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.
11 - Add a section to the doc on Manual cycling of Volumes.
14 - Document default file backup
17 Testing to do: (painful)
18 - that ALL console command line options work and are always implemented
19 - blocksize recognition code.
20 - multiple simultaneous Volumes
23 - Do not ignore SIGCHLD
24 - Fix Bare Metal restore problem.
25 - Fix "access not allowed" for backup of files on WinXP.
26 - Document new transparent Console commands and wait command.
27 - Document new immediate File save capability.
28 - Document OPTIMIZE TABLE in MySQL
29 - Fix watchdog pthread crash on Win32
31 - Figure out how to use ssh or stunnel to protect Bacula communications.
34 - Make | and < work on FD side.
35 - Pass prefix_links to FD.
36 - Implement a M_SECURITY message class.
37 - Implement disk spooling. Two parts: 1. Spool to disk then
38 immediately to tape to speed up tape operations. 2. Spool to
39 disk only when the tape is full, then when a tape is hung move
41 - From Phil Stracchino:
42 It would probably be a per-client option, and would be called
43 something like, say, "Automatically purge obsoleted jobs". What it
44 would do is, when you successfully complete a Differential backup of a
45 client, it would automatically purge all Incremental backups for that
46 client that are rendered redundant by that Differential. Likewise,
47 when a Full backup on a client completed, it would automatically purge
48 all Differential and Incremental jobs obsoleted by that Full backup.
49 This would let people minimize the number of tapes they're keeping on
50 hand without having to master the art of retention times.
52 - Allow multiple Storage specifications (or multiple names on
53 a single Storage specification) in the Job record. Thus a job
54 can be backed up to a number of storage devices.
55 - Implement dump label to UA
56 - Add prefixlinks to where or not where absolute links to FD.
57 - Look at Python for a Bacula scripting language -- www.python.org
58 - Issue message to mount a new tape before the rewind.
59 - Simplified client job initiation for portables.
60 - If SD cannot open a drive, make it periodically retry.
61 - Implement LabelTemplate (at least first cut).
62 - Add more of the config info to the tape label.
63 - Implement a Mount Command and an Unmount Command where
64 the users could specify a system command to be performed
65 to do the mount, after which Bacula could attempt to
66 read the device. This is for Removeable media such as a CDROM.
67 - Most likely, this mount command would be invoked explicitly
68 by the user using the current Console "mount" and "unmount"
69 commands -- the Storage Daemon would do the right thing
70 depending on the exact nature of the device.
71 - As with tape drives, when Bacula wanted a new removable
72 disk mounted, it would unmount the old one, and send a message
73 to the user, who would then use "mount" as described above
74 once he had actually inserted the disk.
76 - Make some way so that if a machine is skipped because it is not up
77 that Bacula will continue retrying for a specified period of time --
79 - If tape is marked read-only, then try opening it read-only rather than
80 failing, and remember that it cannot be written.
81 - Refine SD waiting output:
82 Device is being positioned
83 > Device is being positioned for append
84 > Device is being positioned to file x
86 - Figure out some way to estimate output size and to avoid splitting
87 a backup across two Volumes -- this could be useful for writing CDROMs
88 where you really prefer not to have it split -- not serious.
89 - Add RunBeforeJob and RunAfterJob to the Client program.
90 - Have SD compute MD5 or SHA1 and compare to what FD computes.
91 - Make VolumeToCatalog calculate an MD5 or SHA1 from the
92 actual data on the Volume and compare it.
93 - Implement FileOptions (see end of this document)
94 - Implement Bacula plugins -- design API
95 - Make bcopy read through bad tape records.
96 - Fix read_record to handle multiple sessions.
97 - Program files (i.e. execute a program to read/write files).
98 Pass read date of last backup, size of file last time.
99 - Add Signature type to File DB record.
100 - Make Restore report an error if FD or SD term codes are not OK.
101 - CD into subdirectory when open()ing files for backup to
102 speed up things. Test with testfind().
103 - Priority job to go to top of list.
104 - Find out why Full saves run slower and slower (hashing?)
105 - Why are save/restore of device different sizes (sparse?) Yup! Fix it.
106 - Implement some way for the Console to dynamically create a job.
