4 Documentation to do: (any release a little bit at a time)
5 - Document running a test version.
6 - Document query file format.
7 - Document static linking
8 - Document problems with Verify and pruning.
9 - Document how to use multiple databases.
10 - For FreeBSD typical /dev/nrsa0 and for mtx /dev/pass1
11 - VXA drives have a "cleaning required"
12 indicator, but Exabyte recommends preventive cleaning after every 75
15 In this context, it should be noted that Exabyte has a command-line
16 vxatool utility available for free download. (The current version is
17 vxatool-3.72.) It can get diagnostic info, read, write and erase tapes,
18 test the drive, unload tapes, change drive settings, flash new firmware,
20 Of particular interest in this context is that vxatool <device> -i will
21 report, among other details, the time since last cleaning in tape motion
22 minutes. This information can be retrieved (and settings changed, for
23 that matter) through the generic-SCSI device even when Bacula has the
24 regular tape device locked. (Needless to say, I don't recommend
25 changing tape settings while a job is running.)
26 - Lookup HP cleaning recommendations.
27 - Lookup HP tape replacement recommendations (see trouble shooting autochanger)
28 - Create a man page for each binary (Debian package requirement).
30 Testing to do: (painful)
31 - that ALL console command line options work and are always implemented
32 - blocksize recognition code.
33 - Test if rewind at end of tape waits for tape to rewind.
35 - Test not zeroing Autochanger slot when it is wrong.
36 - Figure out how to use ssh or stunnel to protect Bacula communications.
38 For 1.33 Testing/Documentation:
39 - Document new alias records in Director. SDAddress SDDeviceName, SDPassword.
40 FDPassword, FDAddress, DBAddress, DBPort, DBPassword.
41 - Document new Include/Exclude ...
42 - Add test of exclusion, test multiple Include {} statements.
43 - Add counter variable test.
44 - Document ln -sf /usr/lib/libncurses.so /usr/lib/libtermcap.so
45 and install the esound-dev package for compiling Console on SuSE.
46 - Add an example of using a FIFO in dirdconf.wml
47 - Add an item to the FAQ about running jobs in different timezones.
48 - Add some examples of job editing codes.
49 - Add Dan's 31 Oct 2003 email on tcpwrappers to Tips.
50 - Add to supported autochangers
52 Auto-Changer : QUALSTAR TLS-4210
53 Manufufactur : Qualstar
54 Tapes : 12 (AIT1: 36GB, AIT2: 50GB all uncompressed)
55 Drives : 2xAIT2 (installed in the Qualstar: SONY SDX-500C AIT2)
56 - Setup a standard job that builds a bootstrap file and saves
57 it with the catalog database.
58 - Document Dan's new --with-dir-user, ... options.
59 - Document estimate command in tree.
60 - Document lsmark command in tree.
63 - Fix fast block rejection (stored/read_record.c:118). It passes a null
64 pointer (rec) to try_repositioning().
65 - See if a restore job can add a file to the tape (prohibit this).
66 - Look at extracting Win data from BackupRead.
67 - Having dashes in filenames apparently creates problems for restore
69 - Add data compare on write/read in btape "test".
70 - Make sure that 64 bit I/O packets are used on Cygwin.
71 - Implement RestoreJobRetention? Maybe better "JobRetention" in a Job,
72 which would take precidence over the Catalog "JobRetention".
73 - Implement Label Format in Add and Label console commands.
74 - Possibly up network buffers to 65K. Put on variable.
75 - Put email tape request delays on one or more variables. User wants
76 to cancel the job after a certain time interval. Maximum Mount Wait?
77 Job, Client, Device, Pool, or Volume?
78 Is it possible to make this a directive which is *optional* in multiple
79 resources, like Level? If so, I think I'd make it an optional directive
80 in Job, Client, and Pool, with precedence such that Job overrides Client
81 which in turn overrides Pool.
82 - Fix Ctl-C crashing the Console (readline?).
86 - Restore: Enter Filename: 'C:/Documents and Settings/Comercial/My
87 Documents/MOP/formulário de registro BELAS ARTES.doc' causes Bacula to
89 - Each DVD-RAM disk would be a volume, just like each tape is
90 a volume. It's a 4.7GB media with random access, but there's nothing about
91 it that I can see that makes it so different than a tape from bacula's
92 perspective. Why couldn't I back up to a bare floppy as a volume (ignoring
94 - Make dev->file and dev->block_num signed integers so that -1 can
95 be an invalid value which happens with BSR.
96 - Create VolAddr for disk files in place of VolFile and VolBlock. This
97 is needed to properly specify ranges.
98 - Print bsmtp output to job report so that problems will be seen.
