9 - Document cleaning up the spool files:
10 db, pid, state, bsr, mail, conmsg, spool
11 - Document the multiple-drive-changer.txt script.
12 - Pruning with Admin job.
13 - Does WildFile match against full name? Doc.
14 - %d and %v only valid on Director, not for ClientRunBefore/After.
19 - Fix re-read of last block to check if job has actually written
20 a block, and check if block was written by a different job
21 (i.e. multiple simultaneous jobs writing).
22 - JobStatus and Termination codes.
23 - Some users claim that they must do two prune commands to get a
24 Volume marked as purged.
25 - Print warning message if LANG environment variable does not specify
27 === Migration from David ===
31 Name = "<poolname>-migrate"
35 Migration Selection Type = LowestUtil | OldestVol | PoolOccupancy |
36 Client | PoolResidence | Volume | JobName | SQLquery
37 Migration Selection Pattern = "regexp"
38 Next Pool = <override>
41 There should be no need for a Level (migration is always Full, since you
42 don't calculate differential/incremental differences for migration),
43 Storage should be determined by the volume types in the pool, and Client
44 is really a selection issue. Migration should always occur to the
45 NextPool defined in the pool definition. If no nextpool is defined, the
46 job should end with a reason of "no place to go". If Next Pool statement
47 is present, we override the check in the pool definition and use the
50 Here's how I'd define Migration Selection Types:
53 Client -- Migrate data from selected client only. Migration Selection
54 Pattern regexp provides pattern to select client names, eg ^FS00* makes
55 all client names starting with FS00 eligible for migration.
57 Jobname -- Migration all jobs matching name. Migration Selection Pattern
58 regexp provides pattern to select jobnames existing in pool.
60 Volume -- Migrate all data on specified volumes. Migration Selection
61 Pattern regexp provides selection criteria for volumes to be migrated.
62 Volumes must exist in pool to be eligible for migration.
66 LowestUtil -- Identify the volume in the pool with the least data on it
67 and empty it. No Migration Selection Pattern required.
69 OldestVol -- Identify the LRU volume with data written, and empty it. No
70 Migration Selection Pattern required.
72 PoolOccupancy -- if pool occupancy exceeds <highmig>, migrate volumes
73 (starting with most full volumes) until pool occupancy drops below
74 <lowmig>. Pool highmig and lowmig values are in pool definition, no
75 Migration Selection Pattern required.
79 SQLQuery -- Migrate all jobuids returned by the supplied SQL query.
80 Migration Selection Pattern contains SQL query to execute; should return
81 a list of 1 or more jobuids to migrate.
83 PoolResidence -- Migrate data sitting in pool for longer than
84 PoolResidence value in pool definition. Migration Selection Pattern
85 optional; if specified, override value in pool definition (value in
89 [ possibly a Python event -- kes ]
91 - Document export LDFLAGS=-L/usr/lib64
92 - Don't attempt to restore from "Disabled" Volumes.
93 - Network error on Win32 should set Win32 error code.
94 - What happens when you rename a Disk Volume?
95 - Job retention period in a Pool (and hence Volume). The job would
97 - Detect resource deadlock in Migrate when same job wants to read
98 and write the same device.
99 - Make hardlink code at line 240 of find_one.c use binary search.
100 - Queue warning/error messages during restore so that they
101 are reported at the end of the report rather than being
102 hidden in the file listing ...
103 - Look at -D_FORTIFY_SOURCE=2
104 - Add Win32 FileSet definition somewhere
105 - Look at fixing restore status stats in SD.
106 - Make selection of Database used in restore correspond to
108 - Implement a mode that says when a hard read error is
109 encountered, read many times (as it currently does), and if the
110 block cannot be read, skip to the next block, and try again. If
111 that fails, skip to the next file and try again, ...
113 create table LevelType (LevelType binary(1), LevelTypeLong tinyblob);
114 insert into LevelType (LevelType,LevelTypeLong) values
118 - Add ACL to restore only to original location.
119 - Add a recursive mark command (rmark) to restore.
120 - "Minimum Job Interval = nnn" sets minimum interval between Jobs
121 of the same level and does not permit multiple simultaneous
122 running of that Job (i.e. lets any previous invocation finish
123 before doing Interval testing).
124 - Look at simplifying File exclusions.
125 - New directive "Delete purged Volumes"
126 - new pool XXX with ScratchPoolId = MyScratchPool's PoolId and
127 let it fill itself, and RecyclePoolId = XXX's PoolId so I can
128 see if it become stable and I just have to supervise
130 - If I want to remove this pool, I set RecyclePoolId = MyScratchPool's
131 PoolId, and when it is empty remove it.
132 - Figure out how to recycle Scratch volumes back to the Scratch Pool.
134 - Allow Check Labels to be used with Bacula labels.
135 - "Resuming" a failed backup (lost line for example) by using the
136 failed backup as a sort of "base" job.
138 - Email to the user when the tape is about to need changing x
139 days before it needs changing.
140 - Command to show next tape that will be used for a job even
141 if the job is not scheduled.
142 - From: Arunav Mandal <amandal@trolltech.com>
143 1. When jobs are running and bacula for some reason crashes or if I do a
144 restart it remembers and jobs it was running before it crashed or restarted
145 as of now I loose all jobs if I restart it.
147 2. When spooling and in the midway if client is disconnected for instance a
148 laptop bacula completely discard the spool. It will be nice if it can write
149 that spool to tape so there will be some backups for that client if not all.
151 3. We have around 150 clients machines it will be nice to have a option to
152 upgrade all the client machines bacula version automatically.
154 4. Atleast one connection should be reserved for the bconsole so at heavy load
155 I should connect to the director via bconsole which at sometimes I can't
157 5. Another most important feature that is missing, say at 10am I manually
158 started backup of client abc and it was a full backup since client abc has
159 no backup history and at 10.30am bacula again automatically started backup of
160 client abc as that was in the schedule. So now we have 2 multiple Full
161 backups of the same client and if we again try to start a full backup of
162 client backup abc bacula won't complain. That should be fixed.
164 - Fix bpipe.c so that it does not modify results pointer.
165 ***FIXME*** calling sequence should be changed.
166 - For Windows disaster recovery see http://unattended.sf.net/
167 - regardless of the retention period, Bacula will not prune the
168 last Full, Diff, or Inc File data until a month after the
169 retention period for the last Full backup that was done.
170 - update volume=xxx --- add status=Full
171 - Remove old spool files on startup.
172 - Exclude SD spool/working directory.
173 - Refuse to prune last valid Full backup. Same goes for Catalog.
175 - Make a callback when Rerun failed levels is called.
176 - Give Python program access to Scheduled jobs.
177 - Add setting Volume State via Python.
178 - Python script to save with Python, not save, save with Bacula.
179 - Python script to do backup.
181 - Change the Priority, Client, Storage, JobStatus (error)
182 at the start of a job.
183 - Why is SpoolDirectory = /home/bacula/spool; not reported
184 as an error when writing a DVD?
185 - Make bootstrap file handle multiple MediaTypes (SD)
186 - Remove all old Device resource code in Dir and code to pass it
187 back in SD -- better, rework it to pass back device statistics.
