Bacula on OpenSuse 10.3 - Suse

This is a discussion on Bacula on OpenSuse 10.3 - Suse ; Anyone using backula on OpenSuse 10.3? Just installed it and was looking for the documentation of it for this release....

+ Reply to Thread
Results 1 to 8 of 8

Thread: Bacula on OpenSuse 10.3

  1. Bacula on OpenSuse 10.3

    Anyone using backula on OpenSuse 10.3? Just installed it and was
    looking for the documentation of it for this release.

  2. Re: Bacula on OpenSuse 10.3

    scott wrote:

    > Anyone using backula on OpenSuse 10.3? Just installed it and was
    > looking for the documentation of it for this release.


    Have a look at:

    /usr/share/doc/packages/bacula

    or just type in a terminal:

    man bacula

    or go to the web site:

    http://www.bacula.org/en/?page=documentation

  3. Re: Bacula on OpenSuse 10.3

    On 2008-08-02 15:34, scott wrote:
    > Anyone using backula on OpenSuse 10.3? Just installed it and was
    > looking for the documentation of it for this release.


    I run it on 10.3 and many other linux dists, and even on solaris.

    The time you spend on getting it to work is not wasted, since it
    will run for years by itself later, and you don't have to worry
    about baby sitting your backup.

    /bb

  4. Re: Bacula on OpenSuse 10.3

    birre wrote:
    > The time you spend on getting it to work is not wasted, since it
    > will run for years by itself later, and you don't have to worry
    > about baby sitting your backup.


    Who cares about backups, it is restores I am interested in. ;-)

    As I said earlier somewhere, I am testing storeBackup 2.0 now. In this I
    came to the (perhaps wrong) conclusion that one should not backup Cache
    from e.g. firefox or opera.

    Any ideas on this?

    houghi
    --
    Quote correct (NL) http://www.briachons.org/art/quote/
    Zitiere richtig (DE) http://www.afaik.de/usenet/faq/zitieren
    Quote correctly (EN) http://www.netmeister.org/news/learn2quote.html

  5. Re: Bacula on OpenSuse 10.3

    On 2008-08-04 13:23, houghi wrote:
    > birre wrote:
    >> The time you spend on getting it to work is not wasted, since it
    >> will run for years by itself later, and you don't have to worry
    >> about baby sitting your backup.

    >
    > Who cares about backups, it is restores I am interested in. ;-)
    >
    > As I said earlier somewhere, I am testing storeBackup 2.0 now. In this I
    > came to the (perhaps wrong) conclusion that one should not backup Cache
    > from e.g. firefox or opera.
    >
    > Any ideas on this?
    >
    > houghi


    Well houghi , there is so much useless **** on the backups, as firefox cashe,
    and the horrible beagle indexes if someone use it, and kde cashes, and many
    other applications that make tempfiles and just leave them there, but
    it's a lot of work to identify them all and write excludes for them,
    so I often backup that too even if it's useless.

    You are right about the recover part, that is the first thing to think about
    when designing a system, then how to partition and store the data, so
    it's possible to backup and restore the system as you hope.

    Today I see far to many "admins" that install servers, make gigantic
    filesystems, and mix data,apps,logs,tempfiles in a big mess, take it in
    production and then order someone to fix a backup, and the result can be
    2 days to recover a machine that is critical for the users.
    (if it can be restored at all)

    I have for sure tested bare metal recovery many times, and plan my backups
    so I can restore with more then one data stream, using multiple pools, so I
    can run multiple restores at the same time, and get the important things
    recovered first so the machine can be in service while less important files
    are recovered.

    It's not unusual to see ads for a filer perfect for the small office with a
    price they can afford , giving them a raided storage of about 4TB and up,
    but not a word about how to backup or restore it.

    When they buy that system and start to think they will after a while figure out
    that whatever backup they have it takes too long time to do recover so they
    must buy one more filer with the same space plus an archiver for old files,
    and the final cost is nowhere near what they thought from the beginning.

    Backup is a pain, but can't be ignored.

    I think bacula is very good, but have not yet a gui for click-admins, it must
    be configured by the brain and not the fingers.
    I use it for tape storage and file storage, and wrote an auto-scheduler for it
    so it admin itself, I just read the emails it send daily, and some times recover
    a lost file or directory.


    /bb

  6. Re: Bacula on OpenSuse 10.3

    birre wrote:
    > Well houghi , there is so much useless **** on the backups, as firefox cashe,
    > and the horrible beagle indexes if someone use it, and kde cashes, and many
    > other applications that make tempfiles and just leave them there, but
    > it's a lot of work to identify them all and write excludes for them,
    > so I often backup that too even if it's useless.


    I know I also do this, but I would like not to. I am a sole user, so the
    few extra MB is not even worth mentioning, but I like to learn.
    The more you think about cache, the more I realize it should never be in
    your home directory. It should be placed in /tmp or /var/tmp (obviously
    with thesame rights you protect your home directory)

    Then there is the fact that all Cache directories have a different name.