99 - Pass the number of files to be restored to the FD for reporting
100 - Add progress of files/bytes to SD and FD.
101 - Don't continue Restore if no files selected.
102 - Print warning message if FileId > 4 billion
103 - do a "messages" before the first prompt in Console
104 - Add a date and time stamp at the beginning of every line in the
105 Job report (Volker Sauer).
106 - Client does not show busy during Estimate command.
107 - Implement Console mtx commands.
108 - Look at 2Gb limit for SQLite.
109 - Implement 3 Pools for a Job:
112 Full Backup Pool = xxx
113 Incremental Backup Pool = yyy
114 Differential Backup Pool = zzz
116 - Add a default DB password to MySQL.
117 GRANT all privileges ON bacula.* TO bacula@localhost IDENTIFIED BY
120 - Define week of year for scheduler. W01, W02, ...
121 Week 01 of a year is per definition the first week that has the
122 Thursday in this year, which is equivalent to the week that contains the
123 fourth day of January. In other words, the first week of a new year is
124 the week that has the majority of its days in the new year. Week 01
125 might also contain days from the previous year and the week before week
126 01 of a year is the last week (52 or 53) of the previous year even if it
127 contains days from the new year. A week starts with Monday (day 1) and
128 ends with Sunday (day 7). For example, the first week of the year 1997
129 lasts from 1996-12-30 to 1997-01-05 and can be written in standard
134 The week notation can also be extended by a number indicating the day
135 of the week. For example, the day 1996-12-31, which is the Tuesday (day
136 2) of the first week of 1997, can also be written as
138 1997-W01-2 or 1997W012
140 - Implement a Mount Command and an Unmount Command where
141 the users could specify a system command to be performed
142 to do the mount, after which Bacula could attempt to
143 read the device. This is for Removeable media such as a CDROM.
144 - Most likely, this mount command would be invoked explicitly
145 by the user using the current Console "mount" and "unmount"
146 commands -- the Storage Daemon would do the right thing
147 depending on the exact nature of the device.
148 - As with tape drives, when Bacula wanted a new removable
149 disk mounted, it would unmount the old one, and send a message
150 to the user, who would then use "mount" as described above
151 once he had actually inserted the disk.
152 - Implement dump/print label to UA
153 - Implement disk spooling. Two parts: 1. Spool to disk then
154 immediately to tape to speed up tape operations. 2. Spool to
155 disk only when the tape is full, then when a tape is hung move
157 - Scratch Pool where the volumes can be re-assigned to any Pool.
158 - bextract is sending everything to the log file ****FIXME****
159 - Add Progress command that periodically reports the progress of
161 - Restrict characters permitted in a Resource name, and don't permit
163 - Allow multiple Storage specifications (or multiple names on
164 a single Storage specification) in the Job record. Thus a job
165 can be backed up to a number of storage devices.
166 - Implement some way for the File daemon to contact the Director
167 to start a job or pass its DHCP obtained IP number.
168 - Implement multiple Consoles.
169 - Implement a query tape prompt/replace feature for a console
171 Two jobs ready to go, first one blocked waiting for media
172 Cancel 2nd job ("waiting execution" one)
175 - Copy console @ code to gnome2-console
176 - Make AES the only encryption algorithm see
177 http://csrc.nist.gov/CryptoToolkit/aes/). It's
178 an officially adopted standard, has survived peer
179 review, and provides keys up to 256 bits.
180 - Add ctl-c to console to stop current command and discard buffered
182 - Estimate to Tibs never returns.
183 - Think about how space could be freed up on a tape -- perhaps this
184 is a Merge or Compact feature that is needed.
185 - Modify FileSet, did not upgrade the current Increment job, but
186 waited for the next job to be upgraded.
187 - Take a careful look at SetACL http://setacl.sourceforge.net
188 - Implement a where command for the tree telling where a file
190 - Take a careful look at Level for the estimate command, maybe make
191 it a command line option.
192 - Add Volume name to "I cannot write on this volume because"
193 - Make restore job check if all the files are actually restored.
194 - Make tree walk routines like cd, ls, ... more user friendly
195 by handling spaces better.
196 - Write your PID file and chown root:wheel before drop.
197 - Make sure there is no symlink in a file before creating a
199 - Look at mktemp or mkstemp(3).
200 mktemp and mkstemp create files with predictable names too. That's
201 not the vulnerability. The vulnerability is in creating files without
202 using the O_EXCL flag, which means "only create this file if it doesn't
203 exist, including if the file is a dangling symlink."
205 It is *NOT* enough to do the equivalent of
207 if doesn't exist $filename
208 then create $filename
210 because between the test and the create another process could have
211 gotten the CPU and created the file. You must use atomic functions
212 (those that don't get interrupted by other processes) and O_EXCL is
213 the only way for this particular example.
