconnect
1828 - Building the in memory restore tree is slow.
1829 - Erabt if min_block_size > max_block_size
1830 - Add the ability to consolidate old backup sets (basically do a restore
1831 to tape and appropriately update the catalog). Compress Volume sets.
1832 Might need to spool via file is only one drive is available.
1833 - Why doesn't @"xxx abc" work in a conf file?
1834 - Don't restore Solaris Door files:
1835 #define S_IFDOOR in st_mode.
1836 see: http://docs.sun.com/app/docs/doc/816-5173/6mbb8ae23?a=view#indexterm-360
1837 - Figure out how to recycle Scratch volumes back to the Scratch Pool.
1838 - Implement Despooling data status.
1839 - Use E'xxx' to escape PostgreSQL strings.
1840 - Look at mincore: http://insights.oetiker.ch/linux/fadvise.html
1841 - Unicode input http://en.wikipedia.org/wiki/Byte_Order_Mark
1842 - Look at moving the Storage directive from the Job to the
1843 Pool in the default conf files.
1844 - Look at in src/filed/backup.c
1845 > pm_strcpy(ff_pkt->fname, ff_pkt->fname_save);
1846 > pm_strcpy(ff_pkt->link, ff_pkt->link_save);
1847 - Add Catalog = to Pool resource so that pools will exist
1848 in only one catalog -- currently Pools are "global".
1849 - Add TLS to bat (should be done).
1850 === Duplicate jobs ===
1851 - Done, but implemented somewhat differently than described below!!!
1853 hese apply only to backup jobs.
1855 1. Allow Duplicate Jobs = Yes | No | Higher (Yes)
1857 2. Duplicate Job Interval = <time-interval> (0)
1859 The defaults are in parenthesis and would produce the same behavior as today.
1861 If Allow Duplicate Jobs is set to No, then any job starting while a job of the
1862 same name is running will be canceled.
1864 If Allow Duplicate Jobs is set to Higher, then any job starting with the same
1865 or lower level will be canceled, but any job with a Higher level will start.
1866 The Levels are from High to Low: Full, Differential, Incremental
1868 Finally, if you have Duplicate Job Interval set to a non-zero value, any job
1869 of the same name which starts <time-interval> after a previous job of the
1870 same name would run, any one that starts within <time-interval> would be
1871 subject to the above rules. Another way of looking at it is that the Allow
1872 Duplicate Jobs directive will only apply after <time-interval> of when the
1873 previous job finished (i.e. it is the minimum interval between jobs).
1877 Allow Duplicate Jobs = Yes | No | HigherLevel | CancelLowerLevel (Yes)
1879 Where HigherLevel cancels any waiting job but not any running job.
1880 Where CancelLowerLevel is same as HigherLevel but cancels any running job or
1883 Duplicate Job Proximity = <time-interval> (0)
1885 My suggestion was to define it as the minimum guard time between
1886 executions of a specific job -- ie, if a job was scheduled within Job
1887 Proximity number of seconds, it would be considered a duplicate and
1890 Skip = Do not allow two or more jobs with the same name to run
1891 simultaneously within the proximity interval. The second and subsequent
1892 jobs are skipped without further processing (other than to note the job
1893 and exit immediately), and are not considered errors.
1895 Fail = The second and subsequent jobs that attempt to run during the
1896 proximity interval are cancelled and treated as error-terminated jobs.
1898 Promote = If a job is running, and a second/subsequent job of higher
1899 level attempts to start, the running job is promoted to the higher level
1900 of processing using the resources already allocated, and the subsequent
1901 job is treated as in Skip above.
1907 Allow = yes|no (no = default)
1909 AllowHigherLevel = yes|no (no)
1911 AllowLowerLevel = yes|no (no)
1913 AllowSameLevel = yes|no
1915 Cancel = Running | New (no)
1917 CancelledStatus = Fail | Skip (fail)
1919 Job Proximity = <time-interval> (0)
1920 My suggestion was to define it as the minimum guard time between
1921 executions of a specific job -- ie, if a job was scheduled within Job
1922 Proximity number of seconds, it would be considered a duplicate and
1928 - Fix bpipe.c so that it does not modify results pointer.
1929 ***FIXME*** calling sequence should be changed.