- Kern's ToDo List
- 28 May 2004
+ Kern's ToDo List
+ 18 November 2004
+
+Major development:
+Project Developer
+======= =========
+IPv6_2 Meno Abels
+Data encryption Meno Abels (see projects)
+Communication encryption Meno Abels
+Version 1.37 Kern (see below)
+========================================================
+
+For 1.36.1:
+- Fix orphanned buffers:
+ Orphaned buffer: 24 bytes allocated at line 808 of rufus-dir job.c
+ Orphaned buffer: 40 bytes allocated at line 45 of rufus-dir alist.c
+- Add dump of VolSessionId/Time and FileIndex with bls.
+
+1.37 Projects:
+#3 Migration (Move, Copy, Archive Jobs)
+#4 Embedded Python Scripting
+#5 Events that call a Python program
+#6 Select one from among Multiple Storage Devices for Job
+#7 Single Job Writing to Multiple Storage Devices
+
+Regression tests:
+- Add Pool/Storage override regression test.
+- Add delete JobId to regression.
+- Add a regression test for dbcheck.
+- New test to add bscan to four-concurrent-jobs regression,
+ i.e. after the four-concurrent jobs zap the
+ database as is done in the bscan-test, then use bscan to
+ restore the database, do a restore and compare with the
+ original.
+- Add restore of specific JobId to regression (item 3
+ on the restore prompt)
+- Add IPv6 to regression
+- Add database test to regression. Test each function like delete,
+ purge, ...
+
+
+1.37 Possibilities:
+- Browse generations of files.
+- Add offline command to Bacula console.
+- I've seen an error when my catalog's File table fills up. I
+ then have to recreate the File table with a larger maximum row
+ size. Relevant information is at
+ http://dev.mysql.com/doc/mysql/en/Full_table.html ; I think the
+ "Installing and Configuring MySQL" chapter should talk a bit
+ about this potential problem, and recommend a solution.
+- For Solaris must use POSIX awk.
+- Want speed of writing to tape while despooling.
+- Supported autochanger:
+OS: Linux
+Man.: HP
+Media: LTO-2
+Model: SSL1016
+Slots: 16
+Cap: 200GB
+- Save mount point for directories not traversed with
+ onefs=yes.
+- Implement WildFile and WildDir to solve problem of
+ saving only *.doc files.
+- Add regex from http://www.pcre.org to Bacula for Win32.
+- Cancel command should include JobId in list of Jobs.
+- Require restore via the restore command or make a restore Job
+ get the bootstrap file.
+- Use only shell tools no make in CDROM package.
+- Document PostgreSQL performance problems bug 131.
+- Document testing
+- Add performance testing hooks
+- if 2 concurrent backups are attempted on the same tape
+ drive (autoloader) into different tape pools, one of them will exit
+ fatally instead of halting until the drive is idle
+- Add seconds to start and end times in the Job report output.
+- Add disk seeking on restore.
+- Document that ChangerDevice is used for Alert command.
+- Include within include does it work?
+- Implement a Pool of type Cleaning?
+- Implement VolReadTime and VolWriteTime in SD
+- Modify Backing up Your Database to include a bootstrap file.
+- Think about making certain database errors fatal.
+- Look at correcting the time jump in the sceduler for daylight
+ savings time changes.
+- Add a "real" timer to network connections.
+- Tell the "restore" user when browsing is no longer possible.
+- Write non-optimized bsrs from the JobMedia and Media records,
+ even after Files are pruned.
+- Document that Bootstrap files can be written with cataloging
+ turned off.
+- Look at adding full Volume and Pool information to a Volume
+ label so that bscan can get *all* the info.
+- Promote to Full = Time period
+- Scratch Pool where the volumes can be re-assigned to any Pool.
+- Update StartTime if job held in Job Queue.
+- Despool attributes simultaneously with data in a separate
+ thread, rejoined at end of data spooling.
+- Implement Files/Bytes,... stats for restore job.
+- Implement Total Bytes Written, ... for restore job.
+- Check dates entered by user for correctness (month/day/... ranges)
+- Compress restore Volume listing by date and first file.
+- Look at patches/bacula_db.b2z postgresql that loops during restore.
+ See Gregory Wright.
+- Perhaps add read/write programs and/or plugins to FileSets.
+- How to handle backing up portables ...
+- Add some sort of guaranteed Interval for upgrading jobs.
+- Can we write the state file after every job terminates? On Win32
+ the system crashes and the state file is not updated.
+- Limit bandwidth
Documentation to do: (any release a little bit at a time)
+- Doc to do unmount before removing magazine.
+- Alternative to static linking "ldd prog" save all binaries listed,
+ restore them and point LD_LIBRARY_PATH to them.
+- Document add "</dev/null >/dev/null 2>&1" to the bacula-fd command line
- Document query file format.