188 - Check locking of resources -- be sure to lock devices where previously
189 resources were locked.
190 - The last part is left in the spool dir.
193 - In restore don't compare byte count on a raw device -- directory
194 entry does not contain bytes.
195 - To mark files as deleted, run essentially a Verify to disk, and
196 when a file is found missing (MarkId != JobId), then create
197 a new File record with FileIndex == -1. This could be done
198 by the FD at the same time as the backup.
202 MA = (last_MA * 3 + rate) / 4
203 rate = (bytes - last_bytes) / (runtime - last_runtime)
204 - Max Vols limit in Pool off by one?
205 - Implement Files/Bytes,... stats for restore job.
206 - Implement Total Bytes Written, ... for restore job.
207 - Despool attributes simultaneously with data in a separate
208 thread, rejoined at end of data spooling.
209 - 7. Implement new Console commands to allow offlining/reserving drives,
210 and possibly manipulating the autochanger (much asked for).
211 - Add start/end date editing in messages (%t %T, %e?) ...
212 - Add ClientDefs similar to JobDefs.
213 - Print more info when bextract -p accepts a bad block.
214 - Fix FD JobType to be set before RunBeforeJob in FD.
215 - Look at adding full Volume and Pool information to a Volume
216 label so that bscan can get *all* the info.
217 - If the user puts "Purge Oldest Volume = yes" or "Recycle Oldest Volume = yes"
218 and there is only one volume in the pool, refuse to do it -- otherwise
219 he fills the Volume, then immediately starts reusing it.
220 - Implement copies and stripes.
221 - Add history file to console.
222 - Each file on tape creates a JobMedia record. Peter has 4 million
223 files spread over 10000 tape files and four tapes. A restore takes
224 16 hours to build the restore list.
225 - Add and option to check if the file size changed during backup.
226 - Make sure SD deletes spool files on error exit.
227 - Delete old spool files when SD starts.
228 - When labeling tapes, if you enter 000026, Bacula uses
229 the tape index rather than the Volume name 000026.
230 - Add offline tape command to Bacula console.
232 Enter MediaId or Volume name: 32
233 Enter new Volume name: DLT-20Dec04
234 Automatically selected Pool: Default
235 Connecting to Storage daemon DLTDrive at 192.168.68.104:9103 ...
236 Sending relabel command from "DLT-28Jun03" to "DLT-20Dec04" ...
237 block.c:552 Write error at 0:0 on device /dev/nst0. ERR=Bad file descriptor.
238 Error writing final EOF to tape. This tape may not be readable.
239 dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
240 askdir.c:219 NULL Volume name. This shouldn't happen!!!
241 3912 Failed to label Volume: ERR=dev.c:1207 ioctl MTWEOF error on /dev/nst0. ERR=Permission denied.
242 Label command failed for Volume DLT-20Dec04.
243 Do not forget to mount the drive!!!
244 - Bug: if a job is manually scheduled to run later, it does not appear
245 in any status report and cannot be cancelled.
247 ==== Keeping track of deleted files ====
248 My "trick" for keeping track of deletions is the following.
249 Assuming the user turns on this option, after all the files
250 have been backed up, but before the job has terminated, the
251 FD will make a pass through all the files and send their
252 names to the DIR (*exactly* the same as what a Verify job
253 currently does). This will probably be done at the same
254 time the files are being sent to the SD avoiding a second
255 pass. The DIR will then compare that to what is stored in
256 the catalog. Any files in the catalog but not in what the
257 FD sent will receive a catalog File entry that indicates
258 that at that point in time the file was deleted.
260 During a restore, any file initially picked up by some
261 backup (Full, ...) then subsequently having a File entry
262 marked "delete" will be removed from the tree, so will not
263 be restored. If a file with the same name is later OK it
264 will be inserted in the tree -- this already happens. All
265 will be consistent except for possible changes during the
268 Since I'm on the subject, some of you may be wondering what
269 the utility of the in memory tree is if you are going to
270 restore everything (at least it comes up from time to time
271 on the list). Well, it is still *very* useful because it
272 allows only the last item found for a particular filename
273 (full path) to be entered into the tree, and thus if a file
274 is backed up 10 times, only the last copy will be restored.
275 I recently (last Friday) restored a complete directory, and
276 the Full and all the Differential and Incremental backups
277 spanned 3 Volumes. The first Volume was not even mounted
278 because all the files had been updated and hence backed up
279 since the Full backup was made. In this case, the tree
280 saved me a *lot* of time.
282 Make sure this information is stored on the tape too so
283 that it can be restored directly from the tape.
285 Comments from Martin Simmons (I think they are all covered):
286 Ok, that should cover the basics. There are few issues though:
288 - Restore will depend on the catalog. I think it is better to include the
289 extra data in the backup as well, so it can be seen by bscan and bextract.
291 - I'm not sure if it will preserve multiple hard links to the same inode. Or
292 maybe adding or removing links will cause the data to be dumped again?
294 - I'm not sure if it will handle renamed directories. Possibly it will work
295 by dumping the whole tree under a renamed directory?
297 - It remains to be seen how the backup performance of the DIR's will be
298 affected when comparing the catalog for a large filesystem.
302 How about introducing a Type = MgmtPolicy job type? That job type would
303 be responsible for scanning the Bacula environment looking for specific
304 conditions, and submitting the appropriate jobs for implementing said
308 Name = "Migration-Policy"
310 Policy Selection Job Type = Migrate
311 Scope = "<keyword> <operator> <regexp>"
312 Threshold = "<keyword> <operator> <regexp>"
313 Job Template = <template-name>
316 Where <keyword> is any legal job keyword, <operator> is a comparison
317 operator (=,<,>,!=, logical operators AND/OR/NOT) and <regexp> is a
318 appropriate regexp. I could see an argument for Scope and Threshold
319 being SQL queries if we want to support full flexibility. The
320 Migration-Policy job would then get scheduled as frequently as a site
321 felt necessary (suggested default: every 15 minutes).
326 Name = "Migration-Policy"
328 Policy Selection Job Type = Migration
330 Threshold = "Migration Selection Type = LowestUtil"
331 Job Template = "MigrationTemplate"
334 would select all pools for examination and generate a job based on
335 MigrationTemplate to automatically select the volume with the lowest
336 usage and migrate it's contents to the nextpool defined for that pool.
338 This policy abstraction would be really handy for adjusting the behavior
339 of Bacula according to site-selectable criteria (one thing that pops
340 into mind is Amanda's ability to automatically adjust backup levels
341 depending on various criteria).
347 - Add Pool/Storage override regression test.
348 - Add delete JobId to regression.
349 - Add a regression test for dbcheck.
350 - New test to add bscan to four-concurrent-jobs regression,
351 i.e. after the four-concurrent jobs zap the
352 database as is done in the bscan-test, then use bscan to
353 restore the database, do a restore and compare with the
355 - Add restore of specific JobId to regression (item 3
356 on the restore prompt)
357 - Add IPv6 to regression
358 - Add database test to regression. Test each function like delete,
361 - AntiVir can slow down backups on Win32 systems.