214 - Automatically create pools, but instead of looking for what
215 in in Job records, walk through the pool resources.
216 - Check and double check tree code, why does it take so long?
217 - Add device name to "Current Volume not acceptable because ..."
218 - Make sure that Bacula rechecks the tape after the 20 min wait.
219 - Set IO_NOWAIT on Bacula TCP/IP packets.
220 - Try doing a raw partition backup and restore by mounting a
222 - Report CVS problems to SourceForge.
223 - Implement .consolerc for Console
225 Yes, it would allow to highly automatic the request for new tapes. If a
226 tape is empty, bacula reads the barcodes (native or simulated), and if
227 an unused tape is found, it runs the label command with all the
228 necessary parameters.
230 By the way can bacula automatically "move" an empty/purged volume say
231 in the "short" pool to the "long" pool if this pool runs out of volume
233 - Either restrict the characters in a name, or fix the problem
234 emailing with names containing / (smtp command line breaks).
235 - Eliminate orphaned jobs: dbcheck, normal pruning, delete job command.
236 Hm. Well, there are the remaining orphaned job records:
238 | 105 | Llioness Save | 0000-00-00 00:00:00 | B | D | 0 | 0 | f |
239 | 110 | Llioness Save | 0000-00-00 00:00:00 | B | I | 0 | 0 | f |
240 | 115 | Llioness Save | 2003-09-10 02:22:03 | B | I | 0 | 0 | A |
241 | 128 | Catalog Save | 2003-09-11 03:53:32 | B | I | 0 | 0 | C |
242 | 131 | Catalog Save | 0000-00-00 00:00:00 | B | I | 0 | 0 | f |
244 As you can see, three of the five are failures. I already deleted the
245 one restore and one other failure using the by-client option. Deciding
246 what is an orphaned job is a tricky problem though, I agree. All these
247 records have or had 0 files/ 0 bytes, except for the restore. With no
248 files, of course, I don't know of the job ever actually becomes
249 associated with a Volume.
251 (I'm not sure if this is documented anywhere -- what are the meanings of
252 all the possible JobStatus codes?)
254 Looking at my database, it appears to me as though all the "orphaned"
255 jobs fit into one of two categories:
257 1) The Job record has a StartTime but no EndTime, and the job is not
260 2) The Job record has an EndTime, indicating that it completed, but
261 it has no associated JobMedia record.
264 This does suggest an approach. If failed jobs (or jobs that, for some
265 other reason, write no files) are associated with a volume via a
266 JobMedia record, then they should be purged when the associated volume
267 is purged. I see two ways to handle jobs that are NOT associated with a
270 1) purge them automatically whenever any volume is manually purged;
272 2) add an option to the purge command to manually purge all jobs with
273 no associated volume.
275 I think Restore jobs also fall into category 2 above .... so one might
276 want to make that "The Job record has an EndTime,, but no associated
277 JobMedia record, and is not a Restore job."
278 - make "btape /tmp" work.
279 - Make sure a rescheduled job is properly reported by status.
280 - Walk through the Pool records rather than the Job records
281 in dird.c to create/update pools.
282 - What to do about "list files job=xxx".
283 - Implement scan: for every slot it finds, zero the slot of
284 Volume other volume having that slot.
285 - When job rescheduled, status gives is waiting for Client Rufus
286 to connect to Storage File. Dir needs to inform SD that job
288 - Fix get_storage_from_media_type (ua_restore) to use command line
290 - Don't print "Warning: Wrong Volume mounted ..." if mounting second volume.
291 - Make Dmsg look at global before calling subroutine.
292 - Enable trace output at runtime for Win32
293 - Available volumes for autochangers (see patrick@baanboard.com 3 Sep 03
294 and 4 Sep) scan slots.
295 - Get and test MySQL 4.0
296 - Do a complete audit of all pthreads_mutex, cond, ... to ensure that
297 any that are dynamically initialized are destroyed when no longer used.
298 - Write a mini-readline with history and editing.
299 - Look at how fuser works and /proc/PID/fd that is how Nic found the
300 file descriptor leak in Bacula.
301 - Implement WrapCounters in Counters.
302 - Turn on SIGHUP in dird.c and test.
303 - Use system dependent calls to get more precise info on tape errors.
304 - Add heartbeat from FD to SD if hb interval expires.
305 - Suppress read error on blank tape when doing a label.
306 - Can we dynamically change FileSets?
307 - If pool specified to label command and Label Format is specified,
308 automatically generate the Volume name.
309 - Take a careful look a the Basic recycling algorithm. When Bacula
310 chooses, the order should be:
312 - Look for Recycle or Purged
315 Instead of using lowest media Id, find the least recently used
318 When the tape is mounted and Bacula requests the status
319 - Do everything possible to use it.