+- Add more documentation for bsr files.
- Document problems with Verify and pruning.
- Document how to use multiple databases.
- VXA drives have a "cleaning required"
changing tape settings while a job is running.)
- Lookup HP cleaning recommendations.
- Lookup HP tape replacement recommendations (see trouble shooting autochanger)
-- Add more obvious documentation of @ for conf files.
- Document doing table repair
+
Testing to do: (painful)
-- Test drive polling!
-- blocksize recognition code.
-- Test if rewind at end of tape waits for tape to rewind.
-- Test cancel at EOM.
-For 1.33 Testing/Documentation:
-- Figure out how to use ssh or stunnel to protect Bacula communications.
- Add Dan's work to manual See ssl.txt
-- Add db check test to regression. Test each function like delete,
- purge, ...
+
+For 1.37 Testing/Documentation:
+- Add bandwidth limiting.
+- Add acks every once and a while from the SD to keep
+ the line from timing out.
+- When an error in input occurs and conio beeps, you can back
+ up through the prompt.
+- Detect fixed tape block mode during positioning by looking at
+ block numbers in btape "test". Possibly adjust in Bacula.
+- Fix list volumes to output volume retention in some other
+ units, perhaps via a directive.
+- If opening a tape in read/write mode fails attempt to open
+ it in read-only mode, and mark the tape for read only.
+- Add a read-only mode to the mount option.
+- Allow Simultaneous Priorities = yes => run up to Max concurrent jobs even
+ with multiple priorities.
- If you use restore replace=never, the directory attributes for
non-existent directories will not be restored properly.
-1.34 Cleanup
+- see lzma401.zip in others directory for new compression
+ algorithm/library.
+- Minimal autochanger handling in Bacula and in btape.
+- Look into how tar does not save sockets and the possiblity of
+ not saving them in Bacula (Martin Simmons reported this).
+- Add All Local Partitions = yes to new style saves.
+- localmounts=`awk '/ext/ { print $2 }' /proc/mounts` # or whatever
+ find $localmounts -xdev -type s -ls
+- Fix restore jobs so that multiple jobs can run if they
+ are not using the same tape(s).
+- Allow the user to select JobType for manual pruning/purging.
+- Look at adding Client run command that will use the
+ port opened by the client.
+- bscan does not put first of two volumes back with all info in
+ bscan-test.
+- Implement the FreeBSD nodump flag in chflags.
+- Figure out how to make named console messages go only to that
+ console and to the non-restricted console (new console class?).
+- Make restricted console prompt for password if *ask* is set or
+ perhaps if password is undefined.
+- Implement "from ISO-date/time every x hours/days/weeks/months" in
+ schedules.
+==== from Marc Schoechlin
+- the help-command should be more verbose
+ (it should explain the paramters of the different
+ commands in detail)
+ -> it´s time-comsuming to consult the manual anytime
+ you need a special parameter
+ -> maybe it´s more easy to maintain this, if the
+ descriptions of that commands are outsourced to
+ a ceratin-file
+- the cd-command should allow complete paths
+ i.e. cd /foo/bar/foo/bar
+ -> if a customer mails me the path to a certain file,
+ it´s faster to enter the specified directory
+- if the password is not configured in bconsole.conf
+ you should be asked for it.
+ -> sometimes you like to do restore on a customer-machine
+ which shouldn´t know the password for bacula.
+ -> adding the password to the file favours admins
+ to forget to remove the password after usage
+ -> security-aspects
+ the protection of that file is less important
+- long-listed-output of commands should be scrollable
+ like the unix more/less-command does
+ -> if someone runs 200 and more machines, the lists could
+ be a little long and complex
+- command-output should be shown column by column
+ to reduce scrolling and to increase clarity
+ -> see last item
+- lsmark should list the selected files with full
+ paths
+- wildcards for selecting and file and directories would be nice
+- any actions should be interuptable with STRG+C
+- command-expansion would be pretty cool
+====
+- Compaction of Disk space by "migrating" Volumes that have pruned
+ Jobs (what criteria? size, #jobs, time).
+- Add prune all command
+- Document fact that purge can destroy a part of a restore by purging
+ one volume while others remain valid -- perhaps mark Jobs.
- Add multiple-media-types.txt
- look at mxt-changer.html
-
-1.34.3 Items:
-- Document dbcheck conf changes.
-- Add regression of btape "fill"
-- Document if absolute path specified extension must be specified
- on Win32.
-- Document vendor database repair.
-
-For version 1.35:
-- Do tape alerts -- see tapealert.txt
+- Make ? do a help command (no return needed).
- Implement restore directory.
-- Add All Local Partitions = yes to new style saves.
- Document streams and how to implement them.
-- On Win95
- The error I when I installed 1.34.2 clients:
- The BACULA-FD file is
- linked to missing export KERNEL32.DLL:GetFileAttributesExA.