107 - Restore to a particular time -- e.g. before date, after date.
108 - Solaris -I on tar for include list
109 - Prohibit backing up archive device (findlib/find_one.c:128)
110 - Need a verbose mode in restore, perhaps to bsr.
111 - bscan without -v is too quiet -- perhaps show jobs.
112 - Add code to reject whole blocks if not wanted on restore.
113 - Start working on Base jobs.
114 - Check if we can increase Bacula FD priorty in Win2000
115 - Make sure the MaxVolFiles is fully implemented in SD
116 - Check if both CatalogFiles and UseCatalog are set to SD.
117 - Need return status on read_cb() from read_records(). Need multiple
118 records -- one per Job, maybe a JCR or some other structure with
119 a block and a record.
120 - Figure out how to do a bare metal Windows restore
121 - Put system type returned by FD into catalog.
122 - Possibly add email to Watchdog if drive is unmounted too
123 long and a job is waiting on the drive.
124 - Use read_record.c in SD code.
125 - Why don't we get an error message from Win32 FD when bootstrap
126 file cannot be created for restore command?
127 - When Marking a file in Restore that is a hard link, also
128 mark the link so that the data will be reloaded.
129 - Restore program that errors in SD due to no tape reports
130 OK incorrectly in output.
131 - After unmount, if restore job started, ask to mount.
132 - Convert all %x substitution variables, which are hard to remember
133 and read to %(variable-name). Idea from TMDA.
134 - Add JobLevel in FD status (but make sure it is defined).
135 - Make Pool resource handle Counter resources.
136 - Remove NextId for SQLite. Optimize.
137 - Move all SQL statements into a single location.
138 - Add UA rc and history files.
139 - put termcap (used by console) in ./configure and
140 allow -with-termcap-dir.
141 - Enhance time and size scanning routines.
142 - Fix Autoprune for Volumes to respect need for full save.
143 - Fix Win32 config file definition name on /install
144 - No READLINE_SRC if found in alternate directory.
145 - Test a second language e.g. french.
146 - Compare tape to Client files (attributes, or attributes and data)
147 - Make all database Ids 64 bit.
148 - Write an applet for Linux.
149 - Add estimate to Console commands
150 - Find solution to blank filename (i.e. path only) problem.
151 - Implement new daemon communications protocol.
152 - Remove PoolId from Job table, it exists in Media.
153 - Allow console commands to detach or run in background.
154 - Fix status delay on storage daemon during rewind.
155 - Add SD message variables to control operator wait time
156 - Maximum Operator Wait
157 - Minimum Message Interval
158 - Maximum Message Interval
159 - Send Operator message when cannot read tape label.
160 - Verify level=Volume (scan only), level=Data (compare of data to file).
161 Verify level=Catalog, level=InitCatalog
163 - Add keyword search to show command in Console.
164 - Fix Win2000 error with no messages during startup.
165 - Events : tape has more than xxx bytes.
166 - Restrict characters permitted in a Resource name.
167 - Complete code in Bacula Resources -- this will permit
168 reading a new config file at any time.
169 - Handle ctl-c in Console
170 - Implement script driven addition of File daemon to config files.
171 - Think about how to make Bacula work better with File (non-tape) archives.
172 - Write Unix emulator for Windows.
173 - Implement new serialize subroutines
174 send(socket, "string", &Vol, "uint32", &i, NULL)
175 - Audit all UA commands to ensure that we always prompt where possible.
176 - If ./btape is called without /dev, assume argument is a Storage resource name.
177 - Put memory utilization in Status output of each daemon
178 if full status requested or if some level of debug on.
179 - Make database type selectable by .conf files i.e. at runtime
180 - Set flag for uname -a. Add to Volume label.
181 - Implement throttled work queue.
182 - Check for EOT at ENOSPC or EIO or ENXIO (unix Pc)
183 - Restore files modified after date
184 - Restore file modified before date
185 - Emergency restore info:
187 - Backup working directory
189 - Restore -- do nothing but show what would happen
190 - SET LD_RUN_PATH=$HOME/mysql/lib/mysql
191 - Implement Restore FileSet=
192 - Create a protocol.h and protocol.c where all protocol messages
194 - Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
195 - Timout a job or terminate if link goes down, or reopen link and query.