321 Define a "available" status, which is the currently mounted
322 Volume and all volumes that are currently in the autochanger.
324 - Why can't SQL do the filename sort for restore?
325 - Is a pool specification really needed for a restore? Yes, and
326 you may want to exclude archive Pools.
327 - Look at libkse (man kse) for FreeBSD threading.
328 - Look into Microsoft Volume Shadowcopy Service VSS for backing
329 up system state components (Active Directory, System Volume, ...)
330 - Add ExhautiveRestoreSearch
331 - Look at the possibility of loading only the necessary
332 data into the restore tree (i.e. do it one directory at a
333 time as the user walks through the tree).
334 - Possibly use the hash code if the user selects all for a restore command.
335 - Orphaned Dir buffer at parse_conf.c:373 => store_dir
336 - Add Console usr permissions.
337 - Fix "restore all" to bypass building the tree.
338 - Fix restore to list errors if Invalid block found, and if # files
339 restored does not match # expected.
340 - Prohibit backing up archive device (findlib/find_one.c:128)
341 - Implement Release Device in the Job resource to unmount a drive.
342 - Implement Acquire Device in the Job resource to mount a drive,
343 be sure this works with admin jobs so that the user can get
344 prompted to insert the correct tape. Possibly some way to say to
345 run the job but don't save the files.
346 - Implement FileOptions (see end of this document)
347 - Make things like list where a file is saved case independent for
350 - Implement a PostgreSQL driver.
351 - Bacula needs to propagate SD errors.
352 > > cluster-dir: Start Backup JobId 252, Job=REUTERS.2003-08-11_15.04.12
353 > > prod4-sd: REUTERS.2003-08-11_15.04.12 Error: Write error on device
354 /dev/nst0. ERR=Input/output error.
355 > > prod4-sd: REUTERS.2003-08-11_15.04.12 Error: Re-read of last block failed.
356 Last block=5162 Current block=5164.
357 > > prod4-sd: End of medium on Volume "REU007" Bytes=16,303,521,933
359 - Use autochanger to handle multiple devices.
360 - Add SuSE install doc to list.
361 - Check and rechedk "Invalid block number"
362 - Make bextract release the drive properly between tapes
363 so that an autochanger can be made to work.
364 - User wants to NOT backup up certain big files (email files).
365 - Maybe remove multiple simultaneous devices code in SD.
366 - On Windows with very long path names, it may be impossible to create
367 a file (and thus restore it) because the total length is too long.
368 We must cd into the directory then create the file without the
370 - lstat() is not going to work on Win32 for testing date.
371 - Something is not right in last block of fill command.
372 - Implement a Recycle command
373 - Add FileSet to command line arguments for restore.
374 - Add client name to cram-md5 challenge so Director can immediately
375 verify if it is the correct client.
376 - Add JobLevel in FD status (but make sure it is defined).
377 - Audit all UA commands to ensure that we always prompt where possible.
378 - Check Jmsg in bnet, may not work, must dup bsock.
379 - Suppress Job Name in Jmsg for console
380 - Create Pools that are referenced in a Run statement at startup if possible.
381 - Use runbeforejob to unload, then reload a volume previously used,
382 then the next job run gets an error reading the drive.
383 - Make bootstrap filename unique.
384 - Test a second language e.g. french.
385 - Start working on Base jobs.
386 - Make "make binary-release" work from any directory.
387 - Implement UnsavedFiles DB record.
388 - Implement argc/argv for daemon command line scanning using table driven
390 - Implement table driven single argc/argv scanner to pickup all arguments.
391 Much like xxx_conf.c scan table.
392 keyword, handler(store_routine), store_address, code, flags, default.
393 - From Phil Stracchino:
394 It would probably be a per-client option, and would be called
395 something like, say, "Automatically purge obsoleted jobs". What it
396 would do is, when you successfully complete a Differential backup of a
397 client, it would automatically purge all Incremental backups for that
398 client that are rendered redundant by that Differential. Likewise,
399 when a Full backup on a client completed, it would automatically purge
400 all Differential and Incremental jobs obsoleted by that Full backup.
401 This would let people minimize the number of tapes they're keeping on
402 hand without having to master the art of retention times.
403 - Implement a M_SECURITY message class.
404 - When doing a Backup send all attributes back to the Director, who
405 would then figure out what files have been deleted.
406 - Currently in mount.c:236 the SD simply creates a Volume. It should have
407 explicit permission to do so. It should also mark the tape in error
408 if there is an error.
409 - Make sure all restore counters are working correctly in the FD.
410 - SD Bytes Read is wrong.