-- When restore started from console, report jobid.
-- Feedback while the tree is being built.
-- Possibly implement "Ensure Full Backup = yes" looks for a failed full backup
- and upgrades the current backup if one exists.
-- Check that barcode reading and update slots scan works.
- Try not to re-backup a file if a new hard link is added.
- Add feature to backup hard links only, but not the data.
-- Add "All Local = yes" option to save to include all local partitions.
- Fix stream handling to be simpler.
- Add Priority and Bootstrap to Run a Job.
- Eliminate Restore "Run Restore Job" prompt by allowing new "run command
to be issued"
- Remove View FileSet button from Run a Job dialog.
-- Don't count higher level non-saved directories in Files expected
- in restore report.
- Handle prompt for restore job at end of Restore command.
- Add display of total selected files to Restore window.
- Add tree pane to left of window.
- Add progress meter.
-- Polling does not work for restore. It tries a number of times,
- gives up, and crashes the FD.
-- Lock jcr_chain when doing attach/detach in acquire.c
-- Add assert in free_jcr if attach/detach chain active.
- Max wait time or max run time causes seg fault -- see runtime-bug.txt
- Document writing to a CD/DVD with Bacula.
-- Add check for tape alerts.
- Add a "base" package to the window installer for pthreadsVCE.dll
which is needed by all packages.
- Add message to user to check for fixed block size when the forward
- Possibly implement: Action = Unmount Device="TapeDrive1" in Admin jobs.
- Setup lrrd graphs: (http://www.linpro.no/projects/lrrd/) Mike Acar.
- Revisit the question of multiple Volumes (disk) on a single device.
-- Finish SIGHUP work.
-- Check that all change in wait status in the SD are
- signaled to the Director.
- Add a block copy option to bcopy.
- Investigate adding Mac Resource Forks.
- Finish work on Gnome restore GUI.
- Fix "llist jobid=xx" where no fileset or client exists.
-- Check pruning of restore jobs.
- From Chris Hull:
it seems to be complaining about 12:00pm which should be a valid 12
hour time. I changed the time to 11:59am and everything works fine.
fields.
- Fix option 2 of restore -- list where file is backed up -- require Client,
then list last 20 backups.
-- Add all pools in Dir conf to DB also update them to catch changed
- LabelFormats and such.
-- Update volumes FromPool=xxx does all volumes.
- Pass Director resource name as an option to the Console.
- Add a "batch" mode to the Console (no unsolicited queries, ...).
-- Add code to check for tape alerts -- tapeinfo.
-- Make sure list of Volumes needed is in correct order for restore.
- See havana.
-- Remove paths (and files that reference them) that have no trailing slash
- in dbcheck -- or add a trailing slash.
-- Remove Filenames (and files that reference them) that have a trailing
- slash in dbcheck -- or remove the trailing slash.
-- Remove orphaned paths/filenames by copying them to a new table with a
- reference count, then mark all referenced files/paths and remove unreferenced
- ones.
- Add a .list all files in the restore tree (probably also a list all files)
Do both a long and short form.
- Allow browsing the catalog to see all versions of a file (with
stat data on each file).
- Restore attributes of directory if replace=never set but directory
did not exist.
-- Allow "delete job jobid=xxx,yyy,aaa-bbb" i.e. list + ranges.
- Use SHA1 on authentication if possible.
- See comtest-xxx.zip for Windows code to talk to USB.
-- Make btape accept Device Names in addition to Archive names.
- Add Events and Perl scripting.
- Add John's appended files:
Appended = { /files/server/logs/http/*log }
creation date of files, we can be evensmarter about this and not count
entirely upon the length.On a restore, the full and all incrementals
since it will beapplied in sequence to restore the file.
-- Add a regression test for dbcheck.
-- Add disk seeking on restore. - Allow
- for optional cancelling of SD and FD in case DIR
+- Allow for optional cancelling of SD and FD in case DIR
gets a fatal error. Requested by Jesse Guardiani <jesse@wingnet.net>
-- Bizarre message: Error: Could not open WriteBootstrap file:
-- Build console in client only build.
- Add "limit=n" for "list jobs"
- Check new HAVE_WIN32 open bits.
- Check if the tape has moved before writing.
- Handling removable disks -- see below:
-- Multiple drive autochanger support -- see below.
- Keep track of tape use time, and report when cleaning is necessary.
-- Fix FreeBSD mt_count problem.
- Add FromClient and ToClient keywords on restore command (or
BackupClient RestoreClient).
-- Automatic "update slots" on user configuration directive when a
- slot error occurs.
- Implement a JobSet, which groups any number of jobs. If the
JobSet is started, all the jobs are started together.
Allow Pool, Level, and Schedule overrides.
