BackupPC |
This documentation describes BackupPC version 2.1.0, released on 20 Jun 2004.
BackupPC is a high-performance, enterprise-grade system for backing up Unix, Linux and WinXX PCs, desktops and laptops to a server's disk. BackupPC is highly configurable and easy to install and maintain.
Given the ever decreasing cost of disks and raid systems, it is now practical and cost effective to backup a large number of machines onto a server's local disk or network storage. For some sites this might be the complete backup solution. For other sites additional permanent archives could be created by periodically backing up the server to tape.
Features include:
BackupPC can also be configured to keep a certain number of incremental backups, and to keep a smaller number of very old incremental backups. (BackupPC does not support multi-level incremental backups, although it will in a future version.)
BackupPC's CGI interface ``fills-in'' incremental backups based on the last full backup, giving every backup a ``full'' appearance. This makes browsing and restoring backups easier.
The partial backup may be browsed or used to restore files just like a successful full or incremental backup.
With the rsync transfer method the partial backup is used to resume the next full backup, avoiding the need to retransfer the file data already in the partial backup.
BackupPC saves backups onto disk. Because of pooling you can relatively economically keep several weeks of old backups.
At some sites the disk-based backup will be adequate, without a secondary tape backup. This system is robust to any single failure: if a client disk fails or loses files, the BackupPC server can be used to restore files. If the server disk fails, BackupPC can be restarted on a fresh file system, and create new backups from the clients. The chance of the server disk failing can be made very small by spending more money on increasingly better RAID systems. However, there is still the risk of catastrophic events like fires or earthquakes that can destroy both the BackupPC server and the clients it is backing up if they are physically nearby.
Some sites might choose to do periodic backups to tape or cd/dvd. This backup can be done perhaps weekly using the archive function of BackupPC.
Other users have reported success with removable disks to rotate the BackupPC data drives, or using rsync to mirror the BackupPC data pool offsite.
http://backuppc.sourceforge.net
This page has links to the current documentation, the SourceForge project page and general information.
http://sourceforge.net/projects/backuppc
This page has links to the current releases of BackupPC.
The lists are archived on SourceForge and Gmane. The SourceForge lists are not always up to date and the searching is limited, so Gmane is a good alternative. See:
http://news.gmane.org/index.php?prefix=gmane.comp.sysutils.backup.backuppc http://sourceforge.net/mailarchive/forum.php?forum_id=503
You can subscribe to these lists by visiting:
http://lists.sourceforge.net/lists/listinfo/backuppc-announce http://lists.sourceforge.net/lists/listinfo/backuppc-users http://lists.sourceforge.net/lists/listinfo/backuppc-devel
The backuppc-announce list is moderated and is used only for important announcements (eg: new versions). It is low traffic. You only need to subscribe to one of backuppc-announce and backuppc-users: backuppc-users also receives any messages on backuppc-announce.
The backuppc-devel list is only for developers who are working on BackupPC. Do not post questions or support requests there. But detailed technical discussions should happen on this list.
To post a message to the backuppc-users list, send an email to
backuppc-users@lists.sourceforge.net
Do not send subscription requests to this address!
Unison is a utility that can do two-way, interactive, synchronization. See http://www.cis.upenn.edu/~bcpierce/unison.
Three popular open source packages that do tape backup are Amanda (http://www.amanda.org), afbackup (http://sourceforge.net/projects/afbackup), and Bacula (http://www.bacula.org). Amanda can also backup WinXX machines to tape using samba. These packages can be used as back ends to BackupPC to backup the BackupPC server data to tape.
Various programs and scripts use rsync to provide hardlinked backups. See, for example, Mike Rubel's site (http://www.mikerubel.org/computers/rsync_snapshots), JW Schultz's dirvish (http://www.pegasys.ws/dirvish (although as of June 2004 this link doesn't work)), Ben Escoto's rdiff-backup (http://rdiff-backup.stanford.edu), and John Bowman's rlbackup (http://www.math.ualberta.ca/imaging/rlbackup).
BackupPC provides many additional features, such as compressed storage, hardlinking any matching files (rather than just files with the same name), and storing special files without root privileges. But these other scripts provide simple and effective solutions and are worthy of consideration.
The new features planned for future releases of BackupPC are at http://backuppc.sourceforge.net/faq/roadMap.html.
Comments and suggestions are welcome.
BackupPC is free. I work on BackupPC because I enjoy doing it and I like to contribute to the open source community.
BackupPC already has more than enough features for my own needs. The main compensation for continuing to work on BackupPC is knowing that more and more people find it useful. So feedback is certainly appreciated, both positive and negative.
Beyond being a satisfied user and telling other people about it, everyone is encouraged to add links to http://backuppc.sourceforge.net (I'll see then via Google) or otherwise publicize BackupPC. Unlike the commercial products in this space, I have a zero budget (in both time and money) for marketing, PR and advertising, so it's up to all of you! Feel free to vote for BackupPC at http://freshmeat.net/projects/backuppc.
Also, everyone is encouraged to contribute patches, bug reports, feature and design suggestions, new code, FAQs, and documentation corrections or improvements. Answering questions on the mail list is a big help too.
BackupPC requires:
Several users have reported significantly better performance using reiser compared to ext3 for the BackupPC data file system. It is also recommended you consider either an LVM or raid setup (either in HW or SW; eg: 3Ware RAID5) so that you can expand the file system as necessary.
When BackupPC starts with an empty pool, all the backup data will be written to the pool on disk. After more backups are done, a higher percentage of incoming files will already be in the pool. BackupPC is able to avoid writing to disk new files that are already in the pool. So over time disk writes will reduce significantly (by perhaps a factor of 20 or more), since eventually 95% or more of incoming backup files are typically in the pool. Disk reads from the pool are still needed to do file compares to verify files are an exact match. So, with a mature pool, if a relatively fast client generates data at say 1MB/sec, and you run 4 simultaneous backups, there will be an average server disk load of about 4MB/sec reads and 0.2MB/sec writes (assuming 95% of the incoming files are in the pool). These rates will be perhaps 40% lower if compression is on.
The File::RsyncP module is available from http://perlrsync.sourceforge.net or CPAN. You'll need to install the File::RsyncP module if you want to use Rsync as a transport method.
See http://www.samba.org for source and binaries. It's pretty easy to fetch and compile samba, and just grab smbclient and nmblookup, without doing the installation. Alternatively, http://www.samba.org has binary distributions for most platforms.
For BackupPC to use Rsync you will also need to install the perl File::RsyncP module, which is available from http://perlrsync.sourceforge.net. Version 0.52 or later is required.
Here's one real example for an environment that is backing up 65 laptops with compression off. Each full backup averages 3.2GB. Each incremental backup averages about 0.2GB. Storing one full backup and two incremental backups per laptop is around 240GB of raw data. But because of the pooling of identical files, only 87GB is used. This is without compression.
Another example, with compression on: backing up 95 laptops, where each backup averages 3.6GB and each incremental averages about 0.3GB. Keeping three weekly full backups, and six incrementals is around 1200GB of raw data. Because of pooling and compression, only 150GB is needed.
Here's a rule of thumb. Add up the disk usage of all the machines you want to backup (210GB in the first example above). This is a rough minimum space estimate that should allow a couple of full backups and at least half a dozen incremental backups per machine. If compression is on you can reduce the storage requirements by maybe 30-40%. Add some margin in case you add more machines or decide to keep more old backups.
Your actual mileage will depend upon the types of clients, operating systems and applications you have. The more uniform the clients and applications the bigger the benefit from pooling common files.
For example, the Eudora email tool stores each mail folder in a separate file, and attachments are extracted as separate files. So in the sadly common case of a large attachment emailed to many recipients, Eudora will extract the attachment into a new file. When these machines are backed up, only one copy of the file will be stored on the server, even though the file appears in many different full or incremental backups. In this sense Eudora is a ``friendly'' application from the point of view of backup storage requirements.
An example at the other end of the spectrum is Outlook. Everything (email bodies, attachments, calendar, contact lists) is stored in a single file, which often becomes huge. Any change to this file requires a separate copy of the file to be saved during backup. Outlook is even more troublesome, since it keeps this file locked all the time, so it cannot be read by smbclient whenever Outlook is running. See the Limitations section for more discussion of this problem.
In addition to total disk space, you shold make sure you have plenty of inodes on your BackupPC data partition. Some users have reported running out of inodes on their BackupPC data partition. So even if you have plenty of disk space, BackupPC will report failures when the inodes are exhausted. This is a particular problem with ext2/ext3 file systems that have a fixed number of inodes when the file system is built. Use ``df -i'' to see your inode usage.
Some linux distributions now include BackupPC. The Debian distribution, supprted by Ludovic Drolez, can be found at http://packages.debian.org/backuppc; it should be included in the next stable Debian release. On Debian, BackupPC can be installed with the command:
apt-get install backuppc
In the future there might be packages for Gentoo and other linux flavors. If the packaged version is older than the released version then you will probably want to install the lastest version as described below.
Otherwise, manually fetching and installing BackupPC is easy. Start by downloading the latest version from http://backuppc.sourceforge.net. Hit the ``Code'' button, then select the ``backuppc'' or ``backuppc-beta'' package and download the latest version.
First off, there are three perl modules you should install. These are all optional, but highly recommended:
To build and install these packages, fetch the tar.gz file and then run these commands:
tar zxvf Archive-Zip-1.01.tar.gz cd Archive-Zip-1.01 perl Makefile.PL make make test make install
The same sequence of commands can be used for each module.
