rsync Backup

{page-title}
{html}
<DIV id="vref">
<P class="vtitle">
<IMG src="http://wiki.r1soft.com/download/attachments/4227111/pdf_icon.gif" style="margin: 0pt 10px 0pt 0pt; float: left;"></IMG>
<A href="http://wiki.r1soft.com/pages/doexportpage.action?pageId=5079288&type=TYPE_PDF">rsync Backup</A>
</P>
</DIV>{html} {excerpt}Evaluating rsync Backup for Servers{excerpt}

rsync is an effecient method for synchronizing a set of files over the network.&nbsp; rsync can compute the differences in two file trees effeciently using an ingenious specialized method of [computing deltas|TP:Categories of Backup Software] [using checksums|TP:Computing Deltas - Check Sums]called the rsync algortithm.&nbsp; Deltas are the parts of files that have changed from one version of a file set to another.&nbsp; rsync defines deltas as variable length blocks or chunks of data.&nbsp;

Using check sums to determine the differences between files is nothing new.&nbsp; What rsync added is the idea of a weak and strong checksum.&nbsp; For each file rsync breaks the file into small fixed size chunks of data called blocks.&nbsp; Each block of all files are read into memory where rsync computes the weak check sum ([adler-32|http://en.wikipedia.org/wiki/Adler-32]) of the block.&nbsp; The adler-32 check sum requires very little CPU time to compute.&nbsp; The weak check sum of a new block in a file is compared ot the weak check sum of a block in the older version of a file.&nbsp; If the weak check sum is different the block has changed and is copied from the source to the destination.

By its design the adler-32 check sum is "weak" meaning there is a fair chance two blocks of data that are Not the same could actually compute to the same check sum.&nbsp; This would be bad as if this happened rsync would not detect that the block changed and the file would be corrupt.&nbsp; To overcome this matching weak check sums are verified with a strong check sum ([MD4|http://en.wikipedia.org/wiki/MD4]) which is almost impossible to generate a match on two different blocks.&nbsp; If the strong check sum is the same then the block is determined to not have changed.&nbsp; If it is different the block is updated in the destination file set and is determined to have changed.

h4. What rsync Does Well

The whole idea of the rsync algorithm is to copy as little data across a slow network link as possible to update an older version of a set of files on another computer.&nbsp; rsync does a great job of this as it's guaranteed to only send deltas or different parts of files that have actually changed. &nbsp;

In doing this rsync aims to reduce CPU time needed to compute check sums.&nbsp; This is acomplished as rsync avoids computing the more expensive string check sum (MD4) as much as possible.&nbsp; And remember deltas are just what parts of the file changed.&nbsp; This is a great idea and works very well.

h4. rsync Backup Resources

rsync backup recipes - [http://www.mikerubel.org/computers/rsync_snapshots/]

incremental backups with rdiff - [http://www.gnu.org/savannah-checkouts/non-gnu/rdiff-backup/]

rsync incremental backup script - [http://www.rustyparts.com/ribs.php]

official rsync examples - [http://samba.anu.edu.au/rsync/examples.html]

rsync backup documentation written by Kevin Korb - [http://www.sanitarium.net/golug/rsync_backups.html]

h4. History of rsync

Did you know the author of [rsync|http://samba.anu.edu.au/rsync/] Andrew Tridgell did not intend rsync to be used as a backup application?&nbsp; Instead he developed rsync for general purpose file transfer and mirroring over slow WAN links.&nbsp; Tridgell explains in [his PhD thesis paper|http://samba.org/~tridge/phd_thesis.pdf] that while the rsync algorithm would be highly useful for optimizing tape backup applications the rsync application itself is not intended to be used for backups.

h2. rsync Challenges

*rsync is File Based and Not Block Based*

Another challenge is that rsync works at the [level of files and directories|TP:File Based Backup Technology]like any other program.&nbsp; This might seem confusing considering the previous paragraphs discussed how rsync divides files into blocks.&nbsp; What this means is that rsync examines each file and must create an index of the path to every file and directory on the server.&nbsp; For a server with a fairly large number of files this can be an enormous task all on its own.&nbsp; It's also the reason people often have problems with rsync's consumption of large amounts of memory on larger servers.&nbsp; This is different from [block based backup software|TP:Block Based Backup Technology] that bypasses the file system to read low level disk or volume blocks.