- Enhance cancel to timeout BSOCK packets after a specific delay.
-- When I restore to Windows the Created, Accessed and Modifiedtimes are
- those of the time of the restore, not those of the originalfile.
- The dates you will find in your restore log seem to be the original
- creation dates
-- Volume "add"ed to Pool gets recycled in first use. VolBytes=0
-- If a tape is recycled while it is mounted, Stanislav Tvrudy must do an
- additional mount to deblock the job.
-- From Johan Decock:
- bscan: sql_update.c:65 UPDATE File SET MD5='Ij+5kwN6TFIxK+8l8+/I+A' WHERE FileId=0
- bscan: bscan.c:1074 Could not add MD5/SHA1 to File record. ERR=sql_update.c:65 Update problem: affected_rows=0
- Do scheduling by UTC using gmtime_r() in run_conf, scheduler, and
ua_status.!!! Thanks to Alan Brown for this tip.
- Look at updating Volume Jobs so that Max Volume Jobs = 1 will work
correctly for multiple simultaneous jobs.
- Correct code so that FileSet MD5 is calculated for < and | filename
generation.
-- Mark Volume in error on error from WEOF.
- Implement the Media record flag that indicates that the Volume does disk
addressing.
- Implement VolAddr, which is used when Volume is addressed like a disk,
- Fix fast block rejection (stored/read_record.c:118). It passes a null
pointer (rec) to try_repositioning().
- Look at extracting Win data from BackupRead.
-- Having dashes in filenames apparently creates problems for restore
- by filename??? hard to believe.
- Implement RestoreJobRetention? Maybe better "JobRetention" in a Job,
which would take precidence over the Catalog "JobRetention".
- Implement Label Format in Add and Label console commands.
resources, like Level? If so, I think I'd make it an optional directive
in Job, Client, and Pool, with precedence such that Job overrides Client
which in turn overrides Pool.
-- Print a message when a job starts if the conf file is not current.
-
-- Spooling ideas taken from Volker Sauer's and other's emails:
- > IMHO job spooling should be turned on
- >
- > 1) by job
- > 2) by schedule
- > 3) by sd
- >
- > where and 2) overrides 1) and 3) is independent.
-
- Yes, this is the minimum that I think is necessary.
-
- >
- > Reason(s):
- > It should be switched by job, because the job that backs up the machine
- > with the bacula-sd on doesn't need spooling.
- > It should be switched by schedule, because for full-backups I don't need
- > spooling, so I can switch it off (because the network faster then the
- > tapedrive)
-
- True, with the exception that if you have enough disk spool space,
- and you want to run concurrent jobs, spooling can eliminate the block
- interleaving restore inefficiencies.
-
- > And you should be able to turn it of by sd for sd-machines with low disk
- > capacity or if you just don't need or want this feature.
- >
- > There should be:
- > - definitly the possibility for multipe spool direcories
-
- Having multiple directories is no problem -- having different maximum
- sizes creates specification problems. At some point, I will probably
- have a common SD pool of spool directories as well as a set of
- private spool directories for each device. The first implementation
- will be a set of private spool directories for each device since
- managing a global pool with a bunch of threads writing into the same
- directory is *much* more complicated and prone to error.
-
- > - the ability to spool parts of a backup (not the whole client)
-
- This may change in the future, but for the moment, it will spool
- either to a job high water mark, or until the directory is full
- (reaches max spool size or I/O error). It will then write to tape,
- truncate the spool file, and begin spooling again.
-
- > - spooling while writing to tape
-
- Not within a job, but yes, if you run concurrent jobs -- each is a
- different thread. Within a job could be a feature, but *much* later.
-
- > - parallel spooling (like parallel jobs/ concurrent jobs) of clients
-
- Yes, this is one of my main motivations for doing it (aside from
- eliminating tape "shoe shine" during incremental backups.
-
- > - flushing a backup that only went to disk (like amflush in amanda)
-
- This will be a future feature, since spooling is different from backing
- up to disk. The future feature will be "migration" which will move a job
- from one backup Volume to another.
- New Storage specifications:
+ - Want to write to multiple storage devices simultaneously
+ - Want to write to multiple storage devices sequentially (in one job)
+ - Want to read/write simultaneously
+ - Key is MediaType -- it must match
+
Passed to SD as a sort of BSR record called Storage Specification
Record or SSR.
SSR
MediaType -> Next MediaType
Pool -> Next Pool
Device -> Next Device
- Write Copy Resource that makes a copy of a resource.
Job Resource
Allow multiple Storage specifications
New flags
Storage
Allow Multiple Pool specifications (note, Pool currently
in Job resource).
- Allow Multiple MediaType specifications
- Allow Multiple Device specifications
+ Allow Multiple MediaType specifications in Dir conf
+ Allow Multiple Device specifications in Dir conf
Perhaps keep this in a single SSR
Tie a Volume to a specific device by using a MediaType that
is contained in only one device.