Now let's move onto BackupPC itself. After fetching BackupPC-2.1.0.tar.gz, run these commands as root:
tar zxf BackupPC-2.1.0.tar.gz cd BackupPC-2.1.0 perl configure.pl
In the future this release might also have patches available on the SourceForge site. These patch files are text files, with a name of the form
BackupPC-2.1.0plN.diff
where N is the patch level, eg: pl5 is patch-level 5. These patch files are cumulative: you only need apply the last patch file, not all the earlier patch files. If a patch file is available, eg: BackupPC-2.1.0pl5.diff, you should apply the patch after extracting the tar file:
# fetch BackupPC-2.1.0.tar.gz # fetch BackupPC-2.1.0pl5.diff tar zxf BackupPC-2.1.0.tar.gz cd BackupPC-2.1.0 patch -p0 < ../BackupPC-2.1.0pl5.diff perl configure.pl
A patch file includes comments that describe that bug fixes and changes. Feel free to review it before you apply the patch.
The configure.pl script also accepts command-line options if you wish to run it in a non-interactive manner. It has self-contained documentation for all the command-line options, which you can read with perldoc:
perldoc configure.pl
When you run configure.pl you will be prompted for the full paths of various executables, and you will be prompted for the following information:
On this installation, this is __BACKUPPCUSER__.
On this installation, this is __TOPDIR__.
On this installation, this is __INSTALLDIR__.
On this installation, this is __CGIDIR__.
After running configure.pl, browse through the config file, __INSTALLDIR__/conf/config.pl, and make sure all the default settings are correct. In particular, you will need to decide whether to use smb, tar or rsync transport (or whether to set it on a per-PC basis) and set the relevant parameters for that transport method. See the section Client Setup for more details.
The file __TOPDIR__/conf/hosts contains the list of clients to backup. BackupPC reads this file in three cases:
Whenever you change the hosts file (to add or remove a host) you can either do a kill -HUP BackupPC_pid or simply wait until the next regular wakeup period.
Each line in the hosts file contains three fields, separated by white space:
Please read the section How BackupPC Finds Hosts.
In certain cases you might want several distinct clients to refer to the same physical machine. For example, you might have a database you want to backup, and you want to bracket the backup of the database with shutdown/restart using $Conf{DumpPreUserCmd} and $Conf{DumpPostUserCmd}. But you also want to backup the rest of the machine while the database is still running. In the case you can specify two different clients in the host file, using any mnemonic name (eg: myhost_mysql and myhost), and use $Conf{ClientNameAlias} in myhost_mysql's config.pl to specify the real host name of the machine.
You only need to set DHCP to 1 if your client machine doesn't respond to the NetBios multicast request:
nmblookup myHost
but does respond to a request directed to its IP address:
nmblookup -A W.X.Y.Z
If you do set DHCP to 1 on any client you will need to specify the range of DHCP addresses to search is specified in $Conf{DHCPAddressRanges}.
Note also that the $Conf{ClientNameAlias} feature does not work for clients with DHCP set to 1.
The first non-comment line of the hosts file is special: it contains the names of the columns and should not be edited.
Here's a simple example of a hosts file:
host dhcp user moreUsers farside 0 craig jim,dave larson 1 gary andy
Two methods for getting backup data from a client are supported: smb and tar. Smb or rsync are the preferred methods for WinXX clients and rsync or tar are the preferred methods for linux/unix clients.
The transfer method is set using the $Conf{XferMethod} configuration setting. If you have a mixed environment (ie: you will use smb for some clients and tar for others), you will need to pick the most common choice for $Conf{XferMethod} for the main config.pl file, and then override it in the per-PC config file for those hosts that will use the other method. (Or you could run two completely separate instances of BackupPC, with different data directories, one for WinXX and the other for linux/unix, but then common files between the different machine types will duplicated.)
Here are some brief client setup notes:
If you want to use rsyncd for WinXX clients you can find a pre-packaged zip file on http://backuppc.sourceforge.net. The package is called cygwin-rsync. It contains rsync.exe, template setup files and the minimal set of cygwin libraries for everything to run. The README file contains instructions for running rsync as a service, so it starts automatically everytime you boot your machine.
If you build your own rsync, for rsync 2.6.2 it is strongly recommended you apply the patch in the cygwin-rsync package on http://backuppc.sourceforge.net. This patch adds the --checksum-seed option for checksum caching, and also sends all errors to the client, which is important so BackupPC can log all file access errors.
Otherwise, to use SMB, you need to create shares for the data you want to backup. Open ``My Computer'', right click on the drive (eg: C), and select ``Sharing...'' (or select ``Properties'' and select the ``Sharing'' tab). In this dialog box you can enable sharing, select the share name and permissions. Many machines will be configured by default to share the entire C drive as C$ using the administrator password.
If this machine uses DHCP you will also need to make sure the NetBios name is set. Go to Control Panel|System|Network Identification (on Win2K) or Control Panel|System|Computer Name (on WinXP). Also, you should go to Control Panel|Network Connections|Local Area Connection|Properties|Internet Protocol (TCP/IP)|Properties|Advanced|WINS and verify that NetBios is not disabled.
The relevant configuration settings are $Conf{SmbShareName}, $Conf{SmbShareUserName}, $Conf{SmbSharePasswd}, $Conf{SmbClientPath}, $Conf{SmbClientFullCmd}, $Conf{SmbClientIncrCmd} and $Conf{SmbClientRestoreCmd}.
BackupPC needs to know the smb share user name and password for a client machine that uses smb. The user name is specified in $Conf{SmbShareUserName}. There are four ways to tell BackupPC the smb share password:
Placement and protection of the smb share password is a possible security risk, so please double-check the file and directory permissions. In a future version there might be support for encryption of this password, but a private key will still have to be stored in a protected place. Suggestions are welcome.
As an alternative to setting $Conf{XferMethod} to ``smb'' (using smbclient) for WinXX clients, you can use an smb network filesystem (eg: ksmbfs or similar) on your linux/unix server to mount the share, and then set $Conf{XferMethod} to ``tar'' (use tar on the network mounted file system).
Also, to make sure that file names with 8-bit characters are correctly transferred by smbclient you should add this to samba's smb.conf file for samba 2.x:
[global] # Accept the windows charset client code page = 850 character set = ISO8859-1
For samba 3.x this should instead be:
[global] unix charset = ISO8859-1
This setting should work for western europe. See http://www.oreilly.com/catalog/samba/chapter/book/ch08_03.html for more information about settings for other languages.
You can use either rsync, smb, or tar for linux/unix machines. Smb requires that the Samba server (smbd) be run to provide the shares. Since the smb protocol can't represent special files like symbolic links and fifos, tar and rsync are the better transport methods for linux/unix machines. (In fact, by default samba makes symbolic links look like the file or directory that they point to, so you could get an infinite loop if a symbolic link points to the current or parent directory. If you really need to use Samba shares for linux/unix backups you should turn off the ``follow symlinks'' samba config setting. See the smb.conf manual page.)
The requirements for each Xfer Method are:
The relevant configuration settings are $Conf{TarClientPath}, $Conf{TarShareName}, $Conf{TarClientCmd}, $Conf{TarFullArgs}, $Conf{TarIncrArgs}, and $Conf{TarClientRestoreCmd}.
The relevant configuration settings are $Conf{RsyncClientPath}, $Conf{RsyncClientCmd}, $Conf{RsyncClientRestoreCmd}, $Conf{RsyncShareName}, $Conf{RsyncArgs}, and $Conf{RsyncRestoreArgs}.
The relevant configuration settings are $Conf{RsyncdClientPort}, $Conf{RsyncdUserName}, $Conf{RsyncdPasswd}, $Conf{RsyncdAuthRequired}, $Conf{RsyncShareName}, $Conf{RsyncArgs}, and $Conf{RsyncRestoreArgs}. $Conf{RsyncShareName} is the name of an rsync module (ie: the thing in square brackets in rsyncd's conf file -- see rsyncd.conf), not a file system path.
Be aware that rsyncd will remove the leading '/' from path names in symbolic links if you specify ``use chroot = no'' in the rsynd.conf file. See the rsyncd.conf manual page for more information.
For linux/unix machines you should not backup ``/proc''. This directory contains a variety of files that look like regular files but they are special files that don't need to be backed up (eg: /proc/kcore is a regular file that contains physical memory). See $Conf{BackupFilesExclude}. It is safe to back up /dev since it contains mostly character-special and block-special files, which are correctly handed by BackupPC (eg: backing up /dev/hda5 just saves the block-special file information, not the contents of the disk).
Alternatively, rather than backup all the file systems as a single share (``/''), it is easier to restore a single file system if you backup each file system separately. To do this you should list each file system mount point in $Conf{TarShareName} or $Conf{RsyncShareName}, and add the --one-file-system option to $Conf{TarClientCmd} or add --one-file-system (note the different punctuation) to $Conf{RsyncArgs}. In this case there is no need to exclude /proc explicitly since it looks like a different file system.
Next you should decide whether to run tar over ssh, rsh or nfs. Ssh is the preferred method. Rsh is not secure and therefore not recommended. Nfs will work, but you need to make sure that the BackupPC user (running on the server) has sufficient permissions to read all the files below the nfs mount.
Ssh allows BackupPC to run as a privileged user on the client (eg: root), since it needs sufficient permissions to read all the backup files. Ssh is setup so that BackupPC on the server (an otherwise low privileged user) can ssh as root on the client, without being prompted for a password. There are two common versions of ssh: v1 and v2. Here are some instructions for one way to setup ssh. (Check which version of SSH you have by typing ``ssh'' or ``man ssh''.)