411 - Look at ALL higher level routines that call block.c to be sure
412 they don't expect something in errmsg.
413 - Investigate doing RAW backup of Win32 partition.
414 - Add thread specific data to hold the jcr -- send error messages from
415 low level routines by accessing it and using Jmsg().
416 - Cancel waiting for Client connect in SD if FD goes away.
418 - Examine Bare Metal restore problem (a FD crash exists somewhere ...).
419 - Implement timeout in response() when it should come quickly.
420 - Implement console @echo command.
421 - Implement a Slot priority (loaded/not loaded).
422 - Implement "vacation" Incremental only saves.
423 - Implement single pane restore (much like the Gftp panes).
424 - Implement Automatic Mount even in operator wait.
425 - Implement create "FileSet"?
426 - Fix watchdog pthread crash on Win32 (this is pthread_kill() Cygwin bug)
427 - Implement "scratch pool" where tapes are defined and can be
428 taken by any pool that needs them.
429 - Implement restore "current system", but take all files without
430 doing selection tree -- so that jobs without File records can
432 - Implement a relocatable bacula.spec
433 - Add prefixlinks to where or not where absolute links to FD.
434 - Issue message to mount a new tape before the rewind.
435 - Simplified client job initiation for portables.
436 - If SD cannot open a drive, make it periodically retry.
437 - Add more of the config info to the tape label.
439 - If tape is marked read-only, then try opening it read-only rather than
440 failing, and remember that it cannot be written.
441 - Refine SD waiting output:
442 Device is being positioned
443 > Device is being positioned for append
444 > Device is being positioned to file x
446 - Figure out some way to estimate output size and to avoid splitting
447 a backup across two Volumes -- this could be useful for writing CDROMs
448 where you really prefer not to have it split -- not serious.
449 - Have SD compute MD5 or SHA1 and compare to what FD computes.
450 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
451 actual data on the Volume and compare it.
452 - Implement Bacula plugins -- design API
453 - Make bcopy read through bad tape records.
454 - Program files (i.e. execute a program to read/write files).
455 Pass read date of last backup, size of file last time.
456 - Add Signature type to File DB record.
457 - CD into subdirectory when open()ing files for backup to
458 speed up things. Test with testfind().
459 - Priority job to go to top of list.
460 - Why are save/restore of device different sizes (sparse?) Yup! Fix it.
461 - Implement some way for the Console to dynamically create a job.
462 - Restore to a particular time -- e.g. before date, after date.
463 - Solaris -I on tar for include list
464 - Need a verbose mode in restore, perhaps to bsr.
465 - bscan without -v is too quiet -- perhaps show jobs.
466 - Add code to reject whole blocks if not wanted on restore.
467 - Check if we can increase Bacula FD priorty in Win2000
468 - Make sure the MaxVolFiles is fully implemented in SD
469 - Check if both CatalogFiles and UseCatalog are set to SD.
470 - Figure out how to do a bare metal Windows restore
471 - Possibly add email to Watchdog if drive is unmounted too
472 long and a job is waiting on the drive.
473 - Restore program that errs in SD due to no tape, reports
474 OK incorrectly in output.
475 - After unmount, if restore job started, ask to mount.
476 - Convert all %x substitution variables, which are hard to remember
477 and read to %(variable-name). Idea from TMDA.
478 - Remove NextId for SQLite. Optimize.
479 - Move all SQL statements into a single location.
480 - Add UA rc and history files.
481 - put termcap (used by console) in ./configure and
482 allow -with-termcap-dir.
483 - Enhance time and size scanning routines.
484 - Fix Autoprune for Volumes to respect need for full save.
485 - Fix Win32 config file definition name on /install
486 - Compare tape to Client files (attributes, or attributes and data)
487 - Make all database Ids 64 bit.
488 - Write an applet for Linux.
489 - Allow console commands to detach or run in background.
490 - Fix status delay on storage daemon during rewind.
491 - Add SD message variables to control operator wait time
492 - Maximum Operator Wait
493 - Minimum Message Interval
494 - Maximum Message Interval
495 - Send Operator message when cannot read tape label.
496 - Verify level=Volume (scan only), level=Data (compare of data to file).
497 Verify level=Catalog, level=InitCatalog
499 - Add keyword search to show command in Console.
500 - Events : tape has more than xxx bytes.
501 - Complete code in Bacula Resources -- this will permit
502 reading a new config file at any time.
503 - Handle ctl-c in Console
504 - Implement script driven addition of File daemon to config files.
505 - Think about how to make Bacula work better with File (non-tape) archives.
506 - Write Unix emulator for Windows.
507 - Put memory utilization in Status output of each daemon
508 if full status requested or if some level of debug on.
