<?xml version="1.0" encoding="UTF-8"?><!-- generator="wordpress/2.0.11" -->
<rss version="2.0" 
	xmlns:content="http://purl.org/rss/1.0/modules/content/">
<channel>
	<title>Comments on: Comprehensive Desktop Backup</title>
	<link>http://www.linux-mag.com/id/4489/</link>
	<description>Open Source, Open Standards</description>
	<pubDate>Sun, 23 Nov 2008 10:37:21 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.0.11</generator>

	<item>
		<title>by: Mickey Gee</title>
		<link>http://www.linux-mag.com/id/4489/#comment-662</link>
		<pubDate>Wed, 05 Dec 2007 04:46:52 +0000</pubDate>
		<guid>http://www.linux-mag.com/id/4489/#comment-662</guid>
					<description>&#62;By Alex Rawls November 29th, 2007 at 5:10 pm  	 
&#62;
&#62;I can’t agree with you more about keeping things &#62;simple stupid... Then there is no down time while &#62;you recover.
&#62;
&#62;Also, what are peoples’ thoughts on Unison as a &#62;backup program? 
&#62;
Unison rocks.  I've been using it for years to keep my home dir on my laptop and office computers in sync.  Rock solid.  Quick.  Logical.  I recommend it.</description>
		<content:encoded><![CDATA[<p>&gt;By Alex Rawls November 29th, 2007 at 5:10 pm<br />
&gt;<br />
&gt;I can’t agree with you more about keeping things &gt;simple stupid&#8230; Then there is no down time while &gt;you recover.<br />
&gt;<br />
&gt;Also, what are peoples’ thoughts on Unison as a &gt;backup program?<br />
&gt;<br />
Unison rocks.  I&#8217;ve been using it for years to keep my home dir on my laptop and office computers in sync.  Rock solid.  Quick.  Logical.  I recommend it.
</p>
]]></content:encoded>
				</item>
	<item>
		<title>by: Paul Wiesen</title>
		<link>http://www.linux-mag.com/id/4489/#comment-658</link>
		<pubDate>Fri, 30 Nov 2007 19:53:09 +0000</pubDate>
		<guid>http://www.linux-mag.com/id/4489/#comment-658</guid>
					<description>I've had great success getting back up and running after a hard drive failure with a DVD created using Mondo Restore. Mondo along with rsnapshot has saved me many headaches.</description>
		<content:encoded><![CDATA[<p>I&#8217;ve had great success getting back up and running after a hard drive failure with a DVD created using Mondo Restore. Mondo along with rsnapshot has saved me many headaches.
</p>
]]></content:encoded>
				</item>
	<item>
		<title>by: Daryl Caudill</title>
		<link>http://www.linux-mag.com/id/4489/#comment-655</link>
		<pubDate>Fri, 30 Nov 2007 14:59:14 +0000</pubDate>
		<guid>http://www.linux-mag.com/id/4489/#comment-655</guid>
					<description>&#62;&#62; No backup of OS - its easier to reload than restore

&#62; I use gentoo stage4 backup and it was lot easier to just boot with liveCD and run tar extract than installing gentoo

Agreed.  When I'm done building a server, and it's ready for use, I always boot it with SystemRescueCD live cd, and backup the partition table and partitions (Partimage) to a writable DVD.  Takes about an hour.

I can build a new replacement server in about an hour, online and ready for data restore.</description>
		<content:encoded><![CDATA[<p>&gt;&gt; No backup of OS - its easier to reload than restore</p>
<p>&gt; I use gentoo stage4 backup and it was lot easier to just boot with liveCD and run tar extract than installing gentoo</p>
<p>Agreed.  When I&#8217;m done building a server, and it&#8217;s ready for use, I always boot it with SystemRescueCD live cd, and backup the partition table and partitions (Partimage) to a writable DVD.  Takes about an hour.</p>
<p>I can build a new replacement server in about an hour, online and ready for data restore.
</p>
]]></content:encoded>
				</item>
	<item>
		<title>by: Malahal Naineni</title>
		<link>http://www.linux-mag.com/id/4489/#comment-654</link>
		<pubDate>Fri, 30 Nov 2007 02:16:20 +0000</pubDate>
		<guid>http://www.linux-mag.com/id/4489/#comment-654</guid>
					<description>&#62;&#62; No backup of OS - its easier to reload than restore

I use gentoo stage4 backup and it was lot easier to just boot with liveCD and run tar extract than installing gentoo (maybe installing other distro's too). More over, installing all the packages you need and customizing them or configuring them for you would take lot more time.</description>
		<content:encoded><![CDATA[<p>&gt;&gt; No backup of OS - its easier to reload than restore</p>
<p>I use gentoo stage4 backup and it was lot easier to just boot with liveCD and run tar extract than installing gentoo (maybe installing other distro&#8217;s too). More over, installing all the packages you need and customizing them or configuring them for you would take lot more time.
</p>
]]></content:encoded>
				</item>
	<item>
		<title>by: johneeboy3</title>
		<link>http://www.linux-mag.com/id/4489/#comment-653</link>
		<pubDate>Fri, 30 Nov 2007 01:36:47 +0000</pubDate>
		<guid>http://www.linux-mag.com/id/4489/#comment-653</guid>
					<description>Alex,

Unison was not really designed as a backup tool. It is primarily a tool for maintaining mirrored data.

