4 Documentation to do: (a little bit at a time)
5 - Document running a test version.
6 - Document query file format.
7 - Document static linking
8 - Document how to automatically backup all local partitions
9 - Document problems with Verify and pruning.
10 - Document how to use multiple databases.
13 Testing to do: (painful)
14 - that console command line options work
15 - blocksize recognition code.
16 - multiple simultaneous Volumes
19 - Implement TCP/IP connection for MySQL
20 - Make some way so that if a machine is skipped because it is not up
21 that Bacula will continue retrying for a specified period of time --
23 - Figure out some way to estimate output size and to avoid splitting
24 a backup across two Volumes -- this could be useful for writing CDROMs
25 where you really prefer not to have it split -- not serious.
26 - Add chflags() code for FreeBSD file flags
27 - Add RunBeforeJob and RunAfterJob to the Client program.
28 - Have SD compute MD5 or SHA1 and compare to what FD computes.
29 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
30 actual data on the Volume and compare it.
31 - Implement FileOptions (see end of this document)
32 - Implement Bacula plugins -- design API
33 - Make bcopy read through bad tape records.
34 - Fix read_record to handle multiple sessions.
35 - Program files (i.e. execute a program to read/write files).
36 Pass read date of last backup, size of file last time.
37 - Add Signature type to File DB record.
38 - Make Restore report an error if FD or SD term codes are not OK.
39 - CD into subdirectory when open()ing files for backup to
40 speed up things. Test with testfind().
41 - Add prefixlinks to where or not where absolute links to FD.
42 - Look at handling <> in smtp doesn't work with exim.
43 - Priority job to go to top of list.
44 - Implement Bar code handling
45 - Why is catreq.c:111 Find vol called twice for a job?
46 - Find out why Full saves run slower and slower (hashing?)
47 - Why are save/restore of device different sizes (sparse?) Yup! Fix it.
48 - Implement some way for the Console to dynamically create a job.
49 - Restore to a particular time -- e.g. before date, after date.
50 - Implement disk spooling
51 - Implement finer multiprocessing options.
52 - Solaris -I on tar for include list
53 - Enable avoid backing up archive device (findlib/find_one.c:128)
54 - Need a verbose mode in restore, perhaps to bsr.
55 - bscan without -v is too quiet -- perhaps show jobs.
56 - Add code to reject whole blocks if not wanted on restore.
57 - Start working on Base jobs.
58 - Make sure the MaxVolFiles is fully implemented in SD
59 - Check if both CatalogFiles and UseCatalog are set to SD.
60 - Check if we can increase Bacula FD priorty in Win2000
61 - Need return status on read_cb() from read_records(). Need multiple
62 records -- one per Job, maybe a JCR or some other structure with
64 - Figure out how to do a bare metal Windows restore
65 - Put system type returned by FD into catalog.
66 - Need to specify MaximumConcurrentJobs in the Job resource.
67 - Possibly add email to Watchdog if drive is unmounted too
68 long and a job is waiting on the drive.
69 - Strip trailing slashes from Include directory names in the FD.
70 - Use read_record.c in SD code.
71 - Why don't we get an error message from Win32 FD when bootstrap
72 file cannot be created for restore command?
73 - When Marking a file in Restore that is a hard link, also
74 mark the link so that the data will be reloaded.
75 - Restore program that errors in SD due to no tape reports
76 OK incorrectly in output.
77 - After unmount, if restore job started, ask to mount.
78 - Convert all %x substitution variables, which are hard to remember
79 and read to %(variable-name). Idea from TMDA.
80 - Add JobLevel in FD status (but make sure it is defined).
81 - Make Pool resource handle Counter resources.
82 - Remove NextId for SQLite. Optimize.
83 - Strip trailing / from Include
84 - Move all SQL statements into a single location.
85 - Add UA rc and history files.
86 - put termcap (used by console) in ./configure and
87 allow -with-termcap-dir.
88 - Enhance time and size scanning routines.
89 - Fix Autoprune for Volumes to respect need for full save.
90 - Fix Win32 config file definition name on /install
91 - No READLINE_SRC if found in alternate directory.
92 - Add Client FS/OS id (Linux, Win95/98, ...).
93 - Test a second language e.g. french.