196 - Find general solution for sscanf size problems (as well
197 as sprintf. Do at run time?
198 - Concept of precious tapes (cannot be reused).
199 - Make bcopy copy with a single tape drive.
200 - Permit changing ownership during restore.
202 - Autolabel should be specified by DIR instead of SD.
203 - Find out how to get the system tape block limits, e.g.:
204 Apr 22 21:22:10 polymatou kernel: st1: Block limits 1 - 245760 bytes.
205 Apr 22 21:22:10 polymatou kernel: st0: Block limits 2 - 16777214 bytes.
208 - AutoScan (check checksum of tape)
209 - Format command = "format /dev/nst0"
213 - Seek resolution (usually corresponds to buffer size)
214 - EODErrorCode=ENOSPC or code
215 - Partial Read error code
216 - Partial write error code
217 - Nonformatted read error
218 - Nonformatted write error
219 - WriteProtected error
223 - IgnoreCloseErrors=yes
233 - FD sends unsaved file list to Director at end of job (see
235 - File daemon should build list of files skipped, and then
236 at end of save retry and report any errors.
237 - Write a Storage daemon that uses pipes and
238 standard Unix programs to write to the tape.
240 - Need something that monitors the JCR queue and
241 times out jobs by asking the deamons where they are.
242 - Enhance Jmsg code to permit buffering and saving to disk.
243 - device driver = "xxxx" for drives.
244 - restart: paranoid: read label fsf to
245 eom read append block, and go
246 super-paranoid: read label, read all files
247 in between, read append block, and go
248 verify: backspace, read append block, and go
249 permissive: same as above but frees drive
250 if tape is not valid.
252 - Ensure that /dev/null works
253 - Need report class for messages. Perhaps
254 report resource where report=group of messages
255 - enhance scan_attrib and rename scan_jobtype, and
256 fill in code for "since" option
257 - Director needs a time after which the report status is sent
258 anyway -- or better yet, a retry time for the job.
259 Don't reschedule a job if previous incarnation is still running.
260 - Some way to automatically backup everything is needed????
261 - Need a structure for pending actions:
263 - termination status (part of buffered msgs?)
265 Read, Write, Clean, Delete
266 - Login to Bacula; Bacula users with different permissions:
267 owner, group, user, quotas
268 - Store info on each file system type (probably in the job header on tape.
269 This could be the output of df; or perhaps some sort of /etc/mtab record.
272 - Design at hierarchial storage for Bacula. Migration and Clone.
273 - Implement FSM (File System Modules).
274 - Identify unchanged or "system" files and save them to a
275 special tape thus removing them from the standard
276 backup FileSet -- BASE backup.
277 - Heartbeat between daemons.
278 - Audit M_ error codes to ensure they are correct and consistent.
279 - Add variable break characters to lex analyzer.
280 Either a bit mask or a string of chars so that
281 the caller can change the break characters.
282 - Make a single T_BREAK to replace T_COMMA, etc.
283 - Ensure that File daemon and Storage daemon can
284 continue a save if the Director goes down (this
285 is NOT currently the case). Must detect socket error,
286 buffer messages for later.
287 - Enhance time/duration input to allow multiple qualifiers e.g. 3d2h
288 - Add ability to backup to two Storage devices (two SD sessions) at
289 the same time -- e.g. onsite, offsite.
290 - Add the ability to consolidate old backup sets (basically do a restore
291 to tape and appropriately update the catalog). Compress Volume sets.
292 Might need to spool via file is only one drive is available.
293 - Compress or consolidate Volumes of old possibly deleted files. Perhaps
294 someway to do so with every volume that has less than x% valid
299 Bacula Projects Roadmap
301 last update 5 January 2003
303 Item 1: Multiple simultaneous Jobs. (done)
304 Done -- Restore part needs better implementation to work correctly
306 What: Permit multiple simultaneous jobs in Bacula.
308 Why: An enterprise level solution needs to go fast without the
309 need for the system administrator to carefully tweak
310 timing. Based on the benchmarks, during a full
311 backup, NetWorker typically hit 10 times the bandwidth to
312 the tape compared to Bacula--largely. This is probably due to
313 running parallel jobs and multi-threaded filling of buffers
314 and writing them to tape. This should also make things work
315 better when you have a mix of fast and slow machines backing
318 Notes: Bacula was designed to run multiple simultaneous jobs. Thus
319 implementing this is a matter of some small cleanups and
323 Item 2: Make the Storage daemon use intermediate file storage to buffer data.
324 Deferred -- not necessary yet.
326 What: If data is coming into the SD too fast, buffer it to
327 disk if the user has configured this option.