509 - Make database type selectable by .conf files i.e. at runtime
510 - Set flag for uname -a. Add to Volume label.
511 - Implement throttled work queue.
512 - Restore files modified after date
513 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
514 - Implement Restore FileSet=
515 - Create a protocol.h and protocol.c where all protocol messages
517 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
518 - Timout a job or terminate if link goes down, or reopen link and query.
519 - Concept of precious tapes (cannot be reused).
520 - Make bcopy copy with a single tape drive.
521 - Permit changing ownership during restore.
524 > My suggestion: Add a feature on the systray menu-icon menu to request
525 > an immediate backup now. This would be useful for laptop users who may
526 > not be on the network when the regular scheduled backup is run.
528 > My wife's suggestion: Add a setting to the win32 client to allow it to
529 > shut down the machine after backup is complete (after, of course,
530 > displaying a "System will shut down in one minute, click here to cancel"
531 > warning dialog). This would be useful for sites that want user
532 > woorkstations to be shut down overnight to save power.
535 - From Terry Manderson <terry@apnic.net>
536 jobset { # new structure
537 name = "monthlyUnixBoxen"
540 jobs = "wakame;durian;soy;wasabi;miso" #new!
550 fileset = "genericUnixSet"
556 fileset = "genericUnixSet"
562 fileset = "UnixDevelBoxSet"
567 - Autolabel should be specified by DIR instead of SD.
570 - AutoScan (check checksum of tape)
571 - Format command = "format /dev/nst0"
575 - Seek resolution (usually corresponds to buffer size)
576 - EODErrorCode=ENOSPC or code
577 - Partial Read error code
578 - Partial write error code
579 - Nonformatted read error
580 - Nonformatted write error
581 - WriteProtected error
585 - IgnoreCloseErrors=yes
595 - FD sends unsaved file list to Director at end of job (see
597 - File daemon should build list of files skipped, and then
598 at end of save retry and report any errors.
599 - Write a Storage daemon that uses pipes and
600 standard Unix programs to write to the tape.
602 - Need something that monitors the JCR queue and
603 times out jobs by asking the deamons where they are.
604 - Enhance Jmsg code to permit buffering and saving to disk.
605 - device driver = "xxxx" for drives.
607 - Ensure that /dev/null works
608 - Need report class for messages. Perhaps
609 report resource where report=group of messages
610 - enhance scan_attrib and rename scan_jobtype, and
611 fill in code for "since" option
612 - Director needs a time after which the report status is sent
613 anyway -- or better yet, a retry time for the job.
614 - Don't reschedule a job if previous incarnation is still running.
615 - Some way to automatically backup everything is needed????
616 - Need a structure for pending actions:
618 - termination status (part of buffered msgs?)
620 Read, Write, Clean, Delete
621 - Login to Bacula; Bacula users with different permissions:
622 owner, group, user, quotas
623 - Store info on each file system type (probably in the job header on tape.
624 This could be the output of df; or perhaps some sort of /etc/mtab record.
627 - Design at hierarchial storage for Bacula. Migration and Clone.
628 - Implement FSM (File System Modules).
629 - Audit M_ error codes to ensure they are correct and consistent.
630 - Add variable break characters to lex analyzer.
631 Either a bit mask or a string of chars so that
632 the caller can change the break characters.
633 - Make a single T_BREAK to replace T_COMMA, etc.
634 - Ensure that File daemon and Storage daemon can
635 continue a save if the Director goes down (this
636 is NOT currently the case). Must detect socket error,
637 buffer messages for later.
638 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
639 - Add ability to backup to two Storage devices (two SD sessions) at
640 the same time -- e.g. onsite, offsite.
641 - Add the ability to consolidate old backup sets (basically do a restore
642 to tape and appropriately update the catalog). Compress Volume sets.
643 Might need to spool via file is only one drive is available.
644 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
645 someway to do so with every volume that has less than x% valid
649 Migration: Move a backup from one Volume to another
650 Clone: Copy a backup -- two Volumes
652 Bacula Migration is based on Jobs (apparently Networker is file by file).
654 Migration triggered by:
658 Highwater mark (keep total size)
663 ======================================================
665 It is somewhat like a Full save becomes an incremental since
666 the Base job (or jobs) plus other non-base files.
668 - A Base backup is same as Full backup, just different type.
669 - New BaseFiles table that contains:
671 BaseJobId - Base JobId referenced for this FileId (needed ???)
672 JobId - JobId currently running
673 FileId - File not backed up, exists in Base Job
674 FileIndex - FileIndex from Base Job.
675 i.e. for each base file that exists but is not saved because
676 it has not changed, the File daemon sends the JobId, BaseId,
677 FileId, FileIndex back to the Director who creates the DB entry.
678 - To initiate a Base save, the Director sends the FD
679 the FileId, and full filename for each file in the Base.