362 - Win32 systems with FAT32 can be much slower than NTFS for
363 more than 1000 files per directory.
367 - A HOLD command to stop all jobs from starting.
368 - A PAUSE command to pause all running jobs ==> release the
370 - Media Type = LTO,LTO-2,LTO-3
371 Media Type Read = LTO,LTO2,LTO3
372 Media Type Write = LTO2, LTO3
374 === From Carsten Menke <bootsy52@gmx.net>
376 Following is a list of what I think in the situations where I'm faced with,
377 could be a usefull enhancement to bacula, which I'm certain other users will
378 benefit from as well.
380 1. NextJob/NextJobs Directive within a Job Resource in the form of
381 NextJobs = job1,job2.
384 I currently solved the problem with running multiple jobs each after each
385 by setting the Max Wait Time for a job to 8 hours, and give
386 the jobs different Priorities. However, there scenarios where
387 1 Job is directly depending on another job, so if the former job fails,
388 the job after it needn't to be run
389 while maybe other jobs should run despite of that
392 A Backup Job and a Verify job, if the backup job fails there is no need to run
393 the verify job, as the backup job already failed. However, one may like
394 to backup the Catalog to disk despite of that the main backup job failed.
397 I see that this is related to the Event Handlers which are on the ToDo
398 list, also it is maybe a good idea to check for the return value and
399 execute different actions based on the return value
402 3. offline capability to bconsole
405 Currently I use a script which I execute within the last Job via the
406 RunAfterJob Directive, to release and eject the tape.
407 So I have to call bconsole "release=Storage-Name" and afterwards
408 mt -f /dev/nst0 eject to get the tape out.
410 If I have multiple Storage Devices, than these may not be /dev/nst0 and
411 I have to modify the script or call it with parameters etc.
412 This would actually not be needed, as everything is already defined
413 in bacula-sd.conf and if I can invoke bconsole with the
414 storage name via $1 in the script than I'm done and information is
417 4. %s for Storage Name added to the chars being substituted in "RunAfterJob"
421 For the reason mentioned in 3. to have the ability to call a
422 script with /scripts/foobar %s and in the script use $1
423 to pass the Storage Name to bconsole
425 5. Setting Volume State within a Job Resource
428 Instead of using "Maximum Volume Jobs" in the Pool Resource,
429 I would have the possibilty to define
430 in a Job Resource that after this certain job is run, the Volume State
431 should be set to "Volume State = Used", this give more flexibility (IMHO).
433 6. Localization of Bacula Messages
436 Unfortunatley many,many people I work with don't speak english very well.
437 So if at least the Reporting messages would be localized then they
438 would understand that they have to change the tape,etc. etc.
440 I volunteer to do the german translations, and if I can convince my wife also
441 french and Morre (western african language).
443 7. OK, this is evil, probably bound to security risks and maybe not possible
444 due to the design of bacula.
446 Implementation of Backtics ( `command` ) for shell comand execution to
447 the "Label Format" Directive.
451 Currently I have defined BACULA_DAY_OF_WEEK="day1|day2..." resulting in
452 Label Format = "HolyBackup-${BACULA_DAY_OF_WEEK[${WeekDay}]}". If I could
453 use backticks than I could use "Label Format = HolyBackup-`date +%A` to have
454 the localized name for the day of the week appended to the
455 format string. Then I have the tape labeled automatically with weekday
456 name in the correct language.
458 - Yes, that is surely the case. I probably should turn those into Warning
459 errors. In addition, you just made me think that it might not be bad to
460 add an option to check the file size after backing up the file and
461 report if it changes. This would be done as an option because it would
464 Kern, good idea. If you do do that, mention in the output: file
465 shrunk, or file expanded, just to make it obvious to the user
466 (without having to the refer to file size), just how the file size
469 Would this option be for all file, or just one file? Or a fileset?
470 - Make output from status use html table tags for nicely
471 presenting in a browser.
472 - Can one write tapes faster with 8192 byte block sizes?
473 - Document security problems with the same password for everyone in
474 rpm and Win32 releases.
475 - Browse generations of files.
476 - I've seen an error when my catalog's File table fills up. I
477 then have to recreate the File table with a larger maximum row
478 size. Relevant information is at
479 http://dev.mysql.com/doc/mysql/en/Full_table.html ; I think the
480 "Installing and Configuring MySQL" chapter should talk a bit
481 about this potential problem, and recommend a solution.
482 - For Solaris must use POSIX awk.
483 - Want speed of writing to tape while despooling.
484 - Supported autochanger:
492 Wangtek 6525ES (SCSI-1 QIC drive, 525MB), under Linux 2.4.something,
493 bacula 1.36.0/1 works with blocksize 16k INSIDE bacula-sd.conf.
494 - Add regex from http://www.pcre.org to Bacula for Win32.
495 - Use only shell tools no make in CDROM package.
496 - Include within include does it work?
497 - Implement a Pool of type Cleaning?
498 - Implement VolReadTime and VolWriteTime in SD
499 - Modify Backing up Your Database to include a bootstrap file.
500 - Think about making certain database errors fatal.
501 - Look at correcting the time jump in the scheduler for daylight
502 savings time changes.
503 - Add a "real" timer to network connections.
504 - Promote to Full = Time period
505 - Check dates entered by user for correctness (month/day/... ranges)
506 - Compress restore Volume listing by date and first file.
507 - Look at patches/bacula_db.b2z postgresql that loops during restore.
509 - Perhaps add read/write programs and/or plugins to FileSets.
510 - How to handle backing up portables ...
511 - Add some sort of guaranteed Interval for upgrading jobs.
512 - Can we write the state file after every job terminates? On Win32
513 the system crashes and the state file is not updated.
516 Documentation to do: (any release a little bit at a time)
517 - Doc to do unmount before removing magazine.
518 - Alternative to static linking "ldd prog" save all binaries listed,
519 restore them and point LD_LIBRARY_PATH to them.
520 - Document add "</dev/null >/dev/null 2>&1" to the bacula-fd command line
521 - Document query file format.
522 - Add more documentation for bsr files.
523 - Document problems with Verify and pruning.
524 - Document how to use multiple databases.
525 - VXA drives have a "cleaning required"
526 indicator, but Exabyte recommends preventive cleaning after every 75
529 In this context, it should be noted that Exabyte has a command-line
530 vxatool utility available for free download. (The current version is
531 vxatool-3.72.) It can get diagnostic info, read, write and erase tapes,
532 test the drive, unload tapes, change drive settings, flash new firmware,
534 Of particular interest in this context is that vxatool <device> -i will
535 report, among other details, the time since last cleaning in tape motion
536 minutes. This information can be retrieved (and settings changed, for
537 that matter) through the generic-SCSI device even when Bacula has the
538 regular tape device locked. (Needless to say, I don't recommend
539 changing tape settings while a job is running.)
540 - Lookup HP cleaning recommendations.
541 - Lookup HP tape replacement recommendations (see trouble shooting autochanger)
542 - Document doing table repair
545 ===================================
546 - Add macro expansions in JobDefs.