*rsync Keeps only One Copy of Data*

On its own rsync keeps only one copy of the data.&nbsp; A basic feature of backup and archiving is the ability to keep multiple historical archive points as defined by some policy.&nbsp; There are some very clever scripts that combine&nbsp; rsync with Linux file system hard links to create an effective system for having multiple rsync archive points.&nbsp;&nbsp; A great resource for how to script incremental backups with rsync is [http://www.mikerubel.org/computers/rsync_snapshots/]

*rsync Is Not Able to Get a Consistent Copy of Data*

If you use rsync to copy or update a data set you know for sure is not being written to by any other users then your rsync copy is completely consistent. &nbsp; This is typically Not the case in practice.&nbsp; Typically when rsync is used for backup purposes it is run on a set of files that are being written to by other users and programs on the server.&nbsp; If a file is written to why rsync is reading it the file will likely be corrupt.&nbsp; If files are being added or removed from the server while rsync is running rsync will see some of the changes and miss others causing the file set to not be consistent.&nbsp;

The longer rsync needs to run for the more this is a problem.&nbsp; Consider how many changes are made to a server while rsync runs for 2 or 4 or 8 hours?&nbsp; What writes and changes made it into the backup?&nbsp; Which ones didn't?

h2. How is Continuous Data Protection Different?

Continuous Data Protection as a technology is different in three primary ways.

*CDP Uses a near-Continuous Method of Computing Deltas*

As you read above a major drawback to rysync is the time and Disk I/O work needed to compute deltas using check sums.&nbsp; Continuous Data Protection solves this problem with a proprietary device driver that can [track or see changes to the disk or volume|TP:Computing Deltas - near-Continuous (CDP)] at a very low level while the system is running.&nbsp;

Imagine if you will that the data on your disk is kind of like the earth's surface.&nbsp; Most of the earth's surface is not changing.&nbsp; At the same time forests are being cut down.&nbsp; Buildings and roads constructed.&nbsp; Houses being re-painted and changed.&nbsp; Every time rsync needs to make a map of your files or hard disk it must walk the earth to identify all the little changes.&nbsp; This takes forever of course. &nbsp; Alternatively Continuous Data Protection works as if it was a giant database that got instantly updated every time a change is made.&nbsp; Ask it at any time what the map looks like and it has an instant up to date copy\!

For more information see:&nbsp;[TP:Computing Deltas - near-Continuous (CDP)]

*CDP Is Block Based and Works at the Level of the Linux Kernel*

Continuous Data Protection is [block based|TP:Block Based Backup Technology]bypassing the file system.&nbsp; It is not concerned with the nunber of files.&nbsp; It does not index file on the file system.&nbsp; It only reads deltas.&nbsp; It knows what the deltas are before the backup operation starts because of the [near-Continuous method for computing deltas|TP:Computing Deltas - near-Continuous (CDP)].&nbsp;

*CDP Creates a Point-in-Time Snapshot of the Volume&nbsp;*

In order to read data below the file system at the block level their must be a perfectly consistent point-in-time snapshot of the disk all while the system is running.&nbsp; R1Soft uses Windows Volume Shadow Copy to create snapshots on Windows servers and a proprietary block device snapshot driver in Linux.&nbsp; The Linux block device driver can be inserted into an existing block device and creates point-in-time copy-on-wright snapshots.&nbsp;

This means that the backup appears just as your system did at that point in time.&nbsp; For example your very First backup with R1Soft will take a long time as it must copy all of the data the first time of source&nbsp; During this time on a very large data set it could take hours or even days for users sending data over slow Wide Area networks.&nbsp; During all of that time the Linux and Windows CDP Agents are reading data from a consistent point-in-time snapshot that looks exactly like the disk did when the backup started.

*CDP Uses Virtual Full Backups for Unlimited Recovery Point Archiving*

CDP works by only doing a full backup once.&nbsp; Once and only once.&nbsp; From then an every backup operation is really a "synchronization" where deltas are copied.&nbsp; Through proprietary technology called the "Disk Safe" R1Soft is able to store deltas in a highly efficient manner. &nbsp; Only block level deltas are ever stored.&nbsp; Furthermore they can be optinally compressed.&nbsp; Each backup or synchronization appears as a vritually Full Backup even though it's made up of only deltas.&nbsp; For more details see: [TP:Backup Method - Virtual Full Backup]\\