329 Why: This would be nice, especially if it more or less falls out
330 when implementing (1) above. If not, it probably should not
331 be given a high priority because fundamentally the backup time
332 is limited by the tape bandwidth. Even though you may finish a
333 client job quicker by spilling to disk, you still have to
334 eventually get it onto tape. If intermediate disk buffering
335 allows us to improve write bandwidth to tape, it may make
338 Notes: Whether or not this is implemented will depend upon performance
339 testing after item 1 is implemented.
342 Item 3: Write the bscan program -- also write a bcopy program.
345 What: Write a program that reads a Bacula tape and puts all the
346 appropriate data into the catalog. This allows recovery
347 from a tape that is no longer in the database, or it allows
348 re-creation of a database if lost.
350 Why: This is a fundamental robustness and disaster recovery tool
351 which will increase the comfort level of a sysadmin
352 considering adopting Bacula.
354 Notes: A skeleton of this program already exists, but much work
355 needs to be done. Implementing this will also make apparent
356 any deficiencies in the current Bacula tape format.
359 Item 4: Implement Base jobs.
361 What: A base job is sort of like a Full save except that you
362 will want the FileSet to contain only files that are unlikely
363 to change in the future (i.e. a snapshot of most of your
364 system after installing it). After the base job has been run,
365 when you are doing a Full save, you can specify to exclude
366 all files saved by the base job that have not been modified.
368 Why: This is something none of the competition does, as far as we know
369 (except BackupPC, which is a Perl program that saves to disk
370 only). It is big win for the user, it makes Bacula stand out
371 as offering a unique optimization that immediately saves time
374 Notes: Big savings in tape usage. Will require more resources because
375 the e. DIR must send FD a list of files/attribs, and the FD must
376 search the list and compare it for each file to be saved.
379 Item 5: Implement Label templates
381 What: This is a mechanism whereby Bacula can automatically create
382 a tape label for new tapes according to a detailed specification
383 provided by the user.
385 Why: It is a major convenience item for folks who use automated label
388 Notes: Bacula already has a working form of automatic tape label
389 creation, but it is very crude. The design for the complete
390 tape labeling project is already documented in the manual.
393 Item 6: Write a regression script.
394 Done -- Continue to expand its testing.
396 What: This is an automatic script that runs and tests as many features
397 of Bacula as possible. The output is compared to previous
398 versions of Bacula and any differences are reported.
400 Why: This is an enormous help in preventing introduction of new
401 errors in parts of the program that already work correctly.
403 Notes: This probably should be ranked higher, it's something the typical
404 user doesn't see. Depending on how it's implemented, it may
405 make sense to defer it until the archival tape format and
406 user interface mature.
409 Item 7: GUI for interactive restore
410 Item 8: GUI for interactive backup
412 What: The current interactive restore is implemented with a tty
413 interface. It would be much nicer to be able to "see" the
414 list of files backed up in typical GUI tree format.
415 The same mechanism could also be used for creating
416 ad-hoc backup FileSets (item 8).
418 Why: Ease of use -- especially for the end user.
420 Notes: Rather than implementing in Gtk, we probably should go directly
421 for a Browser implementation, even if doing so meant the
422 capability wouldn't be available until much later. Not only
423 is there the question of Windows sites, most
424 Solaris/HP/IRIX, etc, shops can't currently run Gtk programs
425 without installing lots of stuff admins are very wary about.
426 Real sysadmins will always use the command line anyway, and
427 the user who's doing an interactive restore or backup of his
428 own files will in most cases be on a Windows machine running
432 Item 9: Add SSL to daemon communications.
434 What: This provides for secure communications between the daemons.
436 Why: This would allow doing backup across the Internet without
437 privacy concerns (or with much less concern).
