This captures both the global zone and all the hosted zones simultaneously, and it - so far - works adequately.
But... performance isn't great (~7MB/sec from the older server, ~20MB/sec from the newer, all-gigabit network) and of course this doesn't even get me known-good crash-consistent backups of the (running) Oracle databases.
Is there a "good" way to backup an entire ZFS-based system using, e.g. snapshots? The number of ZFS datasets will fluctuate as zones are added and removed, and I don't want to have to constantly adjust the backup job to handle that.
I particularly would like crash-consistent backups of the (cloned) Oracle databases - snapshots are good enough, but what's an appropriate way to handle this with Bareos? I dislike the notion of having cron jobs running on each host with no coordination w/Bareos; surely there's a better way?
Advice/suggestions welcome both on performance and best-practices.
Thanks,
-Adam
Marco,
I've been looking, but I can't find any listing of what the difference is between the commercial binary subscription and the open-source edition.
This is one difference, I assume there are others?
Thanks,
-Adam
![]() |
Adam Thompson Senior Systems Administrator voice: 204.789.9596 x24 | email: atho...@avant.ca | web: avant.ca |
--
You received this message because you are subscribed to a topic in the Google Groups "bareos-users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/bareos-users/gS8VftvK3lc/unsubscribe.
To unsubscribe from this group and all its topics, send an email to bareos-users...@googlegroups.com.
To post to this group, send email to bareos...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
When I took over this position, I had no idea how to restore from the existing tapes, and no DR recovery procedures were documented, so it seems this is not a major concern for the owner.
> With pre-done Bareos made its question of seconds or minute, in the other case
> hours, days ?
We would have to wait days to get another LTO6 tape drive...
> When you will be on rescue after disaster mode, do you really think
> that you will have time to got the time to recompile the sources,
See previous point... yes, we would have the time. LTO6 drives don't grow on trees. (Unfortunately.)
> And also run all the tests before putting in production.
> The Build, the Test, and quick delivery make the paid product easy.
That is a concern, yes.
> Also if you're doing your own compilation, there's less chance to upgrade
> for direct support in case of need.
Why? At the point of needing support, we would simply purchase the certified binaries first.
> I wouldn't have two different version. This drive free software to "evil core" things.
> and remove the warranty of FOSS. Now with the honest price of a subscription
> you got a build, certified, ready to use binary, and help the project
> to develop.
I agree with that; but I have difficulty justifying the business requirement to spend over C$1000 to get... the same thing I can get for free? FOSS (in this model) still doesn't mesh well with (some) business needs.
It would be much easier to justify spending double the amount on support that merely includes binaries - then I can point to the support and say "*that's* what we're paying for".
> ps : I'd like to check with you if the argument are enough or not.
If it was my own money: maybe. Very much maybe. It would depend on how easily I could afford it at the moment.
When it's not my money and I have to justify a business case: nope, not at all.
-Adam