Note that if you run rsyncd (rsync daemon), ssh is not used. In this case, rsyncd provides its own authentication, but there is no encryption of network data. If you want encryption of network data you can use ssh to create a tunnel, or use a program like stunnel. If someone submits instructions I
Setup instructions for ssh are at http://backuppc.sourceforge.net/faq/ssh.html.
For WinXX machines BackupPC uses the NetBios name server to determine the IP address given the host name. For unix machines you can run nmbd (the NetBios name server) from the Samba distribution so that the machine responds to a NetBios name request. See the manual page and Samba documentation for more information.
Alternatively, you can set $Conf{NmbLookupFindHostCmd} to any command that returns the IP address given the host name.
Please read the section How BackupPC Finds Hosts for more details.
The installation contains an init.d backuppc script that can be copied to /etc/init.d so that BackupPC can auto-start on boot. See init.d/README for further instructions.
BackupPC should be ready to start. If you installed the init.d script, then you should be able to run BackupPC with:
/etc/init.d/backuppc start
(This script can also be invoked with ``stop'' to stop BackupPC and ``reload'' to tell BackupPC to reload config.pl and the hosts file.)
Otherwise, just run
__INSTALLDIR__/bin/BackupPC -d
as user __BACKUPPCUSER__. The -d option tells BackupPC to run as a daemon (ie: it does an additional fork).
Any immediate errors will be printed to stderr and BackupPC will quit. Otherwise, look in __TOPDIR__/log/LOG and verify that BackupPC reports it has started and all is ok.
Note: as of version 1.5.0, BackupPC no longer supports telnet to its TCP port. First off, a unix domain socket is used instead of a TCP port. (The TCP port can still be re-enabled if your installation has apache and BackupPC running on different machines.) Secondly, even if you still use the TCP port, the messages exchanged over this interface are now protected by an MD5 digest based on a shared secret (see $Conf{ServerMesgSecret}) as well as sequence numbers and per-session unique keys, preventing forgery and replay attacks.
You should verify that BackupPC is running by using BackupPC_serverMesg. This sends a message to BackupPC via the unix (or TCP) socket and prints the response.
You can request status information and start and stop backups using this interface. This socket interface is mainly provided for the CGI interface (and some of the BackupPC sub-programs use it too). But right now we just want to make sure BackupPC is happy. Each of these commands should produce some status output:
__INSTALLDIR__/bin/BackupPC_serverMesg status info __INSTALLDIR__/bin/BackupPC_serverMesg status jobs __INSTALLDIR__/bin/BackupPC_serverMesg status hosts
The output should be some hashes printed with Data::Dumper. If it looks cryptic and confusing, and doesn't look like an error message, then all is ok.
The jobs status should initially show just BackupPC_trashClean. The hosts status should produce a list of every host you have listed in __TOPDIR__/conf/hosts as part of a big cryptic output line.
You can also request that all hosts be queued:
__INSTALLDIR__/bin/BackupPC_serverMesg backup all
At this point you should make sure the CGI interface works since it will be much easier to see what is going on. That's our next subject.
The CGI interface script, BackupPC_Admin, is a powerful and flexible way to see and control what BackupPC is doing. It is written for an Apache server. If you don't have Apache, see http://www.apache.org.
There are two options for setting up the CGI interface: standard mode and using mod_perl. Mod_perl provides much higher performance (around 15x) and is the best choice if your Apache was built with mod_perl support. To see if your apache was built with mod_perl run this command:
httpd -l | egrep mod_perl
If this prints mod_perl.c then your Apache supports mod_perl.
Using mod_perl with BackupPC_Admin requires a dedicated Apache to be run as the BackupPC user (__BACKUPPCUSER__). This is because BackupPC_Admin needs permission to access various files in BackupPC's data directories. In contrast, the standard installation (without mod_perl) solves this problem by having BackupPC_Admin installed as setuid to the BackupPC user, so that BackupPC_Admin runs as the BackuPC user.
Here are some specifics for each setup:
You should be very careful about permissions on BackupPC_Admin and the directory __CGIDIR__: it is important that normal users cannot directly execute or change BackupPC_Admin, otherwise they can access backup files for any PC. You might need to change the group ownership of BackupPC_Admin to a group that Apache belongs to so that Apache can execute it (don't add ``other'' execute permission!). The permissions should look like this:
ls -l __CGIDIR__/BackupPC_Admin -swxr-x--- 1 __BACKUPPCUSER__ web 82406 Jun 17 22:58 __CGIDIR__/BackupPC_Admin
The setuid script won't work unless perl on your machine was installed with setuid emulation. This is likely the problem if you get an error saying such as ``Wrong user: my userid is 25, instead of 150'', meaning the script is running as the httpd user, not the BackupPC user. This is because setuid scripts are disabled by the kernel in most flavors of unix and linux.
To see if your perl has setuid emulation, see if there is a program called sperl5.6.0 (or sperl5.8.2 etc, based on your perl version) in the place where perl is installed. If you can't find this program, then you have two options: rebuild and reinstall perl with the setuid emulation turned on (answer ``y'' to the question ``Do you want to do setuid/setgid emulation?'' when you run perl's configure script), or switch to the mod_perl alternative for the CGI script (which doesn't need setuid to work).
To use mod_perl you need to run Apache as user __BACKUPPCUSER__. If you need to run multiple Apache's for different services then you need to create multiple top-level Apache directories, each with their own config file. You can make copies of /etc/init.d/httpd and use the -d option to httpd to point each http to a different top-level directory. Or you can use the -f option to explicitly point to the config file. Multiple Apache's will run on different Ports (eg: 80 is standard, 8080 is a typical alternative port accessed via http://yourhost.com:8080).
Inside BackupPC's Apache http.conf file you should check the settings for ServerRoot, DocumentRoot, User, Group, and Port. See http://httpd.apache.org/docs/server-wide.html for more details.
For mod_perl, BackupPC_Admin should not have setuid permission, so you should turn it off:
chmod u-s __CGIDIR__/BackupPC_Admin
To tell Apache to use mod_perl to execute BackupPC_Admin, add this to Apache's 1.x httpd.conf file:
<IfModule mod_perl.c> PerlModule Apache::Registry PerlTaintCheck On <Location /cgi-bin/BackupPC/BackupPC_Admin> # <--- change path as needed SetHandler perl-script PerlHandler Apache::Registry Options ExecCGI PerlSendHeader On </Location> </IfModule>
Apache 2.0.44 with Perl 5.8.0 on RedHat 7.1, Don Silvia reports that this works (with tweaks from Michael Tuzi):
LoadModule perl_module modules/mod_perl.so PerlModule Apache2
<Directory /path/to/cgi/> SetHandler perl-script PerlResponseHandler ModPerl::Registry PerlOptions +ParseHeaders Options +ExecCGI Order deny,allow Deny from all Allow from 192.168.0 AuthName "Backup Admin" AuthType Basic AuthUserFile /path/to/user_file Require valid-user </Directory>
There are other optimizations and options with mod_perl. For example, you can tell mod_perl to preload various perl modules, which saves memory compared to loading separate copies in every Apache process after they are forked. See Stas's definitive mod_perl guide at http://perl.apache.org/guide.
BackupPC_Admin requires that users are authenticated by Apache. Specifically, it expects that Apache sets the REMOTE_USER environment variable when it runs. There are several ways to do this. One way is to create a .htaccess file in the cgi-bin directory that looks like:
AuthGroupFile /etc/httpd/conf/group # <--- change path as needed AuthUserFile /etc/http/conf/passwd # <--- change path as needed AuthType basic AuthName "access" require valid-user
You will also need ``AllowOverride Indexes AuthConfig'' in the Apache httpd.conf file to enable the .htaccess file. Alternatively, everything can go in the Apache httpd.conf file inside a Location directive. The list of users and password file above can be extracted from the NIS passwd file.
One alternative is to use LDAP. In Apache's http.conf add these lines:
LoadModule auth_ldap_module modules/auth_ldap.so AddModule auth_ldap.c
# cgi-bin - auth via LDAP (for BackupPC) <Location /cgi-binBackupPC/BackupPC_Admin> # <--- change path as needed AuthType Basic AuthName "BackupPC login" # replace MYDOMAIN, PORT, ORG and CO as needed AuthLDAPURL ldap://ldap.MYDOMAIN.com:PORT/o=ORG,c=CO?uid?sub?(objectClass=*) require valid-user </Location>
If you want to disable the user authentication you can set $Conf{CgiAdminUsers} to '*', which allows any user to have full access to all hosts and backups. In this case the REMOTE_USER environment variable does not have to be set by Apache.
Alternatively, you can force a particular user name by getting Apache to set REMOTE_USER, eg, to hardcode the user to www you could add this to Apache's httpd.conf:
<Location /cgi-bin/BackupPC/BackupPC_Admin> # <--- change path as needed Setenv REMOTE_USER www </Location>
Finally, you should also edit the config.pl file and adjust, as necessary, the CGI-specific settings. They're near the end of the config file. In particular, you should specify which users or groups have administrator (privileged) access: see the config settings $Conf{CgiAdminUserGroup} and $Conf{CgiAdminUsers}. Also, the configure.pl script placed various images into $Conf{CgiImageDir} that BackupPC_Admin needs to serve up. You should make sure that $Conf{CgiImageDirURL} is the correct URL for the image directory.
See the section Fixing installation problems for suggestions on debugging the Apache authentication setup.
Starting with v2.0.0 the way hosts are discovered has changed. In most cases you should specify 0 for the DHCP flag in the conf/hosts file, even if the host has a dynamically assigned IP address.