439 Notes: The vast majority of near term potential users will be backing up
440 a single site over a LAN and, correctly or not, they probably
441 won't be concerned with security, at least not enough to go to
442 the trouble to set up keys, etc. to screw things down. We suspect
443 that many users genuinely interested in multi-site backup
444 already run some form of VPN software in their internetwork
445 connections, and are willing to delegate security to that layer.
448 Item 10: Define definitive tape format.
451 What: Define that definitive tape format that will not change
452 for the next millennium.
454 Why: Stability, security.
456 Notes: See notes for item 11 below.
459 Item 11: New daemon communication protocol.
461 What: The current daemon to daemon protocol is basically an ASCII
462 printf() and sending the buffer. On the receiving end, the
463 buffer is sscanf()ed to unpack it. The new scheme would
464 be a binary format that allows quick packing and unpacking
465 of any data type with named fields.
467 Why: Using binary packing would be faster. Named fields will permit
468 error checking to ensure that what is sent is what the
469 receiver really wants.
471 Notes: These are internal improvements in the interest of the
472 long-term stability and evolution of the program. On the one
473 hand, the sooner they're done, the less code we have to rip
474 up when the time comes to install them. On the other hand, they
475 don't bring an immediately perceptible benefit to potential
476 users. Item 10 and possibly item 11 should be deferred until Bacula
477 is well established with a growing user community more or
478 less happy with the feature set. At that time, it will make a
479 good "next generation" upgrade in the interest of data
485 ======================================================
487 It is somewhat like a Full save becomes an incremental since
488 the Base job (or jobs) plus other non-base files.
490 - New BaseFile table that contains:
491 JobId, BaseJobId, FileId (from Base).
492 i.e. for each base file that exists but is not saved because
493 it has not changed, the File daemon sends the JobId, BaseId,
494 and FileId back to the Director who creates the DB entry.
495 - To initiate a Base save, the Director sends the FD
496 the FileId, and full filename for each file in the Base.
497 - When the FD finds a Base file, he requests the Director to
498 send him the full File entry (stat packet plus MD5), or
499 conversely, the FD sends it to the Director and the Director
500 says yes or no. This can be quite rapid if the FileId is kept
501 by the FD for each Base Filename.
502 - It is probably better to have the comparison done by the FD
503 despite the fact that the File entry must be sent across the
505 - An alternative would be to send the FD the whole File entry
506 from the start. The disadvantage is that it requires a lot of
507 space. The advantage is that it requires less communications
509 - The Job record must be updated to indicate that one or more
511 - At end of Job, FD returns:
512 1. Count of base files/bytes not written to tape (i.e. matches)
513 2. Count of base file that were saved i.e. had changed.
514 - No tape record would be written for a Base file that matches, in the
515 same way that no tape record is written for Incremental jobs where
516 the file is not saved because it is unchanged.
517 - On a restore, all the Base file records must explicitly be
518 found from the BaseFile tape. I.e. for each Full save that is marked
519 to have one or more Base Jobs, search the BaseFile for all occurrences
521 - An optimization might be to make the BaseFile have:
527 This would avoid the need to explicitly fetch each File record for
528 the Base job. The Base Job record will be fetched to get the
529 VolSessionId and VolSessionTime.
530 =========================================================
532 ==========================================================
534 For each Incremental job that is run, there may be files that
535 were found but not saved because they were locked (this applies
536 only to Windows). Such a system could send back to the Director
537 a list of Unsaved files.
539 - New UnSavedFiles table that contains:
543 - Then in the next Incremental job, the list of Unsaved Files will be
544 feed to the FD, who will ensure that they are explicitly chosen even
545 if standard date/time check would not have selected them.
546 =============================================================
550 =============================================================
552 Request For Comments For File Backup Options
555 Subject: File Backup Options
558 A few days ago, a Bacula user who is backing up to file volumes and
559 using compression asked if it was possible to suppress compressing
560 all .gz files since it was a waste of CPU time. Although Bacula
561 currently permits using different options (compression, ...) on
562 a directory by directory basis, it cannot do it on a file by
563 file basis, which is clearly what was desired.
565 Proposed Implementation:
566 To solve this problem, I propose the following:
568 - Add a new Director resource type called FileOptions.
570 - The FileOptions resource will have records for all
571 options that can currently be specified on the Include record
572 (in a FileSet). Examples below.