In SD allow Device to have Multiple MediaTypes
-After 1.33:
- Look at www.nu2.nu/pebuilder as a helper for full windows
bare metal restore.
-Ideas from Jerry Scharf:
+- Ideas from Jerry Scharf:
First let's point out some big pluses that bacula has for this
it's open source
more importantly it's active. Thank you so much for that
even more important, it's not flaky
it has an open access catalog, opening many possibilities
it's pushing toward heterogeneous systems capability
- simple things:
- I don't remember an include file directive for config files
- (not filesets, actual config directives)
- can you check the configs without starting the daemon?
- some warnings about possible common mistakes
big things:
- doing the testing and blessing of concurrent backup writes
- this is absolutely necessary in the enterprise
- easy user recovery GUI with full access checking
Macintosh file client
macs are an interesting niche, but I fear a server is a rathole
working bare iron recovery for windows
- much better handling on running config changes
- thinking through the logic of what happens to jobs in progress
the option for inc/diff backups not reset on fileset revision
a) use both change and inode update time against base time
b) do the full catalog check (expensive but accurate)
support for Oracle database ??
===
- Look at adding SQL server and Exchange support for Windows.
-- Restore: Enter Filename: 'C:/Documents and Settings/Comercial/My
- Documents/MOP/formulário de registro BELAS ARTES.doc' causes Bacula to
- crash.
- Each DVD-RAM disk would be a volume, just like each tape is
a volume. It's a 4.7GB media with random access, but there's nothing about
it that I can see that makes it so different than a tape from bacula's
be an invalid value which happens with BSR.
- Create VolAddr for disk files in place of VolFile and VolBlock. This
is needed to properly specify ranges.
-- Print bsmtp output to job report so that problems will be seen.
-- Pass the number of files to be restored to the FD for reporting
- Add progress of files/bytes to SD and FD.
-- Don't continue Restore if no files selected.
- Print warning message if FileId > 4 billion
- do a "messages" before the first prompt in Console
- Add a date and time stamp at the beginning of every line in the
Job report (Volker Sauer).
- Client does not show busy during Estimate command.
- Implement Console mtx commands.
-- Add a default DB password to MySQL.
- GRANT all privileges ON bacula.* TO bacula@localhost IDENTIFIED BY
- 'bacula_password';
- FLUSH PRIVILEGES;
- Implement a Mount Command and an Unmount Command where
the users could specify a system command to be performed
to do the mount, after which Bacula could attempt to
to the user, who would then use "mount" as described above
once he had actually inserted the disk.
- Implement dump/print label to UA
-- Implement disk spooling. Two parts: 1. Spool to disk then
- immediately to tape to speed up tape operations. 2. Spool to
- disk only when the tape is full, then when a tape is hung move
+- Spool to disk only when the tape is full, then when a tape is hung move
it to tape.
-- Scratch Pool where the volumes can be re-assigned to any Pool.
- bextract is sending everything to the log file ****FIXME****
-- Add Progress command that periodically reports the progress of
- a job or all jobs.
-- Restrict characters permitted in a Resource name, and don't permit
- duplicate names.
- Allow multiple Storage specifications (or multiple names on
a single Storage specification) in the Job record. Thus a job
can be backed up to a number of storage devices.
- Implement some way for the File daemon to contact the Director
to start a job or pass its DHCP obtained IP number.
-- Implement multiple Consoles.
- Implement a query tape prompt/replace feature for a console
-- From Johan?
- Two jobs ready to go, first one blocked waiting for media
- Cancel 2nd job ("waiting execution" one)
- Cancel blocked job
- boom - segfault*
- Copy console @ code to gnome2-console
- Make AES the only encryption algorithm see
http://csrc.nist.gov/CryptoToolkit/aes/). It's
an officially adopted standard, has survived peer
review, and provides keys up to 256 bits.
-- Think about how space could be freed up on a tape -- perhaps this
- is a Merge or Compact feature that is needed.
-- Modify FileSet, did not upgrade the current Increment job, but
- waited for the next job to be upgraded.
- Take a careful look at SetACL http://setacl.sourceforge.net
-- Implement a where command for the tree telling where a file
- is located.
- Take a careful look at Level for the estimate command, maybe make
it a command line option.
- Add Volume name to "I cannot write on this volume because"
- Make tree walk routines like cd, ls, ... more user friendly
by handling spaces better.
-- Write your PID file and chown root:wheel before drop.
-- Make sure there is no symlink in a file before creating a
- file (attack).
-- Look at mktemp or mkstemp(3).
- mktemp and mkstemp create files with predictable names too. That's
- not the vulnerability. The vulnerability is in creating files without
- using the O_EXCL flag, which means "only create this file if it doesn't
- exist, including if the file is a dangling symlink."