547 Run Before Job = "SomeFile %{Level} %{Client}"
548 Write Bootstrap="/some/dir/%{JobName}_%{Client}.bsr"
549 - Use non-blocking network I/O but if no data is available, use
551 - Use gather write() for network I/O.
552 - Autorestart on crash.
553 - Add bandwidth limiting.
554 - Add acks every once and a while from the SD to keep
555 the line from timing out.
556 - When an error in input occurs and conio beeps, you can back
557 up through the prompt.
558 - Detect fixed tape block mode during positioning by looking at
559 block numbers in btape "test". Possibly adjust in Bacula.
560 - Fix list volumes to output volume retention in some other
561 units, perhaps via a directive.
562 - If opening a tape in read/write mode fails attempt to open
563 it in read-only mode, and mark the tape for read only.
564 - Allow Simultaneous Priorities = yes => run up to Max concurrent jobs even
565 with multiple priorities.
566 - If you use restore replace=never, the directory attributes for
567 non-existent directories will not be restored properly.
569 - see lzma401.zip in others directory for new compression
571 - Minimal autochanger handling in Bacula and in btape.
572 - Look into how tar does not save sockets and the possiblity of
573 not saving them in Bacula (Martin Simmons reported this).
574 - Fix restore jobs so that multiple jobs can run if they
575 are not using the same tape(s).
576 - Allow the user to select JobType for manual pruning/purging.
577 - bscan does not put first of two volumes back with all info in
579 - Implement the FreeBSD nodump flag in chflags.
580 - Figure out how to make named console messages go only to that
581 console and to the non-restricted console (new console class?).
582 - Make restricted console prompt for password if *ask* is set or
583 perhaps if password is undefined.
584 - Implement "from ISO-date/time every x hours/days/weeks/months" in
587 ==== from Marc Schoechlin
588 - the help-command should be more verbose
589 (it should explain the paramters of the different
591 -> its time-comsuming to consult the manual anytime
592 you need a special parameter
593 -> maybe its more easy to maintain this, if the
594 descriptions of that commands are outsourced to
596 - the cd-command should allow complete paths
597 i.e. cd /foo/bar/foo/bar
598 -> if a customer mails me the path to a certain file,
599 its faster to enter the specified directory
600 - if the password is not configured in bconsole.conf
601 you should be asked for it.
602 -> sometimes you like to do restore on a customer-machine
603 which shouldnt know the password for bacula.
604 -> adding the password to the file favours admins
605 to forget to remove the password after usage
607 the protection of that file is less important
608 - long-listed-output of commands should be scrollable
609 like the unix more/less-command does
610 -> if someone runs 200 and more machines, the lists could
611 be a little long and complex
612 - command-output should be shown column by column
613 to reduce scrolling and to increase clarity
615 - lsmark should list the selected files with full
617 - wildcards for selecting and file and directories would be nice
618 - any actions should be interuptable with STRG+C
619 - command-expansion would be pretty cool
621 - When the replace Never option is set, new directory permissions
622 are not restored. See bug 213. To fix this requires creating a
623 list of newly restored directories so that those directory
624 permissions *can* be restored.
625 - Compaction of Disk space by "migrating" Volumes that have pruned
626 Jobs (what criteria? size, #jobs, time).
627 - Add prune all command
628 - Document fact that purge can destroy a part of a restore by purging
629 one volume while others remain valid -- perhaps mark Jobs.
630 - Add multiple-media-types.txt
631 - look at mxt-changer.html
632 - Make ? do a help command (no return needed).
633 - Implement restore directory.
634 - Document streams and how to implement them.
635 - Try not to re-backup a file if a new hard link is added.
636 - Add feature to backup hard links only, but not the data.
637 - Fix stream handling to be simpler.
638 - Add Priority and Bootstrap to Run a Job.
639 - Eliminate Restore "Run Restore Job" prompt by allowing new "run command
641 - Remove View FileSet button from Run a Job dialog.
642 - Handle prompt for restore job at end of Restore command.
643 - Add display of total selected files to Restore window.
644 - Add tree pane to left of window.
645 - Add progress meter.
646 - Max wait time or max run time causes seg fault -- see runtime-bug.txt
647 - Document writing to a CD/DVD with Bacula.
648 - Add a "base" package to the window installer for pthreadsVCE.dll
649 which is needed by all packages.
650 - Add message to user to check for fixed block size when the forward
651 space test fails in btape.
652 - When unmarking a directory check if all files below are unmarked and
653 then remove the + flag -- in the restore tree.
654 - Possibly implement: Action = Unmount Device="TapeDrive1" in Admin jobs.
655 - Setup lrrd graphs: (http://www.linpro.no/projects/lrrd/) Mike Acar.
656 - Revisit the question of multiple Volumes (disk) on a single device.
657 - Add a block copy option to bcopy.
658 - Investigate adding Mac Resource Forks.
659 - Finish work on Gnome restore GUI.
660 - Fix "llist jobid=xx" where no fileset or client exists.
661 - For each job type (Admin, Restore, ...) require only the really necessary
662 fields.- Pass Director resource name as an option to the Console.
663 - Add a "batch" mode to the Console (no unsolicited queries, ...).
664 - Add a .list all files in the restore tree (probably also a list all files)
665 Do both a long and short form.
666 - Allow browsing the catalog to see all versions of a file (with
667 stat data on each file).
668 - Restore attributes of directory if replace=never set but directory
670 - Use SHA1 on authentication if possible.
671 - See comtest-xxx.zip for Windows code to talk to USB.
672 - Add John's appended files:
673 Appended = { /files/server/logs/http/*log }
674 and such files would be treated as follows.On a FULL backup, they would
675 be backed up like any other file.On an INCREMENTAL backup, where a
676 previous INCREMENTAL or FULL was already in thecatalogue and the length
677 of the file wasgreater than the length of the last backup, only thedata
678 added since the last backup will be dumped.On an INCREMENTAL backup, if
679 the length of the file is less than thelength of the file with the same
680 name last backed up, the completefile is dumped.On Windows systems, with
681 creation date of files, we can be evensmarter about this and not count
682 entirely upon the length.On a restore, the full and all incrementals
683 since it will beapplied in sequence to restore the file.
684 - Check new HAVE_WIN32 open bits.
685 - Check if the tape has moved before writing.
686 - Handling removable disks -- see below:
687 - Keep track of tape use time, and report when cleaning is necessary.
688 - Add FromClient and ToClient keywords on restore command (or
689 BackupClient RestoreClient).
690 - Implement a JobSet, which groups any number of jobs. If the
691 JobSet is started, all the jobs are started together.
692 Allow Pool, Level, and Schedule overrides.
693 - Enhance cancel to timeout BSOCK packets after a specific delay.
694 - Do scheduling by UTC using gmtime_r() in run_conf, scheduler, and
695 ua_status.!!! Thanks to Alan Brown for this tip.
696 - Look at updating Volume Jobs so that Max Volume Jobs = 1 will work
697 correctly for multiple simultaneous jobs.