94 - Compare tape to Client files (attributes, or attributes and data)
95 - Make all database Ids 64 bit.
96 - Write an applet for Linux.
97 - Add estimate to Console commands
98 - Find solution to blank filename (i.e. path only) problem.
99 - Implement new daemon communications protocol.
100 - Remove PoolId from Job table, it exists in Media.
101 - Allow console commands to detach or run in background.
102 - Fix status delay on storage daemon during rewind.
103 - Add SD message variables to control operator wait time
104 - Maximum Operator Wait
105 - Minimum Message Interval
106 - Maximum Message Interval
107 - Send Operator message when cannot read tape label.
108 - Verify level=Volume (scan only), level=Data (compare of data to file).
109 Verify level=Catalog, level=InitCatalog
111 - Add keyword search to show command in Console.
112 - Fix Win2000 error with no messages during startup.
113 - Events : tape has more than xxx bytes.
114 - Restrict characters permitted in a Resource name.
115 - Complete code in Bacula Resources -- this will permit
116 reading a new config file at any time.
117 - Handle ctl-c in Console
118 - Implement LabelTemplate (at least first cut).
119 - Implement script driven addition of File daemon to config files.
120 - Think about how to make Bacula work better with File (non-tape) archives.
121 - Write Unix emulator for Windows.
123 - Implement new serialize subroutines
124 send(socket, "string", &Vol, "uint32", &i, NULL)
125 - Audit all UA commands to ensure that we always prompt where possible.
126 - If ./btape is called without /dev, assume argument is a Storage resource name.
127 - Put memory utilization in Status output of each daemon
128 if full status requested or if some level of debug on.
129 - Make database type selectable by .conf files i.e. at runtime
130 - gethostbyname failure in bnet_connect() continues
131 generating errors -- should stop.
132 - Set flag for uname -a. Add to Volume label.
133 - Implement throttled work queue.
134 - Check for EOT at ENOSPC or EIO or ENXIO (unix Pc)
135 - Allow multiple Storage specifications (or multiple names on
136 a single Storage specification) in the Job record. Thus a job
137 can be backed up to a number of storage devices.
138 - Implement dump label to UA
139 - Concept of VolumeSet during restore which is a list
140 of Volume names needed.
141 - Restore files modified after date
142 - Restore file modified before date
143 - Emergency restore info:
145 - Backup working directory
147 - Restore -- do nothing but show what would happen
148 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
149 - Implement Restore FileSet=
150 - Create a protocol.h and protocol.c where all protocol messages
152 - If SD cannot open a drive, make it periodically retry.
153 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
154 - Timout a job or terminate if link goes down, or reopen link and query.
155 - Find general solution for sscanf size problems (as well
156 as sprintf. Do at run time?
157 - Concept of precious tapes (cannot be reused).
158 - Make bcopy copy with a single tape drive.
159 - Permit changing ownership during restore.
161 - Autolabel should be specified by DIR instead of SD.
162 - Find out how to get the system tape block limits, e.g.:
163 Apr 22 21:22:10 polymatou kernel: st1: Block limits 1 - 245760 bytes.
164 Apr 22 21:22:10 polymatou kernel: st0: Block limits 2 - 16777214 bytes.
167 - AutoScan (check checksum of tape)
168 - Format command = "format /dev/nst0"
172 - Seek resolution (usually corresponds to buffer size)
173 - EODErrorCode=ENOSPC or code
174 - Partial Read error code
175 - Partial write error code
176 - Nonformatted read error
177 - Nonformatted write error
178 - WriteProtected error
182 - IgnoreCloseErrors=yes
192 - FD sends unsaved file list to Director at end of job (see
194 - File daemon should build list of files skipped, and then
195 at end of save retry and report any errors.
196 - Write a Storage daemon that uses pipes and
197 standard Unix programs to write to the tape.
199 - Need something that monitors the JCR queue and
200 times out jobs by asking the deamons where they are.
201 - Enhance Jmsg code to permit buffering and saving to disk.
202 - device driver = "xxxx" for drives.
203 - restart: paranoid: read label fsf to
204 eom read append block, and go
205 super-paranoid: read label, read all files
206 in between, read append block, and go
207 verify: backspace, read append block, and go
208 permissive: same as above but frees drive
209 if tape is not valid.