-
- It is *NOT* enough to do the equivalent of
-
- if doesn't exist $filename
- then create $filename
-
- because between the test and the create another process could have
- gotten the CPU and created the file. You must use atomic functions
- (those that don't get interrupted by other processes) and O_EXCL is
- the only way for this particular example.
-- Automatically create pools, but instead of looking for what
- in in Job records, walk through the pool resources.
-- Check and double check tree code, why does it take so long?
-- Add device name to "Current Volume not acceptable because ..."
- Make sure that Bacula rechecks the tape after the 20 min wait.
- Set IO_NOWAIT on Bacula TCP/IP packets.
- Try doing a raw partition backup and restore by mounting a
By the way can bacula automatically "move" an empty/purged volume say
in the "short" pool to the "long" pool if this pool runs out of volume
space?
-- Eliminate orphaned jobs: dbcheck, normal pruning, delete job command.
- Hm. Well, there are the remaining orphaned job records:
-
- | 105 | Llioness Save | 0000-00-00 00:00:00 | B | D | 0 | 0 | f |
- | 110 | Llioness Save | 0000-00-00 00:00:00 | B | I | 0 | 0 | f |
- | 115 | Llioness Save | 2003-09-10 02:22:03 | B | I | 0 | 0 | A |
- | 128 | Catalog Save | 2003-09-11 03:53:32 | B | I | 0 | 0 | C |
- | 131 | Catalog Save | 0000-00-00 00:00:00 | B | I | 0 | 0 | f |
-
- As you can see, three of the five are failures. I already deleted the
- one restore and one other failure using the by-client option. Deciding
- what is an orphaned job is a tricky problem though, I agree. All these
- records have or had 0 files/ 0 bytes, except for the restore. With no
- files, of course, I don't know of the job ever actually becomes
- associated with a Volume.
-
- (I'm not sure if this is documented anywhere -- what are the meanings of
- all the possible JobStatus codes?)
-
- Looking at my database, it appears to me as though all the "orphaned"
- jobs fit into one of two categories:
-
- 1) The Job record has a StartTime but no EndTime, and the job is not
- currently running;
- or
- 2) The Job record has an EndTime, indicating that it completed, but
- it has no associated JobMedia record.
-
-
- This does suggest an approach. If failed jobs (or jobs that, for some
- other reason, write no files) are associated with a volume via a
- JobMedia record, then they should be purged when the associated volume
- is purged. I see two ways to handle jobs that are NOT associated with a
- specific volume:
-
- 1) purge them automatically whenever any volume is manually purged;
- or
- 2) add an option to the purge command to manually purge all jobs with
- no associated volume.
-
- I think Restore jobs also fall into category 2 above .... so one might
- want to make that "The Job record has an EndTime,, but no associated
- JobMedia record, and is not a Restore job."
-- make "btape /tmp" work.
-- Make sure a rescheduled job is properly reported by status.
-- Walk through the Pool records rather than the Job records
- in dird.c to create/update pools.
- What to do about "list files job=xxx".
-- When job rescheduled, status gives is waiting for Client Rufus
- to connect to Storage File. Dir needs to inform SD that job
- is rescheduled.
- Make Dmsg look at global before calling subroutine.
-- Enable trace output at runtime for Win32
-- Available volumes for autochangers (see patrick@baanboard.com 3 Sep 03
- and 4 Sep) scan slots.
- Get and test MySQL 4.0
-- Do a complete audit of all pthreads_mutex, cond, ... to ensure that
- any that are dynamically initialized are destroyed when no longer used.
- Look at how fuser works and /proc/PID/fd that is how Nic found the
file descriptor leak in Bacula.
- Implement WrapCounters in Counters.
-- Turn on SIGHUP in dird.c and test.
-- Use system dependent calls to get more precise info on tape errors.
- Add heartbeat from FD to SD if hb interval expires.
-- Suppress read error on blank tape when doing a label.
- Can we dynamically change FileSets?
- If pool specified to label command and Label Format is specified,
automatically generate the Volume name.
- Why can't SQL do the filename sort for restore?
-- Look at libkse (man kse) for FreeBSD threading.
- Look into Microsoft Volume Shadowcopy Service VSS for backing
up system state components (Active Directory, System Volume, ...)
- Add ExhautiveRestoreSearch
data into the restore tree (i.e. do it one directory at a
time as the user walks through the tree).
- Possibly use the hash code if the user selects all for a restore command.
-- Orphaned Dir buffer at parse_conf.c:373 => store_dir
- Fix "restore all" to bypass building the tree.
- Prohibit backing up archive device (findlib/find_one.c:128)
- Implement Release Device in the Job resource to unmount a drive.
be sure this works with admin jobs so that the user can get
prompted to insert the correct tape. Possibly some way to say to
run the job but don't save the files.