BackupPC (starting with v2.0.0) looks up hosts with DHCP = 0 in this manner:
gethostbyname()
function. This should succeed for machines
that have fixed IP addresses that are known via DNS. You can manually
see whether a given host have a DNS entry according to perls'
gethostbyname function with this command:
perl -e 'print(gethostbyname("myhost") ? "ok\n" : "not found\n");'
gethostbyname()
fails, BackupPC then attempts a NetBios multicast to
find the host. Provided your client machine is configured properly,
it should respond to this NetBios multicast request. Specifically,
BackupPC runs a command of this form:
nmblookup myhost
If this fails you will see output like:
querying myhost on 10.10.255.255 name_query failed to find name myhost
If this success you will see output like:
querying myhost on 10.10.255.255 10.10.1.73 myhost<00>
Depending on your netmask you might need to specify the -B option to nmblookup. For example:
nmblookup -B 10.10.1.255 myhost
If necessary, experiment on the nmblookup command that will return the IP address of the client given its name. Then update $Conf{NmbLookupFindHostCmd} with any necessary options to nmblookup.
For hosts that have the DHCP flag set to 1, these machines are discovered as follows:
nmblookup -A W.X.Y.Z
where W.X.Y.Z is each candidate address from $Conf{DHCPAddressRanges}. Any host that has a valid NetBIOS name returned by this command (ie: matching an entry in the hosts file) will be backed up. You can modify the specific nmblookup command if necessary via $Conf{NmbLookupCmd}.
nmblookup myHost
but does respond to a request directed to its IP address:
nmblookup -A W.X.Y.Z
To disable backups for a client there are two special values for $Conf{FullPeriod} in that client's per-PC config.pl file:
This will still allow that client's old backups to be browsable and restorable.
To completely remove a client and all its backups, you should remove its entry in the conf/hosts file, and then delete the __TOPDIR__/pc/$host directory. Whenever you change the hosts file, you should send BackupPC a HUP (-1) signal so that it re-reads the hosts file. If you don't do this, BackupPC will automatically re-read the hosts file at the next regular wakeup.
Note that when you remove a client's backups you won't initially recover a lot of disk space. That's because the client's files are still in the pool. Overnight, when BackupPC_nightly next runs, all the unused pool files will be deleted and this will recover the disk space used by the client's backups.
The backup data directories contain large numbers of hardlinks. If you try to copy the pool the target directory will occupy a lot more space if the hardlinks aren't re-established.
The GNU cp program with the -a option is aware of hardlinks and knows to re-establish them. So GNU cp -a is the recommended way to copy the data directory and pool. Don't forget to stop BackupPC while the copy runs.
BackupPC_compressPool compresses all the files in the uncompressed pool (__TOPDIR__/pool) and moves them to the compressed pool (__TOPDIR__/cpool). It rewrites the files in place, so that the existing hardlinks are not disturbed.
The rest of this section discusses how to run BackupPC_compressPool.
BackupPC_compressPool takes three command line options:
Here are the recommended steps for running BackupPC_compressPool:
BackupPC_compressPool -t -r
The -t option (test mode) makes BackupPC_compressPool do all the steps, but not actually change anything. The -r option re-reads the compressed file and compares it against the original.
BackupPC_compressPool gives a status as it completes each 1% of the job. It also shows the cumulative compression ratio and estimated completion time. Once you are comfortable that things look ok, you can kill BackupPC_compressPool or wait for it to finish.
BackupPC_compressPool
You should put the output into a file and tail this file. (The running time could be twice as long as the test mode since the test mode file writes are immediately followed by an unlink, so in test mode it is likely the file writes never make it to disk.)
It is critical that BackupPC_compressPool runs to completion before re-starting BackupPC. Before BackupPC_compressPool completes, none of the existing backups will be in a consistent state. If you must stop BackupPC_compressPool for some reason, send it an INT or TERM signal and give it several seconds (or more) to clean up gracefully. After that, you can re-run BackupPC_compressPool and it will start again where it left off. Once again, it is critical that it runs to 100% completion.
After BackupPC_compressPool completes you should have a complete set of compressed backups (and your disk usage should be lower). You can now re-start BackupPC.
Please see the FAQ at http://backuppc.sourceforge.net/faq for debugging suggestions.
BackupPC supports several different methods for restoring files. The most convenient restore options are provided via the CGI interface. Alternatively, backup files can be restored using manual commands.
By selecting a host in the CGI interface, a list of all the backups for that machine will be displayed. By selecting the backup number you can navigate the shares and directory tree for that backup.
BackupPC's CGI interface automatically fills incremental backups with the corresponding full backup, which means each backup has a filled appearance. Therefore, there is no need to do multiple restores from the incremental and full backups: BackupPC does all the hard work for you. You simply select the files and directories you want from the correct backup vintage in one step.
You can download a single backup file at any time simply by selecting it. Your browser should prompt you with the file name and ask you whether to open the file or save it to disk.
Alternatively, you can select one or more files or directories in the currently selected directory and select ``Restore selected files''. (If you need to restore selected files and directories from several different parent directories you will need to do that in multiple steps.)
If you select all the files in a directory, BackupPC will replace the list of files with the parent directory. You will be presented with a screen that has three options:
Once you select ``Start Restore'' you will be prompted one last time with a summary of the exact source and target files and directories before you commit. When you give the final go ahead the restore operation will be queued like a normal backup job, meaning that it will be deferred if there is a backup currently running for that host. When the restore job is run, smbclient, tar, rsync or rsyncd is used (depending upon $Conf{XferMethod}) to actually restore the files. Sorry, there is currently no option to cancel a restore that has been started.
A record of the restore request, including the result and list of files and directories, is kept. It can be browsed from the host's home page. $Conf{RestoreInfoKeepCnt} specifies how many old restore status files to keep.
Note that for direct restore to work, the $Conf{XferMethod} must be able to write to the client. For example, that means an SMB share for smbclient needs to be writable, and the rsyncd module needs ``read only'' set to ``yes''. This creates additional security risks. If you only create read-only SMB shares (which is a good idea), then the direct restore will fail. You can disable the direct restore option by setting $Conf{SmbClientRestoreCmd}, $Conf{TarClientRestoreCmd} and $Conf{RsyncRestoreArgs} to undef.
When you select ``Download Zip File'' you should be prompted where to save the restore.zip file.
BackupPC does not consider downloading a zip file as an actual restore operation, so the details are not saved for later browsing as in the first case. However, a mention that a zip file was downloaded by a particular user, and a list of the files, does appear in BackupPC's log file.
Apart from the CGI interface, BackupPC allows you to restore files and directories from the command line. The following programs can be used:
__INSTALLDIR__/bin/BackupPC_zcat __TOPDIR__/pc/host/5/fc/fcraig/fexample.txt > example.txt
It's your responsibility to make sure the file is really compressed: BackupPC_zcat doesn't check which backup the requested file is from. BackupPC_zcat returns a non-zero status if it fails to uncompress a file.
The usage is:
BackupPC_tarCreate [-t] [-h host] [-n dumpNum] [-s shareName] [-r pathRemove] [-p pathAdd] [-b BLOCKS] [-w writeBufSz] files/directories...
The command-line files and directories are relative to the specified shareName. The tar file is written to stdout.
The required options are:
Other options are:
The -h, -n and -s options specify which dump is used to generate the tar archive. The -r and -p options can be used to relocate the paths in the tar archive so extracted files can be placed in a location different from their original location.
The usage is:
BackupPC_zipCreate [-t] [-h host] [-n dumpNum] [-s shareName] [-r pathRemove] [-p pathAdd] [-c compressionLevel] files/directories...
The command-line files and directories are relative to the specified shareName. The zip file is written to stdout.
The required options are:
Other options are:
The -h, -n and -s options specify which dump is used to generate the zip archive. The -r and -p options can be used to relocate the paths in the zip archive so extracted files can be placed in a location different from their original location.
Each of these programs reside in __INSTALLDIR__/bin.
BackupPC supports archiving to removable media. For users that require offsite backups, BackupPC can create archives that stream to tape devices, or create files of specified sizes to fit onto cd or dvd media.
Each archive type is specified by a BackupPC host with its XferMethod set to 'archive'. This allows for multiple configurations at sites where there might be a combination of tape and cd/dvd backups being made.
BackupPC provides a menu that allows one or more hosts to be archived. The most recent backup of each host is archived using BackupPC_tarCreate, and the output is optionally compressed and split into fixed-sized files (eg: 650MB).
The archive for each host is done by default using __INSTALLDIR__/BackupPC_archiveHost. This script can be copied and customized as needed.
To create an Archive Host, add it to the hosts file just as any other host and call it a name that best describes the type of archive, e.g. ArchiveDLT
To tell BackupPC that the Host is for Archives, create a config.pl file in the Archive Hosts's pc directory, adding the following line:
$Conf{XferMethod} = 'archive';
To further customise the archive's parameters you can adding the changed parameters in the host's config.pl file. The parameters are explained in the config.pl file. Parameters may be fixed or the user can be allowed to change them (eg: output device).
The per-host archive command is $Conf{ArchiveClientCmd}. By default this invokes
__INSTALLDIR__/BackupPC_archiveHost
which you can copy and customize as necessary.
In the web interface, click on the Archive Host you wish to use. You will see a list of previous archives and a summary on each. By clicking the ``Start Archive'' button you are presented with the list of hosts and the approximate backup size (note this is raw size, not projected compressed size) Select the hosts you wish to archive and press the ``Archive Selected Hosts'' button.
The next screen allows you to adjust the parameters for this archive run. Press the ``Start the Archive'' to start archiving the selected hosts with the parameters displayed.