211 - Ensure that /dev/null works
212 - Need report class for messages. Perhaps
213 report resource where report=group of messages
214 - enhance scan_attrib and rename scan_jobtype, and
215 fill in code for "since" option
216 - Director needs a time after which the report status is sent
217 anyway -- or better yet, a retry time for the job.
218 Don't reschedule a job if previous incarnation is still running.
219 - Some way to automatically backup everything is needed????
220 - Need a structure for pending actions:
222 - termination status (part of buffered msgs?)
223 - Concept of grouping Storage devices and job can use
224 any of a number of devices
226 Read, Write, Clean, Delete
227 - Login to Bacula; Bacula users with different permissions:
228 owner, group, user, quotas
229 - Store info on each file system type (probably in the job header on tape.
230 This could be the output of df; or perhaps some sort of /etc/mtab record.
233 - Design at hierarchial storage for Bacula.
234 - Implement FSM (File System Modules).
235 - Identify unchanged or "system" files and save them to a
236 special tape thus removing them from the standard
237 backup FileSet -- BASE backup.
238 - Turn virutally all sprintfs into snprintfs.
239 - Heartbeat between daemons.
240 - Audit M_ error codes to ensure they are correct and consistent.
241 - Add variable break characters to lex analyzer.
242 Either a bit mask or a string of chars so that
243 the caller can change the break characters.
244 - Make a single T_BREAK to replace T_COMMA, etc.
245 - Ensure that File daemon and Storage daemon can
246 continue a save if the Director goes down (this
247 is NOT currently the case). Must detect socket error,
248 buffer messages for later.
249 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
250 - Add ability to backup to two Storage devices (two SD sessions) at
251 the same time -- e.g. onsite, offsite.
252 - Add the ability to consolidate old backup sets (basically do a restore
253 to tape and appropriately update the catalog). Compress Volume sets.
254 Might need to spool via file is only one drive is available.
255 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
256 someway to do so with every volume that has less than x% valid
261 Bacula Projects Roadmap
263 last update 5 January 2003
265 Item 1: Multiple simultaneous Jobs. (done)
266 Done -- Restore part needs better implementation to work correctly
268 What: Permit multiple simultaneous jobs in Bacula.
270 Why: An enterprise level solution needs to go fast without the
271 need for the system administrator to carefully tweak
272 timing. Based on the benchmarks, during a full
273 backup, NetWorker typically hit 10 times the bandwidth to
274 the tape compared to Bacula--largely. This is probably due to
275 running parallel jobs and multi-threaded filling of buffers
276 and writing them to tape. This should also make things work
277 better when you have a mix of fast and slow machines backing
280 Notes: Bacula was designed to run multiple simultaneous jobs. Thus
281 implementing this is a matter of some small cleanups and
285 Item 2: Make the Storage daemon use intermediate file storage to buffer data.
286 Deferred -- not necessary yet.
288 What: If data is coming into the SD too fast, buffer it to
289 disk if the user has configured this option.
291 Why: This would be nice, especially if it more or less falls out
292 when implementing (1) above. If not, it probably should not
293 be given a high priority because fundamentally the backup time
294 is limited by the tape bandwidth. Even though you may finish a
295 client job quicker by spilling to disk, you still have to
296 eventually get it onto tape. If intermediate disk buffering
297 allows us to improve write bandwidth to tape, it may make
300 Notes: Whether or not this is implemented will depend upon performance
301 testing after item 1 is implemented.
304 Item 3: Write the bscan program -- also write a bcopy program.
307 What: Write a program that reads a Bacula tape and puts all the
308 appropriate data into the catalog. This allows recovery
309 from a tape that is no longer in the database, or it allows
310 re-creation of a database if lost.
312 Why: This is a fundamental robustness and disaster recovery tool
313 which will increase the comfort level of a sysadmin
314 considering adopting Bacula.
316 Notes: A skeleton of this program already exists, but much work
317 needs to be done. Implementing this will also make apparent
318 any deficiencies in the current Bacula tape format.
321 Item 4: Implement Base jobs.
323 What: A base job is sort of like a Full save except that you
324 will want the FileSet to contain only files that are unlikely
325 to change in the future (i.e. a snapshot of most of your
326 system after installing it). After the base job has been run,
327 when you are doing a Full save, you can specify to exclude
328 all files saved by the base job that have not been modified.