574 - The FileOptions resource will permit an exclude option as well
575 as a number of additional options.
577 - The heart of the FileOptions resource is the ability to
578 supply any number of ApplyTo records which specify POSIX
579 regular expressions. These ApplyTo regular expressions are
580 applied to the fully qualified filename (path and all). If
581 one matches, then the FileOptions will be used.
583 - When an ApplyTo specification matches an included file, the
584 options specified in the FileOptions resource will override
585 the default options specified on the Include record.
587 - Include records will be modified to permit referencing one or
588 more FileOptions resources. The FileOptions will be used
589 in the order listed on the Include record and the first
590 one that matches will be applied.
592 - Options (or specifications) currently supplied on the Include
593 record will be deprecated (i.e. removed in a later version a
594 year or so from now).
596 - The Exclude record will be deprecated as the same functionality
597 can be obtained by using an Exclude = yes in the FileOptions.
600 The following records can appear in the FileOptions resource. An
601 asterisk preceding the name indicates a feature not currently
605 - Compression= (GZIP, ...)
606 - Signature= (MD5, SHA1, ...)
608 - OneFs= (yes/no) - remain on one filesystem
609 - Recurse= (yes/no) - recurse into subdirectories
610 - Sparse= (yes/no) - do sparse file backup
611 - *Exclude= (yes/no) - exclude file from being saved
612 - *Reader= (filename) - external read (backup) program
613 - *Plugin= (filename) - read/write plugin module
616 - verify= (ipnougsamc5) - verify options
619 - replace= (always/ifnewer/ifolder/never) - replace options currently
621 - *Writer= (filename) - external write (restore) program
625 Currently options specifying compression, MD5 signatures, recursion,
626 ... of a FileSet are supplied on the Include record. These will now
627 all be collected into a FileOptions resource, which will be
628 specified on the Include in place of the options. Multiple FileOptions
629 may be specified. Since the FileOptions contain regular expressions
630 that are applied to the full filename, this will give the ability
631 to specify backup options on a file by file basis to whatever level
640 Include = compression=GZIP signature=MD5 {
649 Include = FileOptions=Opts {
660 That's a lot more to do the same thing, but it gives the ability to
661 apply options on a file by file basis. For example, suppose you
662 want to compress all files but not any file with extensions .gz or .Z.
663 You could do so as follows:
667 Include = FileOptions=NoCompress FileOptions=Opts {
675 ApplyTo = /*.?*/ # matches all files
680 # Note multiple ApplyTos are ORed
681 ApplyTo = /*.gz/ # matches .gz files */
682 ApplyTo = /*.Z/ # matches .Z files */
685 Now, since the NoCompress FileOptions is specified first on the
686 Include line, any *.gz or *.Z file will have an MD5 signature computed,
687 but will not be compressed. For all other files, the NoCompress will not
688 match, so the Opts options will be used which will include GZIP
692 - Is it necessary to provide some means of ANDing regular expressions
693 and negation? (not currently planned)
695 e.g. ApplyTo = /*.gz/ && !/big.gz/
697 - I see that Networker has a "null" module which, if specified, does not
698 backup the file, but does make an record of the file in the catalog
699 so that the catalog will reflect an exact picture of the filesystem.
700 The result is that the file can be "seen" when "browsing" the save
701 sets, but it cannot be restored.
703 Is this really useful? Should it be implemented in Bacula?
706 After implementing the above, the user will be able to specify
707 on a file by file basis (using regular expressions) what options are
708 applied for the backup.
709 ====================================
711 =========================================
712 Proposal by Bill Sellers
714 Return-Path: <w.a.sellers@larc.nasa.gov>
715 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
716 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)