Identical files on multiples backups are represented by hard links. Hardlinks are used so that identical files all refer to the same physical file on the server's disk. Also, hard links maintain reference counts so that BackupPC knows when to delete unused files from the pool.
For the computer-science majors among you, you can think of the pooling system used by BackupPC as just a chained hash table stored on a (big) file system.
Using the file length and just the first 4096 bytes of the file for the MD5 digest produces some repetitions. One example: with 900,000 unique files in the pool, this hash gives about 7,000 repeated files, and in the worst case 500 files have the same hash. That's not bad: we only have to do a single file compare 99.2% of the time. But in the worst case we have to compare as many as 500 files checking for a match.
With a modest increase in CPU time, if we use the file length and the first 256K of the file we now only have 500 repeated files and in the worst case around 20 files have the same hash. Furthermore, if we instead use the first and last 128K of the file (more specifically, the first and eighth 128K chunks for files larger than 1MB) we get only 300 repeated files and in the worst case around 20 files have the same hash.
Based on this experimentation, this is the hash function used by BackupPC. It is important that you don't change the hash function after files are already in the pool. Otherwise your pool will grow to twice the size until all the old backups (and all the old files with old hashes) eventually expire.
The $Conf{CompressLevel} setting specifies the compression level to use. Zero (0) means no compression. Compression levels can be from 1 (least cpu time, slightly worse compression) to 9 (most cpu time, slightly better compression). The recommended value is 3. Changing it to 5, for example, will take maybe 20% more cpu time and will get another 2-3% additional compression. Diminishing returns set in above 5. See the zlib documentation for more information about compression levels.
BackupPC implements compression with minimal CPU load. Rather than compressing every incoming backup file and then trying to match it against the pool, BackupPC computes the MD5 digest based on the uncompressed file, and matches against the candidate pool files by comparing each uncompressed pool file against the incoming backup file. Since inflating a file takes roughly a factor of 10 less CPU time than deflating there is a big saving in CPU time.
The combination of pooling common files and compression can yield a factor of 8 or more overall saving in backup storage.
BackupPC reads the configuration information from __TOPDIR__/conf/config.pl. It then runs and manages all the backup activity. It maintains queues of pending backup requests, user backup requests and administrative commands. Based on the configuration various requests will be executed simultaneously.
As specified by $Conf{WakeupSchedule}, BackupPC wakes up periodically to queue backups on all the PCs. This is a four step process:
The backup is done using the specified XferMethod. Either samba's smbclient or tar over ssh/rsh/nfs piped into BackupPC_tarExtract, or rsync over ssh/rsh is run, or rsyncd is connected to, with the incoming data extracted to __TOPDIR__/pc/$host/new. The XferMethod output is put into __TOPDIR__/pc/$host/XferLOG.
The letter in the XferLOG file shows the type of object, similar to the first letter of the modes displayed by ls -l:
d -> directory l -> symbolic link b -> block special file c -> character special file p -> pipe file (fifo) nothing -> regular file
The words mean:
As BackupPC_tarExtract extracts the files from smbclient or tar, or as rsync runs, it checks each file in the backup to see if it is identical to an existing file from any previous backup of any PC. It does this without needed to write the file to disk. If the file matches an existing file, a hardlink is created to the existing file in the pool. If the file does not match any existing files, the file is written to disk and the file name is saved in __TOPDIR__/pc/$host/NewFileList for later processing by BackupPC_link. BackupPC_tarExtract and rsync can handle arbitrarily large files and multiple candidate matching files without needing to write the file to disk in the case of a match. This significantly reduces disk writes (and also reads, since the pool file comparison is done disk to memory, rather than disk to disk).
Based on the configuration settings, BackupPC_dump checks each old backup to see if any should be removed. Any expired backups are moved to __TOPDIR__/trash for later removal by BackupPC_trashClean.
BackupPC_link reads the NewFileList written by BackupPC_dump and inspects each new file in the backup. It re-checks if there is a matching file in the pool (another BackupPC_link could have added the file since BackupPC_dump checked). If so, the file is removed and replaced by a hard link to the existing file. If the file is new, a hard link to the file is made in the pool area, so that this file is available for checking against each new file and new backup.
Then, if $Conf{IncrFill} is set (note that the default setting is off), for each incremental backup, hard links are made in the new backup to all files that were not extracted during the incremental backups. The means the incremental backup looks like a complete image of the PC (with the exception that files that were removed on the PC since the last full backup will still appear in the backup directory tree).
The CGI interface knows how to merge unfilled incremental backups will the most recent prior filled (full) backup, giving the incremental backups a filled appearance. The default for $Conf{IncrFill} is off, since there is no need to fill incremental backups. This saves some level of disk activity, since lots of extra hardlinks are no longer needed (and don't have to be deleted when the backup expires).
Also, once each night, BackupPC_nightly is run to complete some additional administrative tasks, such as cleaning the pool. This involves removing any files in the pool that only have a single hard link (meaning no backups are using that file). Again, to avoid race conditions, BackupPC_nightly is only run when there are no BackupPC_dump or BackupPC_link processes running. Therefore, when it is time to run BackupPC_nightly, no new backups are started and BackupPC waits until all backups have finished. Then BackupPC_nightly is run, and until it finishes no new backups are started. If BackupPC_nightly is slow, the settings
BackupPC also listens for TCP connections on $Conf{ServerPort}, which is used by the CGI script BackupPC_Admin for status reporting and user-initiated backup or backup cancel requests.
BackupPC resides in three directories:
Below __TOPDIR__ are several directories:
Each file is stored in a subdirectory X/Y/Z, where X, Y, Z are the first 3 hex digits of the MD5 digest.
For example, if a file has an MD5 digest of 123456789abcdef0, the file is stored in __TOPDIR__/pool/1/2/3/123456789abcdef0.
The MD5 digest might not be unique (especially since not all the file's contents are used for files bigger than 256K). Different files that have the same MD5 digest are stored with a trailing suffix ``_n'' where n is an incrementing number starting at 0. So, for example, if two additional files were identical to the first, except the last byte was different, and assuming the file was larger than 1MB (so the MD5 digests are the same but the files are actually different), the three files would be stored as:
__TOPDIR__/pool/1/2/3/123456789abcdef0 __TOPDIR__/pool/1/2/3/123456789abcdef0_0 __TOPDIR__/pool/1/2/3/123456789abcdef0_1
Both BackupPC_dump (actually, BackupPC_tarExtract) and BackupPC_link are responsible for checking newly backed up files against the pool. For each file, the MD5 digest is used to generate a file name in the pool directory. If the file exists in the pool, the contents are compared. If there is no match, additional files ending in ``_n'' are checked. (Actually, BackupPC_tarExtract compares multiple candidate files in parallel.) If the file contents exactly match, the file is created by simply making a hard link to the pool file (this is done by BackupPC_tarExtract as the backup proceeds). Otherwise, BackupPC_tarExtract writes the new file to disk and a new hard link is made in the pool to the file (this is done later by BackupPC_link).
Therefore, every file in the pool will have at least 2 hard links (one for the pool file and one for the backup file below __TOPDIR__/pc). Identical files from different backups or PCs will all be linked to the same file. When old backups are deleted, some files in the pool might only have one link. BackupPC_nightly checks the entire pool and removes all files that have only a single link, thereby recovering the storage for that file.
One other issue: zero length files are not pooled, since there are a lot of these files and on most file systems it doesn't save any disk space to turn these files into hard links.
The compressed file format is as generated by Compress::Zlib::deflate with one minor, but important, tweak. Since Compress::Zlib::inflate fully inflates its argument in memory, it could take large amounts of memory if it was inflating a highly compressed file. For example, a 200MB file of 0x0 bytes compresses to around 200K bytes. If Compress::Zlib::inflate was called with this single 200K buffer, it would need to allocate 200MB of memory to return the result.
BackupPC watches how efficiently a file is compressing. If a big file
has very high compression (meaning it will use too much memory when it
is inflated), BackupPC calls the flush()
method, which gracefully
completes the current compression. BackupPC then starts another
deflate and simply appends the output file. So the BackupPC compressed
file format is one or more concatenated deflations/flushes. The specific
ratios that BackupPC uses is that if a 6MB chunk compresses to less
than 64K then a flush will be done.
Back to the example of the 200MB file of 0x0 bytes. Adding flushes every 6MB adds only 200 or so bytes to the 200K output. So the storage cost of flushing is negligible.
To easily decompress a BackupPC compressed file, the script BackupPC_zcat can be found in __INSTALLDIR__/bin. For each file name argument it inflates the file and writes it to stdout.
An incremental backup with rsync compares attributes on the client with the last full backup. Any files with identical attributes are skipped. A full backup with rsync sets the --ignore-times option, which causes every file to be examined independent of attributes.
Each file is examined by generating block checksums (default 2K blocks) on the receiving side (that's the BackupPC side), sending those checksums to the client, where the remote rsync matches those checksums with the corresponding file. The matching blocks and new data is sent back, allowing the client file to be reassembled. A checksum for the entire file is sent to as an extra check the the reconstructed file is correct.
This results in significant disk IO and computation for BackupPC: every file in a full backup, or any file with non-matching attributes in an incremental backup, needs to be uncompressed, block checksums computed and sent. Then the receiving side reassembles the file and has to verify the whole-file checksum. Even if the file is identical, prior to 2.1.0, BackupPC had to read and uncompress the file twice, once to compute the block checksums and later to verify the whole-file checksum.