330 Why: This is something none of the competition does, as far as we know
331 (except BackupPC, which is a Perl program that saves to disk
332 only). It is big win for the user, it makes Bacula stand out
333 as offering a unique optimization that immediately saves time
336 Notes: Big savings in tape usage. Will require more resources because
337 the e. DIR must send FD a list of files/attribs, and the FD must
338 search the list and compare it for each file to be saved.
341 Item 5: Implement Label templates
343 What: This is a mechanism whereby Bacula can automatically create
344 a tape label for new tapes according to a detailed specification
345 provided by the user.
347 Why: It is a major convenience item for folks who use automated label
350 Notes: Bacula already has a working form of automatic tape label
351 creation, but it is very crude. The design for the complete
352 tape labeling project is already documented in the manual.
355 Item 6: Write a regression script.
358 What: This is an automatic script that runs and tests as many features
359 of Bacula as possible. The output is compared to previous
360 versions of Bacula and any differences are reported.
362 Why: This is an enormous help in preventing introduction of new
363 errors in parts of the program that already work correctly.
365 Notes: This probably should be ranked higher, it's something the typical
366 user doesn't see. Depending on how it's implemented, it may
367 make sense to defer it until the archival tape format and
368 user interface mature.
371 Item 7: GUI for interactive restore
372 Item 8: GUI for interactive backup
374 What: The current interactive restore is implemented with a tty
375 interface. It would be much nicer to be able to "see" the
376 list of files backed up in typical GUI tree format.
377 The same mechanism could also be used for creating
378 ad-hoc backup FileSets (item 8).
380 Why: Ease of use -- especially for the end user.
382 Notes: Rather than implementing in Gtk, we probably should go directly
383 for a Browser implementation, even if doing so meant the
384 capability wouldn't be available until much later. Not only
385 is there the question of Windows sites, most
386 Solaris/HP/IRIX, etc, shops can't currently run Gtk programs
387 without installing lots of stuff admins are very wary about.
388 Real sysadmins will always use the command line anyway, and
389 the user who's doing an interactive restore or backup of his
390 own files will in most cases be on a Windows machine running
394 Item 9: Add SSL to daemon communications.
396 What: This provides for secure communications between the daemons.
398 Why: This would allow doing backup across the Internet without
399 privacy concerns (or with much less concern).
401 Notes: The vast majority of near term potential users will be backing up
402 a single site over a LAN and, correctly or not, they probably
403 won't be concerned with security, at least not enough to go to
404 the trouble to set up keys, etc. to screw things down. We suspect
405 that many users genuinely interested in multi-site backup
406 already run some form of VPN software in their internetwork
407 connections, and are willing to delegate security to that layer.
410 Item 10: Define definitive tape format.
413 What: Define that definitive tape format that will not change
414 for the next millennium.
416 Why: Stability, security.
418 Notes: See notes for item 11 below.
421 Item 11: New daemon communication protocol.
423 What: The current daemon to daemon protocol is basically an ASCII
424 printf() and sending the buffer. On the receiving end, the
425 buffer is sscanf()ed to unpack it. The new scheme would
426 be a binary format that allows quick packing and unpacking
427 of any data type with named fields.
429 Why: Using binary packing would be faster. Named fields will permit
430 error checking to ensure that what is sent is what the
431 receiver really wants.
433 Notes: These are internal improvements in the interest of the
434 long-term stability and evolution of the program. On the one
435 hand, the sooner they're done, the less code we have to rip
436 up when the time comes to install them. On the other hand, they
437 don't bring an immediately perceptible benefit to potential
438 users. Item 10 and possibly item 11 should be deferred until Bacula
439 is well established with a growing user community more or
440 less happy with the feature set. At that time, it will make a
441 good "next generation" upgrade in the interest of data
447 ======================================================
449 It is somewhat like a Full save becomes an incremental since
450 the Base job (or jobs) plus other non-base files.
452 - New BaseFile table that contains:
453 JobId, BaseJobId, FileId (from Base).
454 i.e. for each base file that exists but is not saved because
455 it has not changed, the File daemon sends the JobId, BaseId,
456 and FileId back to the Director who creates the DB entry.