717 Message-Id: <5.1.0.14.2.20030114153452.028dbae8@pop.larc.nasa.gov>
718 X-Sender: w.a.sellers@pop.larc.nasa.gov
719 X-Mailer: QUALCOMM Windows Eudora Version 5.1
720 Date: Tue, 14 Jan 2003 16:30:18 -0500
721 To: Kern Sibbald <kern@sibbald.com>
722 From: Bill Sellers <w.a.sellers@larc.nasa.gov>
723 Subject: Re: [Bacula-users] Bacula remote storage?
724 In-Reply-To: <1042565382.1845.177.camel@rufus>
725 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>
727 Content-Type: text/plain; charset="us-ascii"; format=flowed
728 X-Annoyance-Filter-Junk-Probability: 0
729 X-Annoyance-Filter-Classification: Mail
730 At 06:29 PM 1/14/2003 +0100, you wrote:
733 >Well, if you cannot put a Bacula client on the machine,
734 >then it is a big problem. If you know of some software
735 >that can do what you want, let me know, because I
736 >really just don't know how to do it -- at least not
742 We have been able to get Amanda to use the HSM as a storage
743 device. Someone here wrote a driver for Amanda. BUT, Amanda doesn't
744 handle Windows systems very well (or at all without Samba). So I am
745 looking for a backup system that has a Windows client. I really like the
746 Windows integration of Bacula.
748 From the command line, its rather trivial to move the data around. We use
751 tar cf - ./files | gzip -c | rsh hsm dd of=path/file.tgz
753 or if you use GNU tar:
755 tar czf hsm:path/file.tgz ./files
757 One idea for you to consider; Sendmail offers pipes in the aliases file;
758 (mailpipe: "|/usr/bin/vacation root") and Perl supports pipes in the
759 "open" statement (open FILE, "|/bin/nroff -man";) Could you could make a
760 pipe available, as a storage device? Then we could use any command that
761 handles stdin as a storage destination.
768 #Password is not used in rsh, but might be used in ftp.
769 Device = "| gzip -c | rsh hsm dd of=path/file.tgz"
776 Password = "foobar&-"
777 Device = "| ncftpput -c hsm /path/file.bacula"
781 >If you have some local storage available, you could
782 >use Bacula to backup to disk volumes, then use some
783 >other software (ftp, scp) to move them to the HSM
784 >machine. However, this is a bit kludgy.
787 It is, but maybe worth a try. Is there some function in Bacula to put
788 variables in filenames? i.e. backup.2003-01-15.root
795 w.a.sellers@larc.nasa.gov
797 ==============================================
798 The Project for the above
800 I finally realized that this is not at all
801 the same as reader/writer programs or plugins,
802 which are alternate ways of accessing the
803 files to be backed up. Rather, it is an alternate
804 form of storage device, and I have always planned
805 that Bacula should be able to handle all sorts
808 So, I propose the following phases:
810 1. OK from you to invest some time in testing
811 this as I implement it (requires that you
812 know how to download from the SourceForge
813 cvs -- which I imagine is a piece of cake
816 2. Dumb implementation by allowing a device to
817 be a fifo for write only.
818 Reason: easy to implement, proof of concept.
820 3. Try reading from fifo but with fixed block
822 Reason: proof of concept, easy to implement.
824 4. Extend reading from fifo (restores) to handle
826 Reason: requires some delicate low level coding
827 which could destabilize all of Bacula.
829 5. Implementation of above but to a program. E.g.
830 Device = "|program" (not full pipeline).
831 Reason: routines already exist, and program can
832 be a shell script which contains anything.
834 6. Add full pipeline as a possibility. E.g.
835 Device = "| gzip -c | rsh hsm dd of=path/file.tgz"
836 Reason: needs additional coding to implement full
837 pipeline (must fire off either a shell or all
838 programs and connect their pipes).
840 There are a good number of details in each step
841 that I have left out, but I will specify them at
842 every stage, and there may be a few changes as things
843 evolve. I expect that to get to stage 5 will take a
844 few weeks, and at that point, you will have
845 everything you need (just inside a script).
846 Stage 6 will probably take longer, but if this
847 project pleases you, what we do for 5 should
848 be adequate for some time.
852 =============================================
854 Done: (see kernsdone for more)
855 - Look into Pruning/purging problems or why there seem to
856 be so many files listed each night.
857 - Fix cancel in find_one -- need jcr.
858 - Cancel does not work for restore in FD.
859 - Write SetJobStatus() function so cancel status not lost.
860 - Add include list to end of chain in findlib
861 - Zap sd_auth_key after use
862 - Add Bar code reading capabilities (new mtx-changer)
863 - Figure out some way to automatically backup all local partitions
864 - Make hash table for linked files in findlib/find_one.c:161
866 - Rewrite find_one.c to use only pool_memory instead of
867 alloca and malloc (probably not necessary).