-- Implement FileOptions (see end of this document)
- Make things like list where a file is saved case independent for
Windows.
- Implement migrate
-- Bacula needs to propagate SD errors.
- > > cluster-dir: Start Backup JobId 252, Job=REUTERS.2003-08-11_15.04.12
- > > prod4-sd: REUTERS.2003-08-11_15.04.12 Error: Write error on device
- /dev/nst0. ERR=Input/output error.
- > > prod4-sd: REUTERS.2003-08-11_15.04.12 Error: Re-read of last block failed.
- Last block=5162 Current block=5164.
- > > prod4-sd: End of medium on Volume "REU007" Bytes=16,303,521,933
-
- Use autochanger to handle multiple devices.
- Add SuSE install doc to list.
- Check and rechedk "Invalid block number"
-- Make bextract release the drive properly between tapes
- so that an autochanger can be made to work.
- User wants to NOT backup up certain big files (email files).
- Maybe remove multiple simultaneous devices code in SD.
- On Windows with very long path names, it may be impossible to create
a file (and thus restore it) because the total length is too long.
We must cd into the directory then create the file without the
full path name.
-- lstat() is not going to work on Win32 for testing date.
- Implement a Recycle command
- Add client name to cram-md5 challenge so Director can immediately
verify if it is the correct client.
-- Add JobLevel in FD status (but make sure it is defined).
- Audit all UA commands to ensure that we always prompt where possible.
-- Check Jmsg in bnet, may not work, must dup bsock.
-- Suppress Job Name in Jmsg for console
-- Create Pools that are referenced in a Run statement at startup if possible.
-- Use runbeforejob to unload, then reload a volume previously used,
- then the next job run gets an error reading the drive.
- Make bootstrap filename unique.
- Test a second language e.g. french.
- Start working on Base jobs.
-- Make "make binary-release" work from any directory.
- Implement UnsavedFiles DB record.
- Implement argc/argv for daemon command line scanning using table driven
stuff below.
-- Implement table driven single argc/argv scanner to pickup all arguments.
- Much like xxx_conf.c scan table.
- keyword, handler(store_routine), store_address, code, flags, default.
- From Phil Stracchino:
It would probably be a per-client option, and would be called
something like, say, "Automatically purge obsoleted jobs". What it
- Currently in mount.c:236 the SD simply creates a Volume. It should have
explicit permission to do so. It should also mark the tape in error
if there is an error.
-- Make sure all restore counters are working correctly in the FD.
- SD Bytes Read is wrong.
-- Look at ALL higher level routines that call block.c to be sure
- they don't expect something in errmsg.
- Investigate doing RAW backup of Win32 partition.
-- Add thread specific data to hold the jcr -- send error messages from
- low level routines by accessing it and using Jmsg().
- Cancel waiting for Client connect in SD if FD goes away.
-- Examine Bare Metal restore problem (a FD crash exists somewhere ...).
- Implement timeout in response() when it should come quickly.
- Implement console @echo command.
- Implement a Slot priority (loaded/not loaded).
- Implement single pane restore (much like the Gftp panes).
- Implement Automatic Mount even in operator wait.
- Implement create "FileSet"?
-- Fix watchdog pthread crash on Win32 (this is pthread_kill() Cygwin bug)
-- Implement "scratch pool" where tapes are defined and can be
- taken by any pool that needs them.
- Implement restore "current system", but take all files without
doing selection tree -- so that jobs without File records can
be restored.
- Figure out how to do a bare metal Windows restore
- Possibly add email to Watchdog if drive is unmounted too
long and a job is waiting on the drive.
-- Restore program that errs in SD due to no tape, reports
- OK incorrectly in output.
- After unmount, if restore job started, ask to mount.
- Convert all %x substitution variables, which are hard to remember
and read to %(variable-name). Idea from TMDA.
if full status requested or if some level of debug on.
- Make database type selectable by .conf files i.e. at runtime
- Set flag for uname -a. Add to Volume label.
-- Implement throttled work queue.
- Restore files modified after date
- SET LD_RUN_PATH=$HOME/mysql/lib/mysql
-- Implement Restore FileSet=
-- Create a protocol.h and protocol.c where all protocol messages
- are concentrated.
- Remove duplicate fields from jcr (e.g. jcr.level and jcr.jr.Level, ...).
- Timout a job or terminate if link goes down, or reopen link and query.
- Concept of precious tapes (cannot be reused).
- Store info on each file system type (probably in the job header on tape.
This could be the output of df; or perhaps some sort of /etc/mtab record.