698 - Correct code so that FileSet MD5 is calculated for < and | filename
700 - Implement the Media record flag that indicates that the Volume does disk
702 - Implement VolAddr, which is used when Volume is addressed like a disk,
703 and form it from VolFile and VolBlock.
704 - Make multiple restore jobs for multiple media types specifying
705 the proper storage type.
706 - Fix fast block rejection (stored/read_record.c:118). It passes a null
707 pointer (rec) to try_repositioning().
708 - Look at extracting Win data from BackupRead.
709 - Implement RestoreJobRetention? Maybe better "JobRetention" in a Job,
710 which would take precidence over the Catalog "JobRetention".
711 - Implement Label Format in Add and Label console commands.
712 - Possibly up network buffers to 65K. Put on variable.
713 - Put email tape request delays on one or more variables. User wants
714 to cancel the job after a certain time interval. Maximum Mount Wait?
715 - Job, Client, Device, Pool, or Volume?
716 Is it possible to make this a directive which is *optional* in multiple
717 resources, like Level? If so, I think I'd make it an optional directive
718 in Job, Client, and Pool, with precedence such that Job overrides Client
719 which in turn overrides Pool.
721 - New Storage specifications:
722 - Want to write to multiple storage devices simultaneously
723 - Want to write to multiple storage devices sequentially (in one job)
724 - Want to read/write simultaneously
725 - Key is MediaType -- it must match
727 Passed to SD as a sort of BSR record called Storage Specification
731 MediaType -> Next MediaType
733 Device -> Next Device
735 Allow multiple Storage specifications
743 Allow Multiple Pool specifications (note, Pool currently
745 Allow Multiple MediaType specifications in Dir conf
746 Allow Multiple Device specifications in Dir conf
747 Perhaps keep this in a single SSR
748 Tie a Volume to a specific device by using a MediaType that
749 is contained in only one device.
750 In SD allow Device to have Multiple MediaTypes
752 - Ideas from Jerry Scharf:
753 First let's point out some big pluses that bacula has for this
755 more importantly it's active. Thank you so much for that
756 even more important, it's not flaky
757 it has an open access catalog, opening many possibilities
758 it's pushing toward heterogeneous systems capability
760 Macintosh file client
761 macs are an interesting niche, but I fear a server is a rathole
762 working bare iron recovery for windows
763 the option for inc/diff backups not reset on fileset revision
764 a) use both change and inode update time against base time
765 b) do the full catalog check (expensive but accurate)
766 sizing guide (how much system is needed to back up N systems/files)
767 consultants on using bacula in building a disaster recovery system
769 or how to get at fancy things that one could do with bacula
770 logwatch code for bacula logs (or similar)
771 linux distro inclusion of bacula (brings good and bad, but necessary)
772 win2k/XP server capability (icky but you asked)
773 support for Oracle database ??
775 - Look at adding SQL server and Exchange support for Windows.
776 - Make dev->file and dev->block_num signed integers so that -1 can
777 be an invalid value which happens with BSR.
778 - Create VolAddr for disk files in place of VolFile and VolBlock. This
779 is needed to properly specify ranges.
780 - Add progress of files/bytes to SD and FD.
781 - Print warning message if FileId > 4 billion
782 - do a "messages" before the first prompt in Console
783 - Client does not show busy during Estimate command.
784 - Implement Console mtx commands.
785 - Implement a Mount Command and an Unmount Command where
786 the users could specify a system command to be performed
787 to do the mount, after which Bacula could attempt to
788 read the device. This is for Removeable media such as a CDROM.
789 - Most likely, this mount command would be invoked explicitly
790 by the user using the current Console "mount" and "unmount"
791 commands -- the Storage Daemon would do the right thing
792 depending on the exact nature of the device.
793 - As with tape drives, when Bacula wanted a new removable
794 disk mounted, it would unmount the old one, and send a message
795 to the user, who would then use "mount" as described above
796 once he had actually inserted the disk.
797 - Implement dump/print label to UA
798 - Spool to disk only when the tape is full, then when a tape is hung move
800 - bextract is sending everything to the log file ****FIXME****
801 - Allow multiple Storage specifications (or multiple names on
802 a single Storage specification) in the Job record. Thus a job
803 can be backed up to a number of storage devices.
804 - Implement some way for the File daemon to contact the Director
805 to start a job or pass its DHCP obtained IP number.
806 - Implement a query tape prompt/replace feature for a console
807 - Copy console @ code to gnome2-console
808 - Make AES the only encryption algorithm see
809 http://csrc.nist.gov/CryptoToolkit/aes/). It's
810 an officially adopted standard, has survived peer
811 review, and provides keys up to 256 bits.
812 - Take a careful look at SetACL http://setacl.sourceforge.net
813 - Make tree walk routines like cd, ls, ... more user friendly
814 by handling spaces better.
815 - Make sure that Bacula rechecks the tape after the 20 min wait.
816 - Set IO_NOWAIT on Bacula TCP/IP packets.
817 - Try doing a raw partition backup and restore by mounting a
820 Yes, it would allow to highly automatic the request for new tapes. If a
821 tape is empty, bacula reads the barcodes (native or simulated), and if
822 an unused tape is found, it runs the label command with all the
823 necessary parameters.
825 By the way can bacula automatically "move" an empty/purged volume say
826 in the "short" pool to the "long" pool if this pool runs out of volume
828 - What to do about "list files job=xxx".
829 - Get and test MySQL 4.0
830 - Look at how fuser works and /proc/PID/fd that is how Nic found the
831 file descriptor leak in Bacula.
832 - Implement WrapCounters in Counters.
833 - Add heartbeat from FD to SD if hb interval expires.
834 - Can we dynamically change FileSets?
835 - If pool specified to label command and Label Format is specified,
836 automatically generate the Volume name.
837 - Why can't SQL do the filename sort for restore?
838 - Add ExhautiveRestoreSearch
839 - Look at the possibility of loading only the necessary
840 data into the restore tree (i.e. do it one directory at a
841 time as the user walks through the tree).
842 - Possibly use the hash code if the user selects all for a restore command.
843 - Fix "restore all" to bypass building the tree.
844 - Prohibit backing up archive device (findlib/find_one.c:128)
845 - Implement Release Device in the Job resource to unmount a drive.
846 - Implement Acquire Device in the Job resource to mount a drive,
847 be sure this works with admin jobs so that the user can get
848 prompted to insert the correct tape. Possibly some way to say to
849 run the job but don't save the files.
850 - Make things like list where a file is saved case independent for
853 - Use autochanger to handle multiple devices.
854 - On Windows with very long path names, it may be impossible to create
855 a file (and thus restore it) because the total length is too long.
856 We must cd into the directory then create the file without the
858 - Implement a Recycle command
859 - Test a second language e.g. french.
860 - Start working on Base jobs.
861 - Implement UnsavedFiles DB record.
862 - From Phil Stracchino:
863 It would probably be a per-client option, and would be called
864 something like, say, "Automatically purge obsoleted jobs". What it
865 would do is, when you successfully complete a Differential backup of a
866 client, it would automatically purge all Incremental backups for that
867 client that are rendered redundant by that Differential. Likewise,
868 when a Full backup on a client completed, it would automatically purge
869 all Differential and Incremental jobs obsoleted by that Full backup.
