[prev in list] [next in list] [prev in thread] [next in thread] 

List:       amanda-users
Subject:    Re: [Amanda-users] Getting started
From:       Gene Heskett <gene.heskett () verizon ! net>
Date:       2010-05-06 18:16:19
Message-ID: 201005061416.19978.gene.heskett () verizon ! net
[Download RAW message or body]

On Thursday 06 May 2010, jasonswett wrote:
>I read through a good portion of W.Curtis Preston's book but I'm afraid I'm
> too much of a beginner to know where to start with my solution. I think I
> want to use Amanda but everything I've read on the Amanda site is too
> narrow and specific.
>
>Here's what I have:
>
>I have two Linux boxes. One is a development server and one is a production
> server. Each is running a database instance that I want to back up.
>
Databases typically are dynamic, often requiring a wrapper script to stop the 
engine long enough to take a snapshot, then the engine is restarted and only 
the snapshot is backed up.  The exact details are going to be database 
dependent.  The wiki may have some help for that.

>I also have two Macs and one PC that are just workstations.

Amanda can easily back those up too as long as they are networked.

>I've looked into Amanda a little bit but just came out confused. I don't
> have any storage media but I'm willing to buy some.

Here is where I'm going to be a little heretical.  As a home user, there is 
no way I can justify spending thousands on a tape library and tapes for it. 
What I need is generally just a deleted file recovery, or worse, a full, bare 
metal ability for 2, maybe eventually 3 machines.  My home movies are the 
biggest single item since a 25 minute wedding can be 8Gb of files when its 
pulled from my TRV-460 camera.  So this gives you a hint of the size I'd 
need.  Your database snapshots might be 200x that size.

In my situation I've worn out several of the seagate dds2 changers, finding 
that neither the tapes nor the drives & changer mechanics really weren't all 
that dependable, so I did what I considered to be the next best thing that I 
could afford, and that was to add another big hard drive, setting it up as a 
30 tape virtual tape library.  It has for me, worked rather phenomenally 
well.  One drive failure occurred, and smartd warned me about it in plenty of 
time to go get another drive and rsync it to the old one.  With terrabyte 
drives being priced as commodity items now, and a raid 5 of them still less 
than a 5 pack of tapes for the high capacity tape drives, it makes perfect 
sense to forgo the tape scene altogether for _me_.

However if your backup needs include offsite storage for fire recovery 
purposes, and for a business this is almost a mandated condition, then I 
think I would still do hard drives in a raid5, but perhaps 3 kits of them, so 
that one can be offsite, 2 can be onsite with one in service continuously, 
and the 2nd and third being rsync'd from the 'in service' raid just before 
being traded with each other for the offsite storage.  In fact, if network 
bandwidth exists between your site and the offsite location, its possible 
that the only time the offsite raids would need to be transported would be 
for quicker fire recovery, bringing the newest rsync'd one back to become the 
main onsite unit.

As for me, the virtual tape is easily 100x faster at a file recovery because 
each file on the virtual tape is in reality a file on the hard drive, and I 
do not have to read through a batch of 4Gb tapes in series to find one file 
that may be in the 33rd file on a tape, its random access, going directly to 
the 33rd file to extract the file I need which is maybe 200 megs into that 
file.

As for setting amanda up, the wiki has many examples, so I won't get into 
that.

I felt your question was more of an "advice requester" than as a "how can I 
do this specific thing."

So the first question you'll need to decide on is "How much data do I need to 
back up in order to assure the viability of my business"   Multiply that by 
at least 3 so that you can have generations of your backups to fall back on 
since it may be weeks before you discover something has gone aglay in the 
databases.  With my 30 vtape and a daily amanda run, I have a month before 
something gone falls off the end of the chain and is truly non-recoverable. 

One thing to consider too, is that amanda does things differently in terms of 
scheduling the backups.  Amanda tries to equalize the media usage on a per 
run basis, and will need to be started slowly, adding more entries to the 
disklist per day, so that means the level 0 backups, whose frequency will be 
set in your amanda.conf, will be scattered about in the backup cycle, the 
idea being that amanda will itself, adjust when these backups are done so as 
to try and use the same amount of media every run.  So while it can be forced 
into the every friday night do it all framework, it will be better in the 
long run to let amanda do its own schedule adjustments.

>Can anyone help point me in the right direction?
>
>Thanks!
>Jason

Welcome to the list Jason.  This list is also read by the authors and 
developers of this great program, another way of saying that all of the above 
can be forgotten in an instant if I am over ruled because in reality, I'm 
just another user, but one who builds and uses the latest bleeding edge, 
sometimes daily snapshots, so I'm sometimes called the "canary".  I try to 
not let my 75 years get in the way too much.

-- 
Cheers, Gene
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Warp 7 -- It's a law we can live with.
[prev in list] [next in list] [prev in thread] [next in thread]