680 - When the FD finds a Base file, he requests the Director to
681 send him the full File entry (stat packet plus MD5), or
682 conversely, the FD sends it to the Director and the Director
683 says yes or no. This can be quite rapid if the FileId is kept
684 by the FD for each Base Filename.
685 - It is probably better to have the comparison done by the FD
686 despite the fact that the File entry must be sent across the
688 - An alternative would be to send the FD the whole File entry
689 from the start. The disadvantage is that it requires a lot of
690 space. The advantage is that it requires less communications
692 - The Job record must be updated to indicate that one or more
694 - At end of Job, FD returns:
695 1. Count of base files/bytes not written to tape (i.e. matches)
696 2. Count of base file that were saved i.e. had changed.
697 - No tape record would be written for a Base file that matches, in the
698 same way that no tape record is written for Incremental jobs where
699 the file is not saved because it is unchanged.
700 - On a restore, all the Base file records must explicitly be
701 found from the BaseFile tape. I.e. for each Full save that is marked
702 to have one or more Base Jobs, search the BaseFile for all occurrences
704 - An optimization might be to make the BaseFile have:
710 This would avoid the need to explicitly fetch each File record for
711 the Base job. The Base Job record will be fetched to get the
712 VolSessionId and VolSessionTime.
713 =========================================================
717 =============================================================
719 Request For Comments For File Backup Options
722 Subject: File Backup Options
725 A few days ago, a Bacula user who is backing up to file volumes and
726 using compression asked if it was possible to suppress compressing
727 all .gz files since it was a waste of CPU time. Although Bacula
728 currently permits using different options (compression, ...) on
729 a directory by directory basis, it cannot do it on a file by
730 file basis, which is clearly what was desired.
732 Proposed Implementation:
733 To solve this problem, I propose the following:
735 - Add a new Director resource type called Options.
737 - The Options resource will have records for all
738 options that can currently be specified on the Include record
739 (in a FileSet). Examples below.
741 - The Options resource will permit an exclude option as well
742 as a number of additional options.
744 - The heart of the Options resource is the ability to
745 supply any number of Match records which specify POSIX
746 regular expressions. These Match regular expressions are
747 applied to the fully qualified filename (path and all). If
748 one matches, then the Options will be used.
750 - When an Match specification matches an included file, the
751 options specified in the Options resource will override
752 the default options specified on the Include record.
754 - Include records will be modified to permit referencing one or
755 more Options resources. The Options will be used
756 in the order listed on the Include record and the first
757 one that matches will be applied.
759 - Options (or specifications) currently supplied on the Include
760 record will be deprecated (i.e. removed in a later version a
761 year or so from now).
763 - The Exclude record will be deprecated as the same functionality
764 can be obtained by using an Exclude = yes in the Options.
767 The following records can appear in the Options resource. An
768 asterisk preceding the name indicates a feature not currently
771 - Regexp "xxx" - Match regular expression
772 - Wild "xxx" - Do a wild card match
775 - Compression= (GZIP, ...)
776 - Signature= (MD5, SHA1, ...)
778 - OneFs= (yes/no) - remain on one filesystem
779 - Recurse= (yes/no) - recurse into subdirectories
780 - Sparse= (yes/no) - do sparse file backup
781 - *Exclude= (yes/no) - exclude file from being saved
782 - *Reader= (filename) - external read (backup) program
783 - *Plugin= (filename) - read/write plugin module
785 - Include= (yes/no) - Include the file matched no additional
786 patterns are applied.
789 - verify= (ipnougsamc5) - verify options
792 - replace= (always/ifnewer/ifolder/never) - replace options currently
794 - *Writer= (filename) - external write (restore) program
798 Currently options specifying compression, MD5 signatures, recursion,
799 ... of a FileSet are supplied on the Include record. These will now
800 all be collected into a Options resource, which will be
801 specified in the Include in place of the options. Multiple Options
802 may be specified. Since the Options may contain regular expressions
803 that are applied to the full filename, this will give the ability
804 to specify backup options on a file by file basis to whatever level
813 Include = compression=GZIP signature=MD5 {
825 Wild = /*.?*/ # matches all files.
830 That's a lot more to do the same thing, but it gives the ability to
831 apply options on a file by file basis. For example, suppose you
832 want to compress all files but not any file with extensions .gz or .Z.