868 - Make sure btraceback goes into /sbin not sysconf directory.
869 - InitVerify is getting pruned and it shouldn't (document it)
870 - Make 1.28c release ??? NO do 1.29 directly
871 - Set timeout on opening fifo for save or restore (findlib)
872 - Document FIFO storage device.
873 - Document fifo and | and <
876 - Get correct error status from run_program or open_bpipe().
877 - Restrict permissions on File Volumes (now 0640).
878 - Umasked 022 daemons
879 - Fix restore of hard linked file.
880 - Figure out how to allow multiple simultaneous file Volumes on a single device.
881 - Implement multiple simultaneous file Volumes on a single device.
882 - Cleanup db_update_media and db_update_pool
883 - Flush all the daemon messages at the end of every job.
884 - Change stat1= fgets()!=NULL to stat1=fgest()==NULL; in
885 run_program -- bpipe.c
886 - Apparently cancel does not work for jobs waiting to be
888 - Implement TCP/IP connection for MySQL
889 - Pull a canceled job from the Scheduling queue.
890 - Implement max_file_size in block.c (already done, just tweaked).
891 - Look at purge jobs volume (at least document it, and see if it is
893 - Add list volumes does all pools. list volumes pool=xxx now works.
894 - Add pool= to "list media" in ua_output.c
895 - Strip trailing slashes from Include directory names in the FD.
896 - Fix Error: bnet.c:408 gethostbyname() for lpmatou failed:
897 ERR=Operation not permited loop.
898 - Add code if there is no mtio.h (cannot do -- too many ioctl defines needed)
899 - Produce better error messages in when error/eof writing block.
900 - Cancelling of a queued job does NOT work!!!!!!
902 rufus-dir: Volume used once. Marking Volume "File0003" as Used.
903 rufus-sd: Recycled volume File0003 on device /home/kern/bacula/working, all previous data lost.
904 rufus-dir: Volume used once. Marking Volume "File0003" as Used.
905 - Ability to backup to a file then later transfer to a tape -- Migration.
906 Migration based on MaxJobs(MinJobs),MaxVols(MinVols),AgeJobs,MaxBytes(MinBytes)
907 (i.e. HighwaterMark, LowwaterMark).
908 - Eugeny Fisher <efischer@vip-rus.com> wants to cycle through a
909 set of volumes recycling the oldest volume when it is needed.
910 - gethostbyname failure in bnet_connect() continues
911 generating errors -- should stop.
912 - Add chflags() code for FreeBSD file flags
913 - Bevan Anderson suggests having a run queue for each device
914 so that multiple simultaneous jobs can run but each writing
915 to a different Volume.
916 - Look at handling <> in smtp doesn't work with exim.
917 - Need to specify MaximumConcurrentJobs in the Job resource.
918 - ***test GetFileAttributexEx, and remove MessageBox at 335 of winservice.cpp ****
919 - Implement finer multiprocessing options.
920 - Implement | and < in Exclude statements.
921 - Figure out some way to specify a retention period for files
922 that no longer exist on the machine -- so that we maintain
923 say backups for 30 days, but if the file is deleted, we maintain
924 the last copy for 1 year. -- answer Volume retention.
925 - Make non-zero status from RunJobBefore/After error the job.
926 - Need define of int_least16_t in sha1.h for SuSE.
927 - Implement bar code reader for autochangers
928 - Document new MaximumConcurrentJob records (Job, Client, Storage)
929 - Write up how to use/manage disk Volume Storage. ******
930 - Remove kern and kelvin from mysql_grant...
931 - Install grant_mysql...
932 - Strip trailing / from Include
933 - add #define ENABLE_NLS for Gnome compile on SuSE.
934 - Add Client FS/OS id (Linux, Win95/98, ...).
935 - Concept of VolumeSet during restore which is a list
936 of Volume names needed.
937 - Turn virutally all sprintfs into snprintfs.
938 - Update volume=Test01 requests pool, then lists volumes.
939 **** Test select_pool_and_media ...
941 - Add IP address to authentication failures.
942 - Add a default File storage so that new users can do backup
943 and restores right away.