Starting in 2.1.0, BackupPC supports optional checksum caching, which means the block and file checksums only need to be computed once for each file. This results in a significant performance improvement. This only works for compressed pool files. It is enabled by adding
'--checksum-seed=32761',
to $Conf{RsyncArgs} and $Conf{RsyncRestoreArgs}.
Rsync versions prior to and including rsync-2.6.2 need a small patch to add support for the --checksum-seed option. This patch is available in the cygwin-rsyncd package at http://backuppc.sourceforge.net. This patch is already included in rsync CVS, so it will be standard in future versions of rsync.
When this option is present, BackupPC will add block and file checksums to the compressed pool file the next time a pool file is used and it doesn't already have cached checksums. The first time a new file is written to the pool, the checksums are not appended. The next time checksums are needed for a file, they are computed and added. So the full performance benefit of checksum caching won't be noticed until the third time a pool file is used (eg: the third full backup).
With checksum caching enabled, there is a risk that should a file's contents in the pool be corrupted due to a disk problem, but the cached checksums are still correct, the corruption will not be detected by a full backup, since the file contents are no longer read and compared. To reduce the chance that this remains undetected, BackupPC can recheck cached checksums for a fraction of the files. This fraction is set with the $Conf{RsyncCsumCacheVerifyProb} setting. The default value of 0.01 means that 1% of the time a file's checksums are read, the checksums are verified. This reduces performance slightly, but, over time, ensures that files contents are in sync with the cached checksums.
The format of the cached checksum data can be discovered by looking at the code. Basically, the first byte of the compressed file is changed to denote that checksums are appended. The block and file checksum data, plus some other information and magic word, are appended to the compressed file. This allows the cache update to be done in-place.
Backup file names are stored in ``mangled'' form. Each node of a path is preceded by ``f'' (mnemonic: file), and special characters (\n, \r, % and /) are URI-encoded as ``%xx'', where xx is the ascii character's hex value. So c:/craig/example.txt is now stored as fc/fcraig/fexample.txt.
This was done mainly so meta-data could be stored alongside the backup files without name collisions. In particular, the attributes for the files in a directory are stored in a file called ``attrib'', and mangling avoids file name collisions (I discarded the idea of having a duplicate directory tree for every backup just to store the attributes). Other meta-data (eg: rsync checksums) could be stored in file names preceded by, eg, ``c''. There are two other benefits to mangling: the share name might contain ``/'' (eg: ``/home/craig'' for tar transport), and I wanted that represented as a single level in the storage tree. Secondly, as files are written to NewFileList for later processing by BackupPC_link, embedded newlines in the file's path will cause problems which are avoided by mangling.
The CGI script undoes the mangling, so it is invisible to the user. Old (unmangled) backups are still supported by the CGI interface.
Linux/unix file systems support several special file types: symbolic links, character and block device files, fifos (pipes) and unix-domain sockets. All except unix-domain sockets are supported by BackupPC (there's no point in backing up or restoring unix-domain sockets since they only have meaning after a process creates them). Symbolic links are stored as a plain file whose contents are the contents of the link (not the file it points to). This file is compressed and pooled like any normal file. Character and block device files are also stored as plain files, whose contents are two integers separated by a comma; the numbers are the major and minor device number. These files are compressed and pooled like any normal file. Fifo files are stored as empty plain files (which are not pooled since they have zero size). In all cases, the original file type is stored in the attrib file so it can be correctly restored.
Hardlinks are also supported. When GNU tar first encounters a file with more than one link (ie: hardlinks) it dumps it as a regular file. When it sees the second and subsequent hardlinks to the same file, it dumps just the hardlink information. BackupPC correctly recognizes these hardlinks and stores them just like symlinks: a regular text file whose contents is the path of the file linked to. The CGI script will download the original file when you click on a hardlink.
Also, BackupPC_tarCreate has enough magic to re-create the hardlinks dynamically based on whether or not the original file and hardlinks are both included in the tar file. For example, imagine a/b/x is a hardlink to a/c/y. If you use BackupPC_tarCreate to restore directory a, then the tar file will include a/b/x as the original file and a/c/y will be a hardlink to a/b/x. If, instead you restore a/c, then the tar file will include a/c/y as the original file, not a hardlink.
The unix attributes for the contents of a directory (all the files and directories in that directory) are stored in a file called attrib. There is a single attrib file for each directory in a backup. For example, if c:/craig contains a single file c:/craig/example.txt, that file would be stored as fc/fcraig/fexample.txt and there would be an attribute file in fc/fcraig/attrib (and also fc/attrib and ./attrib). The file fc/fcraig/attrib would contain a single entry containing the attributes for fc/fcraig/fexample.txt.
The attrib file starts with a magic number, followed by the concatenation of the following information for each file:
The attrib file is also compressed if compression is enabled. See the lib/BackupPC/Attrib.pm module for full details.
Attribute files are pooled just like normal backup files. This saves space if all the files in a directory have the same attributes across multiple backups, which is common.
BackupPC doesn't care about the access time of files in the pool since it saves attribute meta-data separate from the files. Since BackupPC mostly does reads from disk, maintaining the access time of files generates a lot of unnecessary disk writes. So, provided BackupPC has a dedicated data disk, you should consider mounting BackupPC's data directory with the noatime attribute (see mount(1)).
BackupPC isn't perfect (but it is getting better). Please see http://backuppc.sourceforge.net/faq/limitations.html for a discussion of some of BackupPC's limitations.
Please see http://backuppc.sourceforge.net/faq/security.html for a discussion of some of various security issues.
The BackupPC configuration file resides in __TOPDIR__/conf/config.pl. Optional per-PC configuration files reside in __TOPDIR__/pc/$host/config.pl. This file can be used to override settings just for a particular PC.
The configuration file is a perl script that is executed by BackupPC, so you should be careful to preserve the file syntax (punctuation, quotes etc) when you edit it. It is recommended that you use CVS, RCS or some other method of source control for changing config.pl.
BackupPC reads or re-reads the main configuration file and the hosts file in three cases:
Whenever you change the configuration file you can either do a kill -HUP BackupPC_pid or simply wait until the next regular wakeup period.
Each time the configuration file is re-read a message is reported in the LOG file, so you can tail it (or view it via the CGI interface) to make sure your kill -HUP worked. Errors in parsing the configuration file are also reported in the LOG file.
The optional per-PC configuration file (__TOPDIR__/pc/$host/config.pl) is read whenever it is needed by BackupPC_dump, BackupPC_link and others.
If you have a heterogeneous set of clients (eg: a variety of WinXX and linux/unix machines) you will need to create host-specific config.pl files for some or all of these machines to customize the default settings from the master config.pl file (at a minimum to set $Conf{XferMethod}).
Since the config.pl file is just regular perl code, you can include one config file from another. For example, imagine you had three general classes of machines: WinXX desktops, linux machines in the DMZ and linux desktops. You could create three config files in __TOPDIR__/conf:
__TOPDIR__/conf/ConfigWinDesktop.pl __TOPDIR__/conf/ConfigLinuxDMZ.pl __TOPDIR__/conf/ConfigLinuxDesktop.pl
From each client's directory you can either add a symbolic link to the appropriate config file:
cd __TOPDIR__/pc/$host ln -s ../../conf/ConfigWinDesktop.pl config.pl
or, better yet, create a config.pl file in __TOPDIR__/pc/$host that includes the default config.pl file using perl's ``do'' command:
do "__TOPDIR__/conf/ConfigWinDesktop.pl";
This alternative allows you to set other configuration options specific to each host after the ``do'' command (perhaps even overriding the settings in the included file).
Note that you could also include snippets of configuration settings from the main configuration file. However, be aware that the modification-time checking that BackupPC does only applies to the main configuration file: if you change one of the included files, BackupPC won't notice. You will need to either touch the main configuration file too, or send BackupPC a HUP (-1) signal.
The configuration parameters are divided into five general groups. The first group (general server configuration) provides general configuration for BackupPC. The next two groups describe what to backup, when to do it, and how long to keep it. The fourth group are settings for email reminders, and the final group contains settings for the CGI interface.
All configuration settings in the second through fifth groups can be overridden by the per-PC config.pl file.
To avoid possible attacks via the TCP socket interface, every client message is protected by an MD5 digest. The MD5 digest includes four items: - a seed that is sent to the client when the connection opens - a sequence number that increments for each message - a shared secret that is stored in $Conf{ServerMesgSecret} - the message itself.
The message is sent in plain text preceded by the MD5 digest. A snooper can see the plain-text seed sent by BackupPC and plain-text message from the client, but cannot construct a valid MD5 digest since the secret $Conf{ServerMesgSecret} is unknown. A replay attack is not possible since the seed changes on a per-connection and per-message basis.
If the hosts you are backing up are always connected to the network you might have only one or two wakeups each night. This will keep the backup activity after hours. On the other hand, if you are backing up laptops that are only intermittently connected to the network you will want to have frequent wakeups (eg: hourly) to maximized the chance that each laptop is backed up.
Examples:
$Conf{WakeupSchedule} = [22.5]; # once per day at 10:30 pm. $Conf{WakeupSchedule} = [1..23]; # every hour except midnight $Conf{WakeupSchedule} = [2,4,6,8,10,12,14,16,18,20,22]; # every 2 hours
The default value is every hour except midnight.
The first entry of $Conf{WakeupSchedule} is when BackupPC_nightly is run. No other backups can run while BackupPC_nightly is running. You might want to re-arrange the entries in $Conf{WakeupSchedule} (they don't have to be ascending) so that the first entry is when you want BackupPC_nightly to run (eg: when you don't expect a lot of regular backups to run).
