33 lines
3.7 KiB
Markdown
33 lines
3.7 KiB
Markdown
when you buy new tapes, install them in your library and run the
|
|
|
|
*label pool=Scratch storage=Iron-Autochanger slots=3,11,27,28 barcodes
|
|
|
|
command to write the initial label onto the tape (so bacula knows it's allowed to use the tape) and then bacula and scratch pool will do everything else
|
|
libraries have slots, that command says "load the tape from slot 1 into the drive, and write a label to it matching its barcode. then repeat for slot 2. then repeat for slot 3"
|
|
|
|
# mt -f /dev/nst0 erase 1 to quick erase disk
|
|
|
|
*status dir days=15
|
|
|
|
<kenwoodfox> Can i change that behavior?
|
|
<kenwoodfox> https://cdn.discordapp.com/attachments/306060488163328001/715312740876550164/unknown.png
|
|
<kenwoodfox> What if have a job that takes up 4.5 tapes, and later on, i purge the first four tapes that are 100% full will the last tape that only has half of its volume used by the now purged job go back to being Append or will i have to pudge that tape too?
|
|
* Topic for #bacula is "If you can't send to the channel, please register with FreeNode. || Please don't ask if you can ask... just ask || Please use pastebin.ca to paste || auth problems? http://tinyurl.com/47u2lck || SQLite no longer supported http://bit.ly/dUw2zN (login as anonymous/anonymous)"
|
|
* Topic set by dvl!~dvl@freebsd/developer/dvl on 2016-06-08 23:22:21 UTC
|
|
[@ChanServ] This is the support channel for Bacula, an open-source cross-platform enterprise backup solution from one of the principal authors of apcupsd. If you have questions, please just speak up - ve don't bite, and ve von't suck your blut. Honest!
|
|
* Homepage for #bacula is http://www.bacula.org
|
|
*** Mode #bacula +nt by orwell.freenode.net
|
|
* Channel #bacula created on 2005-08-08 07:51:28 UTC
|
|
- {Day changed to Thursday, May 28, 2020}
|
|
<kenwoodfox_> Do my incremental, difs and fulls have to line up? is that the way to do it?
|
|
<kenwoodfox_> Like doing one incremental twice a week, retaining them for four weeks, doing a diff every four weeks, retaining them for three months, and doing four full backups a year
|
|
<kenwoodfox_> I only have so many tapes,
|
|
<kenwoodfox_> Right now i do an incremental twice a week, a full backup four times a year and a diff backup every other month, just starting out so i dont know how much that will be eventually
|
|
<optiz0r> <kenwoodfox> Can i change that behavior? << No. you can only force bacula to reuse a purged volume by making sure it cannot label new volumes. i.e. by disabling autolabelling or by setting a PoolMaximumPoolVolumes to an upper limit on the size of the pool
|
|
<optiz0r> prune/purge only affects volumes in the full or used states. If a volume is half-full and still in append state, deleting the job will not cause that volume to be purged. Instead it will remain in the append state, the next job will continue writing in the second half of the tape, and only when it is filled does the expiration counter start
|
|
<optiz0r> if you know you have half-full volumes and you're about to delete the contents of that volume, you would want to manually set the VolStatus to Full and then prune it
|
|
<optiz0r> <kenwoodfox_> Do my incremental, difs and fulls have to line up? is that the way to do it? << I keep each successive level for longer than the interval of the parent level. e.g. if I take monthly fulls, weekly diffs, and daily incrementals, I would retain the incrementals for at least 5 weeks, and incrementals for at least 10 days
|
|
<optiz0r> also for each level, make sure you are retaining volumes for longer than the interval. If you had monthly fulls in a 30-day pool, you would only be keeping one backup at a time. and if the most recent backup failed (e.g. network interruptionm, server crash) then you no longer have any viable backups
|
|
|
|
Todo for major name-change
|
|
Purge all volumes in pools that have that name |