In that (latter) capacity, it's great.

I used it for a few months as a backup tool, but it has problems. It was a while ago now, so I'm sorry but I can't be more specific as to why - my bad memory!

See my post above about BackupPC. It's fantastic.

Cheers

John</description>
		<content:encoded><![CDATA[<p>Alex,</p>
<p>Unison was not really designed as a backup tool. It is primarily a tool for maintaining mirrored data.</p>
<p>In that (latter) capacity, it&#8217;s great.</p>
<p>I used it for a few months as a backup tool, but it has problems. It was a while ago now, so I&#8217;m sorry but I can&#8217;t be more specific as to why - my bad memory!</p>
<p>See my post above about BackupPC. It&#8217;s fantastic.</p>
<p>Cheers</p>
<p>John
</p>
]]></content:encoded>
				</item>
	<item>
		<title>by: johneeboy3</title>
		<link>http://www.linux-mag.com/id/4489/#comment-652</link>
		<pubDate>Fri, 30 Nov 2007 01:34:48 +0000</pubDate>
		<guid>http://www.linux-mag.com/id/4489/#comment-652</guid>
					<description>BackupPC is a fantastic backup tool, which gets precious little air time.

http://backuppc.sourceforge.net/

Key points:
  * It's GPL
  * It's rsync-based, and written in perl.
  * It will back up Windows, Linux or OSX machines
  * It 'pools' identical files, saving massive amounts of space on the backup host
  * It compresses all backed up files.
  * It's extremely flexible. Basically anything can be (re)configured.
  * It facilitates backing up fixed (desktop) and often disconnected (laptop) machines.
  * It has a great web-based interface, allowing you to drill down to individual files in the backup as well as see the complete change history of a backed up file. On top of that, restores only take a few clicks and the file(s) are restored automatically back onto the target.

I've been using backuppc in a production environment for 3+ years and it 'just works' and never fails.

I currently have over 920GB of files backed up across more than 10 machines. Due to pooling and compression, the backup pool is only about 160GB!!!

Simply fantastic and highly recommended. It's saved my IT life on numerous occasions. I've reviewed most backup systems and nothing else comes close to BackupPC.

Cheers

John</description>
		<content:encoded><![CDATA[<p>BackupPC is a fantastic backup tool, which gets precious little air time.</p>
<p><a href="http://backuppc.sourceforge.net/" rel="nofollow">http://backuppc.sourceforge.net/</a></p>
<p>Key points:<br />
  * It&#8217;s GPL<br />
  * It&#8217;s rsync-based, and written in perl.<br />
  * It will back up Windows, Linux or OSX machines<br />
  * It &#8216;pools&#8217; identical files, saving massive amounts of space on the backup host<br />
  * It compresses all backed up files.<br />
  * It&#8217;s extremely flexible. Basically anything can be (re)configured.<br />
  * It facilitates backing up fixed (desktop) and often disconnected (laptop) machines.<br />
  * It has a great web-based interface, allowing you to drill down to individual files in the backup as well as see the complete change history of a backed up file. On top of that, restores only take a few clicks and the file(s) are restored automatically back onto the target.</p>
<p>I&#8217;ve been using backuppc in a production environment for 3+ years and it &#8216;just works&#8217; and never fails.</p>
<p>I currently have over 920GB of files backed up across more than 10 machines. Due to pooling and compression, the backup pool is only about 160GB!!!</p>
<p>Simply fantastic and highly recommended. It&#8217;s saved my IT life on numerous occasions. I&#8217;ve reviewed most backup systems and nothing else comes close to BackupPC.</p>
<p>Cheers</p>
<p>John
</p>
]]></content:encoded>
				</item>
	<item>
		<title>by: Alex Rawls</title>
		<link>http://www.linux-mag.com/id/4489/#comment-651</link>
		<pubDate>Fri, 30 Nov 2007 00:10:09 +0000</pubDate>
		<guid>http://www.linux-mag.com/id/4489/#comment-651</guid>
					<description>I can't agree with you more about keeping things simple stupid; however, RAID has its place in KISS backups.  If you're only talking about mission critical files (meaning relatively little storage space required), then RAID-1 is your best friend.  Though it costs more (a $15 PCI card for older machines, and another hard drive), you get a lot of redundancy, in fact 100% times the number of drives in the RAID.  And if you RAID the entirety of both (or all) drives including your OS, then if a drive dies you can just take that bad one out and keep on chugging while you order another drive.  Then there is no down time while you recover.