457 - To initiate a Base save, the Director sends the FD
458 the FileId, and full filename for each file in the Base.
459 - When the FD finds a Base file, he requests the Director to
460 send him the full File entry (stat packet plus MD5), or
461 conversely, the FD sends it to the Director and the Director
462 says yes or no. This can be quite rapid if the FileId is kept
463 by the FD for each Base Filename.
464 - It is probably better to have the comparison done by the FD
465 despite the fact that the File entry must be sent across the
467 - An alternative would be to send the FD the whole File entry
468 from the start. The disadvantage is that it requires a lot of
469 space. The advantage is that it requires less communications
471 - The Job record must be updated to indicate that one or more
473 - At end of Job, FD returns:
474 1. Count of base files/bytes not written to tape (i.e. matches)
475 2. Count of base file that were saved i.e. had changed.
476 - No tape record would be written for a Base file that matches, in the
477 same way that no tape record is written for Incremental jobs where
478 the file is not saved because it is unchanged.
479 - On a restore, all the Base file records must explicitly be
480 found from the BaseFile tape. I.e. for each Full save that is marked
481 to have one or more Base Jobs, search the BaseFile for all occurrences
483 - An optimization might be to make the BaseFile have:
489 This would avoid the need to explicitly fetch each File record for
490 the Base job. The Base Job record will be fetched to get the
491 VolSessionId and VolSessionTime.
492 =========================================================
494 ==========================================================
496 For each Incremental job that is run, there may be files that
497 were found but not saved because they were locked (this applies
498 only to Windows). Such a system could send back to the Director
499 a list of Unsaved files.
501 - New UnSavedFiles table that contains:
505 - Then in the next Incremental job, the list of Unsaved Files will be
506 feed to the FD, who will ensure that they are explicitly chosen even
507 if standard date/time check would not have selected them.
508 =============================================================
512 =============================================================
514 Request For Comments For File Backup Options
517 Subject: File Backup Options
520 A few days ago, a Bacula user who is backing up to file volumes and
521 using compression asked if it was possible to suppress compressing
522 all .gz files since it was a waste of CPU time. Although Bacula
523 currently permits using different options (compression, ...) on
524 a directory by directory basis, it cannot do it on a file by
525 file basis, which is clearly what was desired.
527 Proposed Implementation:
528 To solve this problem, I propose the following:
530 - Add a new Director resource type called FileOptions.
532 - The FileOptions resource will have records for all
533 options that can currently be specified on the Include record
534 (in a FileSet). Examples below.
536 - The FileOptions resource will permit an exclude option as well
537 as a number of additional options.
539 - The heart of the FileOptions resource is the ability to
540 supply any number of ApplyTo records which specify POSIX
541 regular expressions. These ApplyTo regular expressions are
542 applied to the fully qualified filename (path and all). If
543 one matches, then the FileOptions will be used.
545 - When an ApplyTo specification matches an included file, the
546 options specified in the FileOptions resource will override
547 the default options specified on the Include record.
549 - Include records will be modified to permit referencing one or
550 more FileOptions resources. The FileOptions will be used
551 in the order listed on the Include record and the first
552 one that matches will be applied.
554 - Options (or specifications) currently supplied on the Include
555 record will be deprecated (i.e. removed in a later version a
556 year or so from now).
558 - The Exclude record will be deprecated as the same functionality
559 can be obtained by using an Exclude = yes in the FileOptions.
562 The following records can appear in the FileOptions resource. An
563 asterisk preceding the name indicates a feature not currently
567 - Compression= (GZIP, ...)
568 - Signature= (MD5, SHA1, ...)
570 - OneFs= (yes/no) - remain on one filesystem
571 - Recurse= (yes/no) - recurse into subdirectories
572 - Sparse= (yes/no) - do sparse file backup
573 - *Exclude= (yes/no) - exclude file from being saved
574 - *Reader= (filename) - external read (backup) program
575 - *Plugin= (filename) - read/write plugin module
578 - verify= (ipnougsamc5) - verify options
581 - replace= (always/ifnewer/ifolder/never) - replace options currently
583 - *Writer= (filename) - external write (restore) program
587 Currently options specifying compression, MD5 signatures, recursion,
588 ... of a FileSet are supplied on the Include record. These will now
589 all be collected into a FileOptions resource, which will be
590 specified on the Include in place of the options. Multiple FileOptions
591 may be specified. Since the FileOptions contain regular expressions
592 that are applied to the full filename, this will give the ability
593 to specify backup options on a file by file basis to whatever level
602 Include = compression=GZIP signature=MD5 {
611 Include = FileOptions=Opts {
622 That's a lot more to do the same thing, but it gives the ability to
623 apply options on a file by file basis. For example, suppose you
624 want to compress all files but not any file with extensions .gz or .Z.