870 This would let people minimize the number of tapes they're keeping on
871 hand without having to master the art of retention times.
872 - When doing a Backup send all attributes back to the Director, who
873 would then figure out what files have been deleted.
874 - Currently in mount.c:236 the SD simply creates a Volume. It should have
875 explicit permission to do so. It should also mark the tape in error
876 if there is an error.
877 - Cancel waiting for Client connect in SD if FD goes away.
879 - Implement timeout in response() when it should come quickly.
880 - Implement a Slot priority (loaded/not loaded).
881 - Implement "vacation" Incremental only saves.
882 - Implement create "FileSet"?
883 - Add prefixlinks to where or not where absolute links to FD.
884 - Issue message to mount a new tape before the rewind.
885 - Simplified client job initiation for portables.
886 - If SD cannot open a drive, make it periodically retry.
887 - Add more of the config info to the tape label.
889 - If tape is marked read-only, then try opening it read-only rather than
890 failing, and remember that it cannot be written.
891 - Refine SD waiting output:
892 Device is being positioned
893 > Device is being positioned for append
894 > Device is being positioned to file x
896 - Figure out some way to estimate output size and to avoid splitting
897 a backup across two Volumes -- this could be useful for writing CDROMs
898 where you really prefer not to have it split -- not serious.
899 - Have SD compute MD5 or SHA1 and compare to what FD computes.
900 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
901 actual data on the Volume and compare it.
902 - Implement Bacula plugins -- design API
903 - Make bcopy read through bad tape records.
904 - Program files (i.e. execute a program to read/write files).
905 Pass read date of last backup, size of file last time.
906 - Add Signature type to File DB record.
907 - CD into subdirectory when open()ing files for backup to
908 speed up things. Test with testfind().
909 - Priority job to go to top of list.
910 - Why are save/restore of device different sizes (sparse?) Yup! Fix it.
911 - Implement some way for the Console to dynamically create a job.
912 - Solaris -I on tar for include list
913 - Need a verbose mode in restore, perhaps to bsr.
914 - bscan without -v is too quiet -- perhaps show jobs.
915 - Add code to reject whole blocks if not wanted on restore.
916 - Check if we can increase Bacula FD priorty in Win2000
917 - Make sure the MaxVolFiles is fully implemented in SD
918 - Check if both CatalogFiles and UseCatalog are set to SD.
919 - Possibly add email to Watchdog if drive is unmounted too
920 long and a job is waiting on the drive.
921 - After unmount, if restore job started, ask to mount.
922 - Add UA rc and history files.
923 - put termcap (used by console) in ./configure and
924 allow -with-termcap-dir.
925 - Fix Autoprune for Volumes to respect need for full save.
926 - Compare tape to Client files (attributes, or attributes and data)
927 - Make all database Ids 64 bit.
928 - Allow console commands to detach or run in background.
929 - Fix status delay on storage daemon during rewind.
930 - Add SD message variables to control operator wait time
931 - Maximum Operator Wait
932 - Minimum Message Interval
933 - Maximum Message Interval
934 - Send Operator message when cannot read tape label.
935 - Verify level=Volume (scan only), level=Data (compare of data to file).
936 Verify level=Catalog, level=InitCatalog
938 - Add keyword search to show command in Console.
939 - Events : tape has more than xxx bytes.
940 - Complete code in Bacula Resources -- this will permit
941 reading a new config file at any time.
942 - Handle ctl-c in Console
943 - Implement script driven addition of File daemon to config files.
944 - Think about how to make Bacula work better with File (non-tape) archives.
945 - Write Unix emulator for Windows.
946 - Put memory utilization in Status output of each daemon
947 if full status requested or if some level of debug on.
948 - Make database type selectable by .conf files i.e. at runtime
949 - Set flag for uname -a. Add to Volume label.
950 - Restore files modified after date
951 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
952 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
953 - Timout a job or terminate if link goes down, or reopen link and query.
954 - Concept of precious tapes (cannot be reused).
955 - Make bcopy copy with a single tape drive.
956 - Permit changing ownership during restore.
959 > My suggestion: Add a feature on the systray menu-icon menu to request
960 > an immediate backup now. This would be useful for laptop users who may
961 > not be on the network when the regular scheduled backup is run.
963 > My wife's suggestion: Add a setting to the win32 client to allow it to
964 > shut down the machine after backup is complete (after, of course,
965 > displaying a "System will shut down in one minute, click here to cancel"
966 > warning dialog). This would be useful for sites that want user
967 > woorkstations to be shut down overnight to save power.
970 - Autolabel should be specified by DIR instead of SD.
973 - AutoScan (check checksum of tape)
974 - Format command = "format /dev/nst0"
978 - Seek resolution (usually corresponds to buffer size)
979 - EODErrorCode=ENOSPC or code
980 - Partial Read error code
981 - Partial write error code
982 - Nonformatted read error
983 - Nonformatted write error
984 - WriteProtected error
988 - IgnoreCloseErrors=yes
998 - FD sends unsaved file list to Director at end of job (see
1000 - File daemon should build list of files skipped, and then
1001 at end of save retry and report any errors.
1002 - Write a Storage daemon that uses pipes and
1003 standard Unix programs to write to the tape.
1005 - Need something that monitors the JCR queue and
1006 times out jobs by asking the deamons where they are.
1007 - Enhance Jmsg code to permit buffering and saving to disk.
1008 - device driver = "xxxx" for drives.
1009 - Verify from Volume
1010 - Ensure that /dev/null works
1011 - Need report class for messages. Perhaps
1012 report resource where report=group of messages
1013 - enhance scan_attrib and rename scan_jobtype, and
1014 fill in code for "since" option
1015 - Director needs a time after which the report status is sent
1016 anyway -- or better yet, a retry time for the job.
1017 - Don't reschedule a job if previous incarnation is still running.
1018 - Some way to automatically backup everything is needed????
1019 - Need a structure for pending actions:
1021 - termination status (part of buffered msgs?)
1023 Read, Write, Clean, Delete
1024 - Login to Bacula; Bacula users with different permissions:
1025 owner, group, user, quotas
1026 - Store info on each file system type (probably in the job header on tape.
1027 This could be the output of df; or perhaps some sort of /etc/mtab record.
1029 ========= ideas ===============
1030 From: "Jerry K. Schieffer" <jerry@skylinetechnology.com>
1031 To: <kern@sibbald.com>
1032 Subject: RE: [Bacula-users] future large programming jobs
1033 Date: Thu, 26 Feb 2004 11:34:54 -0600
1035 I noticed the subject thread and thought I would offer the following
1036 merely as sources of ideas, i.e. something to think about, not even as
1037 strong as a request. In my former life (before retiring) I often
1038 dealt with backups and storage management issues/products as a
1039 developer and as a consultant. I am currently migrating my personal
1040 network from amanda to bacula specifically because of the ability to
1041 cross media boundaries during storing backups.