+========= ideas ===============
+From: "Jerry K. Schieffer" <jerry@skylinetechnology.com>
+To: <kern@sibbald.com>
+Subject: RE: [Bacula-users] future large programming jobs
+Date: Thu, 26 Feb 2004 11:34:54 -0600
+
+I noticed the subject thread and thought I would offer the following
+merely as sources of ideas, i.e. something to think about, not even as
+strong as a request. In my former life (before retiring) I often
+dealt with backups and storage management issues/products as a
+developer and as a consultant. I am currently migrating my personal
+network from amanda to bacula specifically because of the ability to
+cross media boundaries during storing backups.
+Are you familiar with the commercial product called ADSM (I think IBM
+now sells it under the Tivoli label)? It has a couple of interesting
+ideas that may apply to the following topics.
+
+1. Migration: Consider that when you need to restore a system, there
+may be pressure to hurry. If all the information for a single client
+can eventually end up on the same media (and in chronological order),
+the restore is facillitated by not having to search past information
+from other clients. ADSM has the concept of "client affinity" that
+may be associated with it's storage pools. It seems to me that this
+concept (as an optional feature) might fit in your architecture for
+migration.
+
+ADSM also has the concept of defining one or more storage pools as
+"copy pools" (almost mirrors, but only in the sense of contents).
+These pools provide the ability to have duplicte data stored both
+onsite and offsite. The copy process can be scheduled to be handled
+by their storage manager during periods when there is no backup
+activity. Again, the migration process might be a place to consider
+implementing something like this.
+
+>
+> It strikes me that it would be very nice to be able to do things
+like
+> have the Job(s) backing up the machines run, and once they have all
+> completed, start a migration job to copy the data from disks Volumes
+to
+> a tape library and then to offsite storage. Maybe this can already
+be
+> done with some careful scheduling and Job prioritzation; the events
+> mechanism described below would probably make it very easy.
+
+This is the goal. In the first step (before events), you simply
+schedule
+the Migration to tape later.
+
+2. Base jobs: In ADSM, each copy of each stored file is tracked in
+the database. Once a file (unique by path and metadata such as dates,
+size, ownership, etc.) is in a copy pool, no more copies are made. In
+other words, when you start ADSM, it begins like your concept of a
+base job. After that it is in the "incremental" mode. You can
+configure the number of "generations" of files to be retained, plus a
+retention date after which even old generations are purged. The
+database tracks the contents of media and projects the percentage of
+each volume that is valid. When the valid content of a volume drops
+below a configured percentage, the valid data are migrated to another
+volume and the old volume is marked as empty. Note, this requires
+ADSM to have an idea of the contents of a client, i.e. marking the
+database when an existing file was deleted, but this would solve your
+issue of restoring a client without restoring deleted files.
+
+This is pretty far from what bacula now does, but if you are going to
+rip things up for Base jobs,.....
+Also, the benefits of this are huge for very large shops, especially
+with media robots, but are a pain for shops with manual media
+mounting.
+
+>
+> Base jobs sound pretty useful, but I'm not dying for them.
+
+Nobody is dying for them, but when you see what it does, you will die
+without it.
+
+3. Restoring deleted files: Since I think my comments in (2) above
+have low probability of implementation, I'll also suggest that you
+could approach the issue of deleted files by a mechanism of having the
+fd report to the dir, a list of all files on the client for every
+backup job. The dir could note in the database entry for each file
+the date that the file was seen. Then if a restore as of date X takes
+place, only files that exist from before X until after X would be
+restored. Probably the major cost here is the extra date container in
+each row of the files table.
+
+Thanks for "listening". I hope some of this helps. If you want to
+contact me, please send me an email - I read some but not all of the
+mailing list traffic and might miss a reply there.
+
+Please accept my compliments for bacula. It is doing a great job for
+me!! I sympathize with you in the need to wrestle with excelence in
+execution vs. excelence in feature inclusion.
+
+Regards,
+Jerry Schieffer
+
+==============================
+
Longer term to do:
- Design at hierarchial storage for Bacula. Migration and Clone.
- Implement FSM (File System Modules).
====
-=== Done in 1.35
-- To pass Include 1 or two letter commands
- I Name Include name - first record
- B Name Base name - repeat
- R "xxx" Regexp
- W "xxx" Wild Card
- E zzz Exclude expression (wild card)
- P "plugin" Plugin
- D "reader" Reader program
- T "writer" Writer program
- O Options In current commpressed format
- (compression, signature, onefs, recurse, sparse,
- replace, verify options, ...)
- N End option set
- B BaseName Start second option set any letter
- ...
- E
- F Number Number of filenames to follow
- <file-name>
- B Name
- ...
- N End option set
- F Number Number of filenames to follow
- <file-name>
- ...
+=== Done
+- Add "Rerun failed levels = yes/no" to Job resource.
+- Fix CDROM make script to permit picking up already installed
+ static FD.
+- Add date/time to each Jmsg.
projects / bacula / bacula / blobdiff