625 You could do so as follows:
629 Include = FileOptions=NoCompress FileOptions=Opts {
637 ApplyTo = /*.?*/ # matches all files
642 # Note multiple ApplyTos are ORed
643 ApplyTo = /*.gz/ # matches .gz files */
644 ApplyTo = /*.Z/ # matches .Z files */
647 Now, since the NoCompress FileOptions is specified first on the
648 Include line, any *.gz or *.Z file will have an MD5 signature computed,
649 but will not be compressed. For all other files, the NoCompress will not
650 match, so the Opts options will be used which will include GZIP
654 - Is it necessary to provide some means of ANDing regular expressions
655 and negation? (not currently planned)
657 e.g. ApplyTo = /*.gz/ && !/big.gz/
659 - I see that Networker has a "null" module which, if specified, does not
660 backup the file, but does make an record of the file in the catalog
661 so that the catalog will reflect an exact picture of the filesystem.
662 The result is that the file can be "seen" when "browsing" the save
663 sets, but it cannot be restored.
665 Is this really useful? Should it be implemented in Bacula?
668 After implementing the above, the user will be able to specify
669 on a file by file basis (using regular expressions) what options are
670 applied for the backup.
671 ====================================
673 =========================================
674 Proposal by Bill Sellers
676 Return-Path: <w.a.sellers@larc.nasa.gov>
677 Received: from post.larc.nasa.gov (post.larc.nasa.gov [128.155.4.45]) by matou.sibbald.com (8.11.6/8.11.6) with ESMTP id h0ELUIm07622 for <kern@sibbald.com>; Tue, 14 Jan 2003 22:30:18 +0100
678 Received: from Baron.larc.nasa.gov (baron.larc.nasa.gov [128.155.40.132]) by post.larc.nasa.gov (pohub4.6) with ESMTP id QAA09768 for <kern@sibbald.com>; Tue, 14 Jan 2003 16:30:14 -0500 (EST)
679 Message-Id: <5.1.0.14.2.20030114153452.028dbae8@pop.larc.nasa.gov>
680 X-Sender: w.a.sellers@pop.larc.nasa.gov
681 X-Mailer: QUALCOMM Windows Eudora Version 5.1
682 Date: Tue, 14 Jan 2003 16:30:18 -0500
683 To: Kern Sibbald <kern@sibbald.com>
684 From: Bill Sellers <w.a.sellers@larc.nasa.gov>
685 Subject: Re: [Bacula-users] Bacula remote storage?
686 In-Reply-To: <1042565382.1845.177.camel@rufus>
687 References: <5.1.0.14.2.20030114113004.0293a210@pop.larc.nasa.gov> <5.1.0.14.2.20030113170650.028dad88@pop.larc.nasa.gov> <5.1.0.14.2.20030113170650.028dad88@pop.larc.nasa.gov> <5.1.0.14.2.20030114113004.0293a210@pop.larc.nasa.gov>
689 Content-Type: text/plain; charset="us-ascii"; format=flowed
690 X-Annoyance-Filter-Junk-Probability: 0
691 X-Annoyance-Filter-Classification: Mail
692 At 06:29 PM 1/14/2003 +0100, you wrote:
695 >Well, if you cannot put a Bacula client on the machine,
696 >then it is a big problem. If you know of some software
697 >that can do what you want, let me know, because I
698 >really just don't know how to do it -- at least not
704 We have been able to get Amanda to use the HSM as a storage
705 device. Someone here wrote a driver for Amanda. BUT, Amanda doesn't
706 handle Windows systems very well (or at all without Samba). So I am
707 looking for a backup system that has a Windows client. I really like the
708 Windows integration of Bacula.
710 From the command line, its rather trivial to move the data around. We use
713 tar cf - ./files | gzip -c | rsh hsm dd of=path/file.tgz
715 or if you use GNU tar:
717 tar czf hsm:path/file.tgz ./files
719 One idea for you to consider; Sendmail offers pipes in the aliases file;
720 (mailpipe: "|/usr/bin/vacation root") and Perl supports pipes in the
721 "open" statement (open FILE, "|/bin/nroff -man";) Could you could make a
722 pipe available, as a storage device? Then we could use any command that
723 handles stdin as a storage destination.