h2. Technology Comparison

|| || rsync \\ || R1Soft CDP \\ ||
| Backup Software Technology Category \\ | [Legacy|TP:Categories of Backup Software] | [Continuous Data Protection|TP:Categories of Backup Software] \\ |
| Backup Method Used \\ | [Incremental|TP:Backup Method - Incremental Backup]&nbsp; (only one copy is kept) \\ | [Virtual Full|TP:Backup Method - Virtual Full Backup]\\ |
| Data Lost In Disaster \\ | Days \\ | Minutes \\ |
| File Or Block Level \\ | [File Level|TP:File Based Backup Technology]\\ | [Block Level|TP:Block Based Backup Technology] \\ |
| Method for Computing Deltas \\ | [Check Sums|TP:Computing Deltas - Check Sums]\\ | [CDP|TP:Computing Deltas - near-Continuous (CDP)]\\ |
| Backup Window Length \\ | {color:#000000}Hours or Days{color}&nbsp; \\ | {color:#000000}Minutes when CDP is in Sync{color}&nbsp; \\ |
| Online Backup (Snapshots)&nbsp; \\ | No \\ | {color:#000000}Yes Integral Part of the Product{color}&nbsp; \\ |
| Bare-Metal Restore \\ | No \\ | Yes \\ |
| Backup Window Critical Path \\ | {color:#000000}= Time to Read All Data from Disks{color} \\
&nbsp; (needed to compute checksums) \\ | {color:#000000}= Only Time to Read Deltas{color}&nbsp; \\ |
| Date Invented&nbsp; \\ | 1999&nbsp;&nbsp; \\ | 2005&nbsp; \\ |
| Core Technological Concept \\ | Copy small set of files \\
Use on slow Wide Area Network links \\
Written by Andrew Tridgell \\
Project for [PhD Thesis Paper|http://samba.org/~tridge/phd_thesis.pdf]\\ | Minimal Backup Window \\
Little or No Load on Production Server \\
Virtual Full Backups \\
Continuous Data Protection \\
Proprietary Device Drivers \\
Web Interface \\
Easy of Deployment \\
Manageability \\
Data Centers \\

|| || rsync || R1Soft CDP ||
| Easy To Use Web Interface \\ | No&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; \\ | Yes |
| Point-in-Time Snapshots \\ | No \\ | Yes (does *NOT* require LVM) \\ |
| Bare-Metal Restore \\ | No \\ | Yes \\ |
| Real-Time Backup (backup without interrupting your work)&nbsp; \\ | No \\ | Yes \\ |
| Hourly or Minutely Backups \\ | No \\ | Yes \\ |
| Backup Archiving \\ | No \\ | Yes \\ |
| Virtual Full Backups \\ | No \\ | Yes \\ |
| Significantly Reduce Disk I/O Load with Linux Kernel Level CDP \\ | No \\ | Yes \\ |

h2. Windows Server Data Protection

|| || rsync || R1Soft CDP ||
| Easy To Use Web Interface \\ | No&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; \\ | Yes |
| Point-in-Time Snapshots \\ | No \\ | Yes \\ |
| Bare-Metal Restore \\ | No \\ | Yes \\ |
| Real-Time Backup (backup without interrupting your work)&nbsp; \\ | No \\ | Yes \\ |
| Hourly or Minutely Backups \\ | No \\ | Yes \\ |
| Backup Archiving \\ | No \\ | Yes \\ |
| Virtual Full Backups \\ | No \\ | Yes \\ |
| Significantly Reduce Disk I/O Load with R1Soft Windows CDP Volume Filter Driver \\ | No \\ | Yes \\ |
| Windows Volume Shadow Copy Service Integration \\ | No | Yes \\ |
| Copy Windows Files from CIFS / Samba Network Share to Linux \\ | Yes \\ | No \\ |

h2. Special Features for Hosting Service Providers

|| || rsync \\ || R1Soft CDP \\ ||
| Easy To Use Web Interface \\ | No &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;&nbsp; \\ | Yes&nbsp;&nbsp; \\ |
| Multi-User System for Backup and Restore with Granular Permissions \\ | No \\ | Yes \\ |
| Customer self-managed backup and restore \\ | No \\ | Yes \\ |
| Quotas for Backup Storage \\ | No \\ | Yes \\ |
| Hosting Control Panel Integration \\ | No \\ | Yes \\ |
| Billing API for Service Providers \\ | No \\ | Yes \\ |
\\