1042 Are you familiar with the commercial product called ADSM (I think IBM
1043 now sells it under the Tivoli label)? It has a couple of interesting
1044 ideas that may apply to the following topics.
1046 1. Migration: Consider that when you need to restore a system, there
1047 may be pressure to hurry. If all the information for a single client
1048 can eventually end up on the same media (and in chronological order),
1049 the restore is facillitated by not having to search past information
1050 from other clients. ADSM has the concept of "client affinity" that
1051 may be associated with it's storage pools. It seems to me that this
1052 concept (as an optional feature) might fit in your architecture for
1055 ADSM also has the concept of defining one or more storage pools as
1056 "copy pools" (almost mirrors, but only in the sense of contents).
1057 These pools provide the ability to have duplicte data stored both
1058 onsite and offsite. The copy process can be scheduled to be handled
1059 by their storage manager during periods when there is no backup
1060 activity. Again, the migration process might be a place to consider
1061 implementing something like this.
1064 > It strikes me that it would be very nice to be able to do things
1066 > have the Job(s) backing up the machines run, and once they have all
1067 > completed, start a migration job to copy the data from disks Volumes
1069 > a tape library and then to offsite storage. Maybe this can already
1071 > done with some careful scheduling and Job prioritzation; the events
1072 > mechanism described below would probably make it very easy.
1074 This is the goal. In the first step (before events), you simply
1076 the Migration to tape later.
1078 2. Base jobs: In ADSM, each copy of each stored file is tracked in
1079 the database. Once a file (unique by path and metadata such as dates,
1080 size, ownership, etc.) is in a copy pool, no more copies are made. In
1081 other words, when you start ADSM, it begins like your concept of a
1082 base job. After that it is in the "incremental" mode. You can
1083 configure the number of "generations" of files to be retained, plus a
1084 retention date after which even old generations are purged. The
1085 database tracks the contents of media and projects the percentage of
1086 each volume that is valid. When the valid content of a volume drops
1087 below a configured percentage, the valid data are migrated to another
1088 volume and the old volume is marked as empty. Note, this requires
1089 ADSM to have an idea of the contents of a client, i.e. marking the
1090 database when an existing file was deleted, but this would solve your
1091 issue of restoring a client without restoring deleted files.
1093 This is pretty far from what bacula now does, but if you are going to
1094 rip things up for Base jobs,.....
1095 Also, the benefits of this are huge for very large shops, especially
1096 with media robots, but are a pain for shops with manual media
1100 > Base jobs sound pretty useful, but I'm not dying for them.
1102 Nobody is dying for them, but when you see what it does, you will die
1105 3. Restoring deleted files: Since I think my comments in (2) above
1106 have low probability of implementation, I'll also suggest that you
1107 could approach the issue of deleted files by a mechanism of having the
1108 fd report to the dir, a list of all files on the client for every
1109 backup job. The dir could note in the database entry for each file
1110 the date that the file was seen. Then if a restore as of date X takes
1111 place, only files that exist from before X until after X would be
1112 restored. Probably the major cost here is the extra date container in
1113 each row of the files table.
1115 Thanks for "listening". I hope some of this helps. If you want to
1116 contact me, please send me an email - I read some but not all of the
1117 mailing list traffic and might miss a reply there.
1119 Please accept my compliments for bacula. It is doing a great job for
1120 me!! I sympathize with you in the need to wrestle with excelence in
1121 execution vs. excelence in feature inclusion.
1126 ==============================
1129 - Design at hierarchial storage for Bacula. Migration and Clone.
1130 - Implement FSM (File System Modules).
1131 - Audit M_ error codes to ensure they are correct and consistent.
1132 - Add variable break characters to lex analyzer.
1133 Either a bit mask or a string of chars so that
1134 the caller can change the break characters.
1135 - Make a single T_BREAK to replace T_COMMA, etc.
1136 - Ensure that File daemon and Storage daemon can
1137 continue a save if the Director goes down (this
1138 is NOT currently the case). Must detect socket error,
1139 buffer messages for later.
1140 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
1141 - Add ability to backup to two Storage devices (two SD sessions) at
1142 the same time -- e.g. onsite, offsite.
1143 - Add the ability to consolidate old backup sets (basically do a restore
1144 to tape and appropriately update the catalog). Compress Volume sets.
1145 Might need to spool via file is only one drive is available.
1146 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
1147 someway to do so with every volume that has less than x% valid
1151 Migration: Move a backup from one Volume to another
1152 Clone: Copy a backup -- two Volumes
1154 Bacula Migration is based on Jobs (apparently Networker is file by file).
1156 Migration triggered by:
1160 Highwater mark (keep total size)
1165 ======================================================
1167 It is somewhat like a Full save becomes an incremental since
1168 the Base job (or jobs) plus other non-base files.
1170 - A Base backup is same as Full backup, just different type.
1171 - New BaseFiles table that contains:
1173 BaseJobId - Base JobId referenced for this FileId (needed ???)
1174 JobId - JobId currently running
1175 FileId - File not backed up, exists in Base Job
1176 FileIndex - FileIndex from Base Job.
1177 i.e. for each base file that exists but is not saved because
1178 it has not changed, the File daemon sends the JobId, BaseId,
1179 FileId, FileIndex back to the Director who creates the DB entry.
1180 - To initiate a Base save, the Director sends the FD
1181 the FileId, and full filename for each file in the Base.
1182 - When the FD finds a Base file, he requests the Director to
1183 send him the full File entry (stat packet plus MD5), or
1184 conversely, the FD sends it to the Director and the Director
1185 says yes or no. This can be quite rapid if the FileId is kept
1186 by the FD for each Base Filename.
1187 - It is probably better to have the comparison done by the FD
1188 despite the fact that the File entry must be sent across the
1190 - An alternative would be to send the FD the whole File entry
1191 from the start. The disadvantage is that it requires a lot of
1192 space. The advantage is that it requires less communications
1194 - The Job record must be updated to indicate that one or more
1196 - At end of Job, FD returns:
1197 1. Count of base files/bytes not written to tape (i.e. matches)
1198 2. Count of base file that were saved i.e. had changed.
1199 - No tape record would be written for a Base file that matches, in the
1200 same way that no tape record is written for Incremental jobs where
1201 the file is not saved because it is unchanged.
1202 - On a restore, all the Base file records must explicitly be
1203 found from the BaseFile tape. I.e. for each Full save that is marked
1204 to have one or more Base Jobs, search the BaseFile for all occurrences
1206 - An optimization might be to make the BaseFile have:
1212 This would avoid the need to explicitly fetch each File record for
1213 the Base job. The Base Job record will be fetched to get the
1214 VolSessionId and VolSessionTime.
1215 =========================================================
1218 ==========================================================
1220 For each Incremental job that is run, there may be files that
1221 were found but not saved because they were locked (this applies
1222 only to Windows). Such a system could send back to the Director
1223 a list of Unsaved files.
1225 - New UnSavedFiles table that contains:
1229 - Then in the next Incremental job, the list of Unsaved Files will be
1230 feed to the FD, who will ensure that they are explicitly chosen even
1231 if standard date/time check would not have selected them.