730 #Password is not used in rsh, but might be used in ftp.
731 Device = "| gzip -c | rsh hsm dd of=path/file.tgz"
738 Password = "foobar&-"
739 Device = "| ncftpput -c hsm /path/file.bacula"
743 >If you have some local storage available, you could
744 >use Bacula to backup to disk volumes, then use some
745 >other software (ftp, scp) to move them to the HSM
746 >machine. However, this is a bit kludgy.
749 It is, but maybe worth a try. Is there some function in Bacula to put
750 variables in filenames? i.e. backup.2003-01-15.root
757 w.a.sellers@larc.nasa.gov
759 ==============================================
760 The Project for the above
762 I finally realized that this is not at all
763 the same as reader/writer programs or plugins,
764 which are alternate ways of accessing the
765 files to be backed up. Rather, it is an alternate
766 form of storage device, and I have always planned
767 that Bacula should be able to handle all sorts
770 So, I propose the following phases:
772 1. OK from you to invest some time in testing
773 this as I implement it (requires that you
774 know how to download from the SourceForge
775 cvs -- which I imagine is a piece of cake
778 2. Dumb implementation by allowing a device to
779 be a fifo for write only.
780 Reason: easy to implement, proof of concept.
782 3. Try reading from fifo but with fixed block
784 Reason: proof of concept, easy to implement.
786 4. Extend reading from fifo (restores) to handle
788 Reason: requires some delicate low level coding
789 which could destabilize all of Bacula.
791 5. Implementation of above but to a program. E.g.
792 Device = "|program" (not full pipeline).
793 Reason: routines already exist, and program can
794 be a shell script which contains anything.
796 6. Add full pipeline as a possibility. E.g.
797 Device = "| gzip -c | rsh hsm dd of=path/file.tgz"
798 Reason: needs additional coding to implement full
799 pipeline (must fire off either a shell or all
800 programs and connect their pipes).
802 There are a good number of details in each step
803 that I have left out, but I will specify them at
804 every stage, and there may be a few changes as things
805 evolve. I expect that to get to stage 5 will take a
806 few weeks, and at that point, you will have
807 everything you need (just inside a script).
808 Stage 6 will probably take longer, but if this
809 project pleases you, what we do for 5 should
810 be adequate for some time.
814 =============================================
816 Done: (see kernsdone for more)
817 - Look into Pruning/purging problems or why there seem to
818 be so many files listed each night.
819 - Fix cancel in find_one -- need jcr.
820 - Cancel does not work for restore in FD.
821 - Write SetJobStatus() function so cancel status not lost.
822 - Add include list to end of chain in findlib
823 - Zap sd_auth_key after use
824 - Add Bar code reading capabilities (new mtx-changer)
825 - Figure out some way to automatically backup all local partitions
826 - Make hash table for linked files in findlib/find_one.c:161
828 - Rewrite find_one.c to use only pool_memory instead of
829 alloca and malloc (probably not necessary).
830 - Make sure btraceback goes into /sbin not sysconf directory.
831 - InitVerify is getting pruned and it shouldn't (document it)
832 - Make 1.28c release ??? NO do 1.29 directly
833 - Set timeout on opening fifo for save or restore (findlib)
834 - Document FIFO storage device.
835 - Document fifo and | and <
838 - Get correct error status from run_program or open_bpipe().
839 - Restrict permissions on File Volumes (now 0640).
840 - Umasked 022 daemons
841 - Fix restore of hard linked file.
842 - Figure out how to allow multiple simultaneous file Volumes on a single device.
843 - Implement multiple simultaneous file Volumes on a single device.
844 - Cleanup db_update_media and db_update_pool
845 - Flush all the daemon messages at the end of every job.
846 - Change stat1= fgets()!=NULL to stat1=fgest()==NULL; in
847 run_program -- bpipe.c