Each night, at the first wakeup listed in $Conf{WakeupSchedule}, BackupPC_nightly is run. Its job is to remove unneeded files in the pool, ie: files that only have one link. To avoid race conditions, BackupPC_nightly runs only when there are no backups running, and no backups will start while it runs.
So to reduce the elapsed time, you might want to increase this setting to run several BackupPC_nightly processes in parallel (eg: 4, or even 8).
Other valid values are 2, 4, 8, 16. This causes BackupPC_nightly to traverse 1/2, 1/4, 1/8 or 1/16th of the pool each night, meaning it takes 2, 4, 8 or 16 days to completely traverse the pool. The advantage is that each night the running time of BackupPC_nightly is reduced roughly in proportion, since the total job is split over multiple days. The disadvantage is that unused pool files take longer to get deleted, which will slightly increase disk usage.
Note that even when $Conf{BackupPCNightlyPeriod} > 1, BackupPC_nightly still runs every night. It just does less work each time it runs.
Examples:
$Conf{BackupPCNightlyPeriod} = 1; # entire pool is checked every night
$Conf{BackupPCNightlyPeriod} = 2; # two days to complete pool check # (different half each night)
$Conf{BackupPCNightlyPeriod} = 4; # four days to complete pool check # (different quarter each night)
If you decrease this number after BackupPC has been running for a while you will have to manually remove the older log files.
$dfPath path to df ($Conf{DfPath}) $topDir top-level BackupPC data directory
Examples:
# to specify 192.10.10.20 to 192.10.10.250 as the DHCP address pool $Conf{DHCPAddressRanges} = [ { ipAddrBase => '192.10.10', first => 20, last => 250, }, ]; # to specify two pools (192.10.10.20-250 and 192.10.11.10-50) $Conf{DHCPAddressRanges} = [ { ipAddrBase => '192.10.10', first => 20, last => 250, }, { ipAddrBase => '192.10.11', first => 10, last => 50, }, ];
$Conf{SmbShareName} = 'c'; # backup 'c' share $Conf{SmbShareName} = ['c', 'd']; # backup 'c' and 'd' shares
This setting only matters if $Conf{XferMethod} = 'smb'.
This setting only matters if $Conf{XferMethod} = 'smb'.
This setting only matters if $Conf{XferMethod} = 'smb'.
$Conf{TarShareName} = '/'; # backup everything $Conf{TarShareName} = '/home'; # only backup /home $Conf{TarShareName} = ['/home', '/src']; # backup /home and /src
The fact this parameter is called 'TarShareName' is for historical consistency with the Smb transport options. You can use any valid directory on the client: there is no need for it to correspond to any Smb share or device mount point.
Note also that you can also use $Conf{BackupFilesOnly} to specify a specific list of directories to backup. It's more efficient to use this option instead of $Conf{TarShareName} since a new tar is run for each entry in $Conf{TarShareName}.
On the other hand, if you add --one-file-system to $Conf{TarClientCmd} you can backup each file system separately, which makes restoring one bad file system easier. In this case you would list all of the mount points here, since you can't get the same result with $Conf{BackupFilesOnly}:
$Conf{TarShareName} = ['/', '/var', '/data', '/boot'];
This setting only matters if $Conf{XferMethod} = 'tar'.
Typically this is set slightly less than an integer number of days. The time taken for the backup, plus the granularity of $Conf{WakeupSchedule} will make the actual backup interval a bit longer.
There are two special values for $Conf{FullPeriod}:
-1 Don't do any regular backups on this machine. Manually requested backups (via the CGI interface) will still occur.
-2 Don't do any backups on this machine. Manually requested backups (via the CGI interface) will be ignored.
These special settings are useful for a client that is no longer being backed up (eg: a retired machine), but you wish to keep the last backups available for browsing or restoring to other machines.
Typically this is set slightly less than an integer number of days. The time taken for the backup, plus the granularity of $Conf{WakeupSchedule} will make the actual backup interval a bit longer.
In the steady state, each time a full backup completes successfully the oldest one is removed. If this number is decreased, the extra old backups will be removed.
If filling of incremental dumps is off the oldest backup always has to be a full (ie: filled) dump. This might mean one or two extra full dumps are kept until the oldest incremental backups expire.
Exponential backup expiry is also supported. This allows you to specify:
- num fulls to keep at intervals of 1 * $Conf{FullPeriod}, followed by - num fulls to keep at intervals of 2 * $Conf{FullPeriod}, - num fulls to keep at intervals of 4 * $Conf{FullPeriod}, - num fulls to keep at intervals of 8 * $Conf{FullPeriod}, - num fulls to keep at intervals of 16 * $Conf{FullPeriod},
and so on. This works by deleting every other full as each expiry boundary is crossed.
Exponential expiry is specified using an array for $Conf{FullKeepCnt}:
$Conf{FullKeepCnt} = [4, 2, 3];
Entry #n specifies how many fulls to keep at an interval of 2^n * $Conf{FullPeriod} (ie: 1, 2, 4, 8, 16, 32, ...).
The example above specifies keeping 4 of the most recent full backups (1 week interval) two full backups at 2 week intervals, and 3 full backups at 4 week intervals, eg:
full 0 19 weeks old \ full 1 15 weeks old >--- 3 backups at 4 * $Conf{FullPeriod} full 2 11 weeks old / full 3 7 weeks old \____ 2 backups at 2 * $Conf{FullPeriod} full 4 5 weeks old / full 5 3 weeks old \ full 6 2 weeks old \___ 4 backups at 1 * $Conf{FullPeriod} full 7 1 week old / full 8 current /
On a given week the spacing might be less than shown as each backup ages through each expiry period. For example, one week later, a new full is completed and the oldest is deleted, giving:
full 0 16 weeks old \ full 1 12 weeks old >--- 3 backups at 4 * $Conf{FullPeriod} full 2 8 weeks old / full 3 6 weeks old \____ 2 backups at 2 * $Conf{FullPeriod} full 4 4 weeks old / full 5 3 weeks old \ full 6 2 weeks old \___ 4 backups at 1 * $Conf{FullPeriod} full 7 1 week old / full 8 current /
You can specify 0 as a count (except in the first entry), and the array can be as long as you wish. For example:
$Conf{FullKeepCnt} = [4, 0, 4, 0, 0, 2];
This will keep 10 full dumps, 4 most recent at 1 * $Conf{FullPeriod}, followed by 4 at an interval of 4 * $Conf{FullPeriod} (approx 1 month apart), and then 2 at an interval of 32 * $Conf{FullPeriod} (approx 7-8 months apart).
Example: these two settings are equivalent and both keep just the four most recent full dumps:
$Conf{FullKeepCnt} = 4; $Conf{FullKeepCnt} = [4];
Note that $Conf{FullAgeMax} will be increased to $Conf{FullAgeMax} times $Conf{FullPeriod} if $Conf{FullAgeMax} specifies enough full backups to exceed $Conf{FullAgeMax}.
In the steady state, each time an incr backup completes successfully the oldest one is removed. If this number is decreased, the extra old backups will be removed.
The default setting of 3 days means that a partial older than 3 days is ignored when the next full backup is done.
BackupPC, and the cgi interface in particular, do the right thing on un-filled incremental backups. It will correctly display the merged incremental backup with the most recent filled backup, giving the un-filled incremental backups a filled appearance. That means it invisible to the user whether incremental dumps are filled or not.
Filling backups takes a little extra disk space, and it does cost some extra disk activity for filling, and later removal. Filling is no longer useful, since file mangling and compression doesn't make a filled backup very useful. It's likely the filling option will be removed from future versions: filling will be delegated to the display and extraction of backup data.
If filling is off, BackupPC makes sure that the oldest backup is a full, otherwise the following incremental backups will be incomplete. This might mean an extra full backup has to be kept until the following incremental backups expire.
The default is off. You can turn this on or off at any time without affecting existing backups.
Note: files/dirs delivered via Zip or Tar downloads don't count as restores. Only the first restore option (where the files and dirs are written to the host) count as restores that are logged.
For Smb, only one of $Conf{BackupFilesExclude} and $Conf{BackupFilesOnly} can be specified per share. If both are set for a particular share, then $Conf{BackupFilesOnly} takes precedence and $Conf{BackupFilesExclude} is ignored.
This can be set to a string, an array of strings, or, in the case of multiple shares, a hash of strings or arrays. A hash is used to give a list of directories or files to backup for each share (the share name is the key). If this is set to just a string or array, and $Conf{SmbShareName} contains multiple share names, then the setting is assumed to apply all shares.
Examples:
$Conf{BackupFilesOnly} = '/myFiles'; $Conf{BackupFilesOnly} = ['/myFiles']; # same as first example $Conf{BackupFilesOnly} = ['/myFiles', '/important']; $Conf{BackupFilesOnly} = { 'c' => ['/myFiles', '/important'], # these are for 'c' share 'd' => ['/moreFiles', '/archive'], # these are for 'd' share };
This can be set to a string, an array of strings, or, in the case of multiple shares, a hash of strings or arrays. A hash is used to give a list of directories or files to exclude for each share (the share name is the key). If this is set to just a string or array, and $Conf{SmbShareName} contains multiple share names, then the setting is assumed to apply to all shares.
The exact behavior is determined by the underlying transport program, smbclient or tar. For smbclient the exlclude file list is passed into the X option. Simple shell wild-cards using ``*'' or ``?'' are allowed.