1232 =============================================================
1236 Multiple drive autochanger data: see Alan Brown
1237 mtx -f xxx unloadStorage Element 1 is Already Full(drive 0 was empty)
1238 Unloading Data Transfer Element into Storage Element 1...source Element
1239 Address 480 is Empty
1241 (drive 0 was empty and so was slot 1)
1242 > mtx -f xxx load 15 0
1243 no response, just returns to the command prompt when complete.
1244 > mtx -f xxx status Storage Changer /dev/changer:2 Drives, 60 Slots ( 2 Import/Export )
1245 Data Transfer Element 0:Full (Storage Element 15 Loaded):VolumeTag = HX001
1246 Data Transfer Element 1:Empty
1247 Storage Element 1:Empty
1248 Storage Element 2:Full :VolumeTag=HX002
1249 Storage Element 3:Full :VolumeTag=HX003
1250 Storage Element 4:Full :VolumeTag=HX004
1251 Storage Element 5:Full :VolumeTag=HX005
1252 Storage Element 6:Full :VolumeTag=HX006
1253 Storage Element 7:Full :VolumeTag=HX007
1254 Storage Element 8:Full :VolumeTag=HX008
1255 Storage Element 9:Full :VolumeTag=HX009
1256 Storage Element 10:Full :VolumeTag=HX010
1257 Storage Element 11:Empty
1258 Storage Element 12:Empty
1259 Storage Element 13:Empty
1260 Storage Element 14:Empty
1261 Storage Element 15:Empty
1262 Storage Element 16:Empty....
1263 Storage Element 28:Empty
1264 Storage Element 29:Full :VolumeTag=CLNU01L1
1265 Storage Element 30:Empty....
1266 Storage Element 57:Empty
1267 Storage Element 58:Full :VolumeTag=NEX261L2
1268 Storage Element 59 IMPORT/EXPORT:Empty
1269 Storage Element 60 IMPORT/EXPORT:Empty
1271 Unloading Data Transfer Element into Storage Element 15...done
1273 (just to verify it remembers where it came from, however it can be
1274 overrriden with mtx unload {slotnumber} to go to any storage slot.)
1276 There needs to be a table of drive # to devices somewhere - If there are
1277 multiple changers or drives there may not be a 1:1 correspondance between
1278 changer drive number and system device name - and depending on the way the
1279 drives are hooked up to scsi busses, they may not be linearly numbered
1280 from an offset point either.something like
1282 Autochanger drives = 2
1283 Autochanger drive 0 = /dev/nst1
1284 Autochanger drive 1 = /dev/nst2
1285 IMHO, it would be _safest_ to use explicit mtx unload commands at all
1286 times, not just for multidrive changers. For a 1 drive changer, that's
1292 MTX's manpage (1.2.15):
1293 unload [<slotnum>] [ <drivenum> ]
1294 Unloads media from drive <drivenum> into slot
1295 <slotnum>. If <drivenum> is omitted, defaults to
1296 drive 0 (as do all commands). If <slotnum> is
1297 omitted, defaults to the slot that the drive was
1298 loaded from. Note that there's currently no way
1299 to say 'unload drive 1's media to the slot it
1300 came from', other than to explicitly use that
1301 slot number as the destination.AB
1307 undef# camcontrol devlist
1308 <WANGTEK 51000 SCSI M74H 12B3> at scbus0 target 2 lun 0 (pass0,sa0)
1309 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 0 (pass1,sa1)
1310 <ARCHIVE 4586XX 28887-XXX 4BGD> at scbus0 target 4 lun 1 (pass2)
1312 tapeinfo -f /dev/sg0 with a bad tape in drive 1:
1313 [kern@rufus mtx-1.2.17kes]$ ./tapeinfo -f /dev/sg0
1314 Product Type: Tape Drive
1316 Product ID: 'C5713A '
1318 Attached Changer: No
1319 TapeAlert[3]: Hard Error: Uncorrectable read/write error.
1320 TapeAlert[20]: Clean Now: The tape drive neads cleaning NOW.
1327 Medium Type: Not Loaded
1330 DataCompEnabled: yes
1331 DataCompCapable: yes
1332 DataDeCompEnabled: yes
1339 Handling removable disks
1341 From: Karl Cunningham <karlc@keckec.com>
1343 My backups are only to hard disk these days, in removable bays. This is my
1344 idea of how a backup to hard disk would work more smoothly. Some of these
1345 things Bacula does already, but I mention them for completeness. If others
1346 have better ways to do this, I'd like to hear about it.
1348 1. Accommodate several disks, rotated similar to how tapes are. Identified
1349 by partition volume ID or perhaps by the name of a subdirectory.
1350 2. Abort & notify the admin if the wrong disk is in the bay.
1351 3. Write backups to different subdirectories for each machine to be backed
1353 4. Volumes (files) get created as needed in the proper subdirectory, one
1355 5. When a disk is recycled, remove or zero all old backup files. This is
1356 important as the disk being recycled may be close to full. This may be
1357 better done manually since the backup files for many machines may be
1358 scattered in many subdirectories.
1363 - Make sure that all do_prompt() calls in Dir check for
1364 -1 (error) and -2 (cancel) returns.
1365 - Fix foreach_jcr() to have free_jcr() inside next().
1366 jcr=jcr_walk_start();
1367 for ( ; jcr; (jcr=jcr_walk_next(jcr)) )
1370 - A Volume taken from Scratch should take on the retention period
1372 - Correct doc for Maximum Changer Wait (and others) accepting only
1374 - Implement status that shows why a job is being held in reserve, or
1375 rather why none of the drives are suitable.
1376 - Implement a way to disable a drive (so you can use the second
1377 drive of an autochanger, and the first one will not be used or
1379 - Make sure Maximum Volumes is respected in Pools when adding
1380 Volumes (e.g. when pulling a Scratch volume).
1381 - Keep same dcr when switching device ...
1382 - Implement code that makes the Dir aware that a drive is an
1383 autochanger (so the user doesn't need to use the Autochanger = yes
1385 - Make catalog respect ACL.
1386 - Add recycle count to Media record.
1387 - Add initial write date to Media record.
1388 - Fix store_yesno to be store_bitmask.
1389 --- create_file.c.orig Fri Jul 8 12:13:05 2005
1390 +++ create_file.c Fri Jul 8 12:13:07 2005
1392 attr->ofname, be.strerror());
1395 + } else if(S_ISSOCK(attr->statp.st_mode)) {
1396 + Dmsg1(200, "Skipping socket: %s\n", attr->ofname);
1398 Dmsg1(200, "Restore node: %s\n", attr->ofname);
1399 if (mknod(attr->ofname, attr->statp.st_mode, attr->statp.st_rdev) != 0 && errno != EEXIST) {
1400 - Add true/false to conf same as yes/no
1401 - Reserve blocks other restore jobs when first cannot connect to SD.
1402 - Fix Maximum Changer Wait, Maximum Open Wait, Maximum Rewind Wait to
1403 accept time qualifiers.
1404 - Does ClientRunAfterJob fail the job on a bad return code?