    > You are right about the recover part, that is the first thing to think about
    > when designing a system, then how to partition and store the data, so
    > it's possible to backup and restore the system as you hope.


    I learned that cp is the best and easiest and most often used recovery
    tool. People always talk about backup, I like to talk about recovery
    also because it makes things a lot easier.

    There are two ways of recovery:
    1) System recovery
    2) Data recovery

    The second is the most obvious and most used one. Some idiot (most of
    the time myself) deleted a file or directory that should not have been
    deleted. I use `cp` to put that back as in `cp ~/backup/dir/file
    ~/dir/file` or use any GUI tool.

    > Backup is a pain, but can't be ignored.


    The backup can be ignored if the restore is.

    > I think bacula is very good, but have not yet a gui for click-admins, it must
    > be configured by the brain and not the fingers.


    GUI for a backup is the worst, I think. OTOH I can imagine that it can
    be a help as well if there are many options to select from.

    > I use it for tape storage and file storage, and wrote an auto-scheduler for it
    > so it admin itself,


    I use cron

    > I just read the emails it send daily, and some times recover a lost
    > file or directory.



    Only the errors if there are any.

    houghi
    --
    Quote correct (NL) http://www.briachons.org/art/quote/
    Zitiere richtig (DE) http://www.afaik.de/usenet/faq/zitieren
    Quote correctly (EN) http://www.netmeister.org/news/learn2quote.html

  7. Re: Bacula on OpenSuse 10.3

    On 2008-08-04 19:30, houghi wrote:

    >
    > There are two ways of recovery:
    > 1) System recovery
    > 2) Data recovery


    Yes, that is what I do too, but maybe wrote it another way :-)
    and I first install so much so the **** boot, and then restore
    multiple pools at the same time, so I stream from 4 AIT tapes,
    or multiple backup disks.
    >
    > The second is the most obvious and most used one. Some idiot (most of
    > the time myself) deleted a file or directory that should not have been
    > deleted. I use `cp` to put that back as in `cp ~/backup/dir/file
    > ~/dir/file` or use any GUI tool.


    I never use cp for backup/restore, it don't preserve ACL's and also reset atime.
    (at least what I know)

    >
    >> Backup is a pain, but can't be ignored.

    >
    > The backup can be ignored if the restore is.
    >


    >> I think bacula is very good, but have not yet a gui for click-admins, it must
    >> be configured by the brain and not the fingers.

    >
    > GUI for a backup is the worst, I think. OTOH I can imagine that it can
    > be a help as well if there are many options to select from.
    >
    >> I use it for tape storage and file storage, and wrote an auto-scheduler for it
    >> so it admin itself,

    >
    > I use cron


    My scheduler is just a bash script started from cron that check the database for
    jobs that has less then 2 level 0 backups, and start a Full backup if needed.

    example:
    SQL="/opt/as/mysql/0/bin/mysql -r --skip-column-names bacula"
    jobnames() {
    $SQL << EOF
    select distinct Name from Job;
    EOF
    }

    numberfull() {
    $SQL << EOF
    select count(JobID) from Job where Name="$1" and Level="F" and JobStatus="T";
    EOF
    }

    I use the returned values to figure out what to run.

    >
    >> I just read the emails it send daily, and some times recover a lost
    >> file or directory.

    >
    >
    > Only the errors if there are any.
    >


    I need some saved info about the latest bsr files, so I know which
    tapes/file devices to use if I lost the database or the system.


    > houghi


    /bb

  8. Re: Bacula on OpenSuse 10.3

    birre wrote:
    >> There are two ways of recovery:
    >> 1) System recovery
    >> 2) Data recovery

    >
    > Yes, that is what I do too, but maybe wrote it another way :-)
    > and I first install so much so the **** boot, and then restore
    > multiple pools at the same time, so I stream from 4 AIT tapes,
    > or multiple backup disks.


    I only do datarecovery. I do not need the system recovery as long as I
    can revocer the data. A new installation will take be about an hour max.

    > I never use cp for backup/restore, it don't preserve ACL's and also reset atime.
    > (at least what I know)


    Only cp for the restore of individual files or directories. e.g. I
    delete ~/importand/file.xls instead of moving it by accident. I then
    just do `cp ~/backup/importand/file.xls ~/importand/file.xls` and I have
    the version of maximimum 12 hours ago as do two backups per day.

    >> I use cron

    >
    > My scheduler is just a bash script started from cron that check the database for
    > jobs that has less then 2 level 0 backups, and start a Full backup if needed.


    I would't call that a scheduler. I would call that a script. ;-)
    Also my backup takes care of what needs to be backedup, I don't.
    To me that part should be implemented into the backup software. If it
    isn't, I would be looking for something else.

    >> Only the errors if there are any.
    >>

    >
    > I need some saved info about the latest bsr files, so I know which
    > tapes/file devices to use if I lost the database or the system.


    Obviously I have saved info. 5 latets backups I have the full logfiles
    in /var/log and in each backup I have the logs as well.

    houghi
    --



    This space left blank intentionaly

+ Reply to Thread