Also, what are peoples’ thoughts on Unison as a backup program?</description>
		<content:encoded><![CDATA[<p>I can&#8217;t agree with you more about keeping things simple stupid; however, RAID has its place in KISS backups.  If you&#8217;re only talking about mission critical files (meaning relatively little storage space required), then RAID-1 is your best friend.  Though it costs more (a $15 PCI card for older machines, and another hard drive), you get a lot of redundancy, in fact 100% times the number of drives in the RAID.  And if you RAID the entirety of both (or all) drives including your OS, then if a drive dies you can just take that bad one out and keep on chugging while you order another drive.  Then there is no down time while you recover.</p>
<p>Also, what are peoples’ thoughts on Unison as a backup program?
</p>
]]></content:encoded>
				</item>
	<item>
		<title>by: John Fisher</title>
		<link>http://www.linux-mag.com/id/4489/#comment-649</link>
		<pubDate>Thu, 29 Nov 2007 21:07:00 +0000</pubDate>
		<guid>http://www.linux-mag.com/id/4489/#comment-649</guid>
					<description>Robin, I had a bad experience at a previous job with an wildly-overcomplicated RAID system with failover, but I also just plain think its too complex for simple stuff like backups. It has a place of course for high-speed servers and such, but my thinking about backups is keep it simple and dumb. I prefer hot spares to RAID enabled boxes for smaller servers too. When in doubt pick the simpler low-performance low-maintenance solution.
JF</description>
		<content:encoded><![CDATA[<p>Robin, I had a bad experience at a previous job with an wildly-overcomplicated RAID system with failover, but I also just plain think its too complex for simple stuff like backups. It has a place of course for high-speed servers and such, but my thinking about backups is keep it simple and dumb. I prefer hot spares to RAID enabled boxes for smaller servers too. When in doubt pick the simpler low-performance low-maintenance solution.<br />
JF
</p>
]]></content:encoded>
				</item>
	<item>
		<title>by: Robin Bowes</title>
		<link>http://www.linux-mag.com/id/4489/#comment-647</link>
		<pubDate>Thu, 29 Nov 2007 20:05:30 +0000</pubDate>
		<guid>http://www.linux-mag.com/id/4489/#comment-647</guid>
					<description>Why avoid RAID?</description>
		<content:encoded><![CDATA[<p>Why avoid RAID?
</p>
]]></content:encoded>
				</item>
	<item>
		<title>by: John Fisher</title>
		<link>http://www.linux-mag.com/id/4489/#comment-645</link>
		<pubDate>Thu, 29 Nov 2007 18:35:48 +0000</pubDate>
		<guid>http://www.linux-mag.com/id/4489/#comment-645</guid>
					<description>This is quite similar to the strategy I use at work to backup a small software office. Once we dumped the big tape drive, I took some old PIII generic boxes and created two backup servers, one for local and one at the home office in another town. (We have no IT dept, thank goodness, so individuals have to be somewhat responsible.)

There is a canonical user and filename regex for all machines. Each development server tarballs the important areas and ships them over to the local backup server. 
Each user is allowed a home on the backup server. (desktop data is not mission critical) users are responsible for tarballing their own data and shipping it to their home on the backup server. I use a Ubuntu/Debian package called "simple network backup" to do incremental backups of my own desktop, but a simple cron job would also do.
The backup server untars the mission critical tarballs and rsyncs them to the backup server offsite over an encrypted link. We do daily, weekly, monthly, yearly archiving using cron jobs.

Important KISS principles:
Use standard Linux components- perl, ftp, ssh, tar, rsync, cron
Use (otherwise) obsolete machines with cheap non-RAID hard drives
Avoid automatic restore - just save directory trees and expect to manually restore if necessary
Critical data is stored in three places, including off-site
No tape
No IT dept
No backup of OS - its easier to reload than restore
If a machine is awake and has a file of the right filename at the right location, back it up, else ignore
Avoid RAID

JF</description>
		<content:encoded><![CDATA[<p>This is quite similar to the strategy I use at work to backup a small software office. Once we dumped the big tape drive, I took some old PIII generic boxes and created two backup servers, one for local and one at the home office in another town. (We have no IT dept, thank goodness, so individuals have to be somewhat responsible.)</p>
<p>There is a canonical user and filename regex for all machines. Each development server tarballs the important areas and ships them over to the local backup server.<br />
Each user is allowed a home on the backup server. (desktop data is not mission critical) users are responsible for tarballing their own data and shipping it to their home on the backup server. I use a Ubuntu/Debian package called &#8220;simple network backup&#8221; to do incremental backups of my own desktop, but a simple cron job would also do.<br />
The backup server untars the mission critical tarballs and rsyncs them to the backup server offsite over an encrypted link. We do daily, weekly, monthly, yearly archiving using cron jobs.</p>
<p>Important KISS principles:<br />
Use standard Linux components- perl, ftp, ssh, tar, rsync, cron<br />
Use (otherwise) obsolete machines with cheap non-RAID hard drives<br />
Avoid automatic restore - just save directory trees and expect to manually restore if necessary<br />
Critical data is stored in three places, including off-site<br />
No tape<br />
No IT dept<br />
No backup of OS - its easier to reload than restore<br />
If a machine is awake and has a file of the right filename at the right location, back it up, else ignore<br />
Avoid RAID</p>
<p>JF
</p>
]]></content:encoded>
				</item>
</channel>
</rss>