833 In that case, you will need to group two sets of options using
834 the Options resource as follows:
842 # Note multiple Matches are ORed
843 Wild = "*.gz" # matches .gz files
844 Wild = "*.Z" # matches .Z files
849 Wild = "*.?*" # matches all files
855 Now, since the no Compression option is specified in the
856 first group of Options, *.gz or *.Z file will have an MD5 signature computed,
857 but will not be compressed. For all other files, the *.gz *.Z will not
858 match, so the second group of options will be used which will include GZIP
862 - Is it necessary to provide some means of ANDing regular expressions
863 and negation? (not currently planned)
865 e.g. Wild = /*.gz/ && !/big.gz/
867 - I see that Networker has a "null" module which, if specified, does not
868 backup the file, but does make an record of the file in the catalog
869 so that the catalog will reflect an exact picture of the filesystem.
870 The result is that the file can be "seen" when "browsing" the save
871 sets, but it cannot be restored.
873 Is this really useful? Should it be implemented in Bacula?
876 After implementing the above, the user will be able to specify
877 on a file by file basis (using regular expressions) what options are
878 applied for the backup.
881 =============================================
883 ==========================================================
885 For each Incremental job that is run, there may be files that
886 were found but not saved because they were locked (this applies
887 only to Windows). Such a system could send back to the Director
888 a list of Unsaved files.
890 - New UnSavedFiles table that contains:
894 - Then in the next Incremental job, the list of Unsaved Files will be
895 feed to the FD, who will ensure that they are explicitly chosen even
896 if standard date/time check would not have selected them.
897 =============================================================
900 Done: (see kernsdone for more)
903 - John's Full save failed with 1.32c FD and 1.31 Dir no FD status,
904 and no error message.
905 - Add fd and st as Console keywords.
906 - Recycling volume with a Slot requires an operator intervention:
907 rufus-dir: Start Backup JobId 18, Job=kernsave.2003-11-01_21.23.52
908 rufus-dir: Pruned 1 Job on Volume Vol01 from catalog.
909 rufus-dir: There are no Jobs associated with Volume Vol01. Marking it purged.
910 rufus-dir: Recycled volume "Vol01"
911 rufus-sd: Please mount Volume "Vol01" on Storage Device "DDS-4" for Job kernsave.2003-11-01_21.23.52
912 Use "mount" command to release Job.
913 - Implement Dan's bacula script (email of 26 Oct).
914 - Add JobName= to VerifyToCatalog so that all verifies can be done at the end.
915 - Edit the Client/Storage name into authentication failure messages.
916 - Fix packet too big problem. This is most likely a Windows TCP stack
918 - Implement ClientRunBeforeJob and ClientRunAfterJob.
919 - Implement forward spacing block/file: position_device(bsr) --
920 just before read_block_from_device();
922 - Change console to bconsole.
923 - Change smtp to bsmtp.
924 - Fix time difference problem between Bacula and Client
925 so that everything is in GMT.
926 - Fix TimeZone problem!
927 - Mount a tape that is not right for the job (wrong # files on tape)
928 Bacula asks for another tape, fix problems with first tape and
929 say "mount". All works OK, but status shows:
930 Device /dev/nst0 open but no Bacula volume is mounted.
931 Total Bytes=1,153,820,213 Blocks=17,888 Bytes/block=64,502
932 Positioned at File=9 Block=3,951
933 Full Backup job Rufus.2003-10-26_16.45.31 using Volume "DLT-24Oct03" on device /dev/nst0
934 Files=21,003 Bytes=253,954,408 Bytes/sec=2,919,016
935 FDReadSeqNo=192,134 in_msg=129830 out_msg=5 fd=7
936 - Upgrade to cygwin 1.5
937 - Optimize fsf not to read.
938 - Use ioctl() fsf if it exists. Figure out where we are from
939 the mt_status command. Use slow fsf only if other does not work.
940 - Enhance "update slots" to include a "scan" feature
941 scan 1; scan 1-5; scan 1,2,4 ... to update the catalog
942 - Allow a slot or range of slots on the label barcodes command.
943 - Finish implementation of Verify=DiskToCatalog
944 - Make sure that Volumes are recycled based on "Least recently used"
945 rather than lowest MediaId.
946 - Add flag to write only one EOF mark on the tape.
947 - Implement autochanger testing in btape "test" command.
948 - Implement lmark to list everyfile marked.
949 - Make mark/unmark report how many files marked/unmarked.
950 - Keep last 5 or 10 completed jobs and show them in a similar list.
951 - Make a Running Jobs: output similar to current Scheduled Jobs:
952 - Change "create_media_record in bscan to use Archive instead of Full.
953 - Have some way to estimate the restore size or have it printed.
954 - Volume problems occurs if you have valid volume, written, then it is
955 truncated. You get 12-Nov-2003 11:48 rufus-sd: kernsave.2003-11-12_11.48.09 Warning: mount.c:228 Volume on /tmp is not a Bacula labeled Volume, because:
956 block.c:640 Read zero bytes on device /tmp.