For tar, if the exclude file contains a ``/'' it is assumed to be anchored at the start of the string. Since all the tar paths start with ``./'', BackupPC prepends a ``.'' if the exclude file starts with a ``/''. Note that GNU tar version >= 1.13.7 is required for the exclude option to work correctly. For linux or unix machines you should add ``/proc'' to $Conf{BackupFilesExclude} unless you have specified --one-file-system in $Conf{TarClientCmd} or --one-file-system in $Conf{RsyncArgs}. Also, for tar, do not use a trailing ``/'' in the directory name: a trailing ``/'' causes the name to not match and the directory will not be excluded.
Users report that for smbclient you should specify a directory followed by ``/*'', eg: ``/proc/*'', instead of just ``/proc''.
Examples:
$Conf{BackupFilesExclude} = '/temp'; $Conf{BackupFilesExclude} = ['/temp']; # same as first example $Conf{BackupFilesExclude} = ['/temp', '/winnt/tmp']; $Conf{BackupFilesExclude} = { 'c' => ['/temp', '/winnt/tmp'], # these are for 'c' share 'd' => ['/junk', '/dont_back_this_up'], # these are for 'd' share };
To allow for periodic rebooting of a PC or other brief periods when a PC is not on the network, a number of consecutive bad pings is allowed before the good ping count is reset. This parameter is $Conf{BlackoutBadPingLimit}.
Note that bad and good pings don't occur with the same interval. If a machine is always on the network, it will only be pinged roughly once every $Conf{IncrPeriod} (eg: once per day). So a setting for $Conf{BlackoutGoodCnt} of 7 means it will take around 7 days for a machine to be subject to blackout. On the other hand, if a ping is failed, it will be retried roughly every time BackupPC wakes up, eg, every one or two hours. So a setting for $Conf{BlackoutBadPingLimit} of 3 means that the PC will lose its blackout status after 3-6 hours of unavailability.
To disable the blackout feature set $Conf{BlackoutGoodCnt} to a negative value. A value of 0 will make all machines subject to blackout. But if you don't want to do any backups during the day it would be easier to just set $Conf{WakeupSchedule} to a restricted schedule.
For example:
$Conf{BlackoutPeriods} = [ { hourBegin => 7.0, hourEnd => 19.5, weekDays => [1, 2, 3, 4, 5], }, ];
specifies one blackout period from 7:00am to 7:30pm local time on Mon-Fri.
The blackout period can also span midnight by setting hourBegin > hourEnd, eg:
$Conf{BlackoutPeriods} = [ { hourBegin => 7.0, hourEnd => 19.5, weekDays => [1, 2, 3, 4, 5], }, { hourBegin => 23, hourEnd => 5, weekDays => [5, 6], }, ];
This specifies one blackout period from 7:00am to 7:30pm local time on Mon-Fri, and a second period from 11pm to 5am on Friday and Saturday night.
The valid values are:
- 'smb': backup and restore via smbclient and the SMB protocol. Easiest choice for WinXX.
- 'rsync': backup and restore via rsync (via rsh or ssh). Best choice for linux/unix. Good choice also for WinXX.
- 'rsyncd': backup and restre via rsync daemon on the client. Best choice for linux/unix if you have rsyncd running on the client. Good choice also for WinXX.
- 'tar': backup and restore via tar, tar over ssh, rsh or nfs. Good choice for linux/unix.
- 'archive': host is a special archive host. Backups are not done. An archive host is used to archive other host's backups to permanent media, such as tape, CDR or DVD.
smbclient is from the Samba distribution. smbclient is used to actually extract the incremental or full dump of the share filesystem from the PC.
This setting only matters if $Conf{XferMethod} = 'smb'.
The following variables are substituted at run-time:
$smbClientPath same as $Conf{SmbClientPath} $host host to backup/restore $hostIP host IP address $shareName share name $userName user name $fileList list of files to backup (based on exclude/include) $I_option optional -I option to smbclient $X_option exclude option (if $fileList is an exclude list) $timeStampFile start time for incremental dump
Same variable substitutions are applied as $Conf{SmbClientFullCmd}.
Same variable substitutions are applied as $Conf{SmbClientFullCmd}.
If your smb share is read-only then direct restores will fail. You should set $Conf{SmbClientRestoreCmd} to undef and the corresponding CGI restore option will be removed.
See the documentation for more information about setting up ssh2 keys.
If you plan to use NFS then tar just runs locally and ssh2 is not needed. For example, assuming the client filesystem is mounted below /mnt/hostName, you could use something like:
$Conf{TarClientCmd} = '$tarPath -c -v -f - -C /mnt/$host/$shareName' . ' --totals';
In the case of NFS or rsh you need to make sure BackupPC's privileges are sufficient to read all the files you want to backup. Also, you will probably want to add ``/proc'' to $Conf{BackupFilesExclude}.
The following variables are substituted at run-time:
$host host name $hostIP host's IP address $incrDate newer-than date for incremental backups $shareName share name to backup (ie: top-level directory path) $fileList specific files to backup or exclude $tarPath same as $Conf{TarClientPath} $sshPath same as $Conf{SshPath}
If a variable is followed by a ``+'' it is shell escaped. This is necessary for the command part of ssh or rsh, since it ends up getting passed through the shell.
This setting only matters if $Conf{XferMethod} = 'tar'.
If you are running tar locally (ie: without rsh or ssh) then remove the ``+'' so that the argument is no longer shell escaped.
This setting only matters if $Conf{XferMethod} = 'tar'.
Note that GNU tar has several methods for specifying incremental backups, including:
--newer-mtime $incrDate+ This causes a file to be included if the modification time is later than $incrDate (meaning its contents might have changed). But changes in the ownership or modes will not qualify the file to be included in an incremental.
--newer=$incrDate+ This causes the file to be included if any attribute of the file is later than $incrDate, meaning either attributes or the modification time. This is the default method. Do not use --atime-preserve in $Conf{TarClientCmd} above, otherwise resetting the atime (access time) counts as an attribute change, meaning the file will always be included in each new incremental dump.
If you are running tar locally (ie: without rsh or ssh) then remove the ``+'' so that the argument is no longer shell escaped.
This setting only matters if $Conf{XferMethod} = 'tar'.
See $Conf{TarClientCmd} for full details.
This setting only matters if $Conf{XferMethod} = ``tar''.
If you want to disable direct restores using tar, you should set $Conf{TarClientRestoreCmd} to undef and the corresponding CGI restore option will be removed.
This setting only matters if $Conf{XferMethod} = 'tar'.
$host host name being backed up $hostIP host's IP address $shareName share name to backup (ie: top-level directory path) $rsyncPath same as $Conf{RsyncClientPath} $sshPath same as $Conf{SshPath} $argList argument list, built from $Conf{RsyncArgs}, $shareName, $Conf{BackupFilesExclude} and $Conf{BackupFilesOnly}
This setting only matters if $Conf{XferMethod} = 'rsync'.
$host host name being backed up $hostIP host's IP address $shareName share name to backup (ie: top-level directory path) $rsyncPath same as $Conf{RsyncClientPath} $sshPath same as $Conf{SshPath} $argList argument list, built from $Conf{RsyncArgs}, $shareName, $Conf{BackupFilesExclude} and $Conf{BackupFilesOnly}
This setting only matters if $Conf{XferMethod} = 'rsync'.
For $Conf{XferMethod} = ``rsyncd'' this should be the name of the module to backup (ie: the name from /etc/rsynd.conf).
This can also be a list of multiple file system paths or modules. For example, by adding --one-file-system to $Conf{RsyncArgs} you can backup each file system separately, which makes restoring one bad file system easier. In this case you would list all of the mount points:
$Conf{RsyncShareName} = ['/', '/var', '/data', '/boot'];
This setting is the probability (0 means never and 1 means always) that a file will be rechecked. Setting it to 0 means the checksums will not be rechecked (unless there is a phase 0 failure). Setting it to 1 (ie: 100%) means all files will be checked, but that is not a desirable setting since you are better off simply turning caching off (ie: remove the --checksum-seed option).
The default of 0.01 means 1% (on average) of the files during a full backup will have their cached checksum re-checked.
This setting has no effect unless checksum caching is turned on.
Examples of additional arguments that should work are --exclude/--include, eg:
$Conf{RsyncArgs} = [ # original arguments here '-v', '--exclude', '/proc', '--exclude', '*.tmp', ];
If you want to disable direct restores using rsync (eg: is the module is read-only), you should set $Conf{RsyncRestoreArgs} to undef and the corresponding CGI restore option will be removed.
The Destination of the archive e.g. /tmp for file archive or /dev/nst0 for device archive
The valid values are:
- 'none': No Compression
- 'gzip': Medium Compression. Recommended.
- 'bzip2': High Compression but takes longer.
The amount of Parity data to generate, as a percentage of the archive size. Uses the commandline par2 (par2cmdline) available from http://parchive.sourceforge.net
Only useful for file dumps.
Set to 0 to disable this feature.
Only for file archives. Splits the output into the specified size * 1,000,000. e.g. to split into 650,000,000 bytes, specify 650 below.
If the value is 0, or if $Conf{ArchiveDest} is an existing file or device (e.g. a streaming tape drive), this feature is disabled.
This is the command that is called to actually run the archive process for each host. The following variables are substituted at run-time:
$Installdir The installation directory of BackupPC $tarCreatePath The path to BackupPC_tarCreate $splitpath The path to the split program $parpath The path to the par2 program $host The host to archive $backupnumber The backup number of the host to archive $compression The path to the compression program $compext The extension assigned to the compression type $splitsize The number of bytes to split archives into $archiveloc The location to put the archive $parfile The amount of parity data to create (percentage)
nmblookup is from the Samba distribution. nmblookup is used to get the netbios name, necessary for DHCP hosts.
$nmbLookupPath path to nmblookup ($Conf{NmbLookupPath}) $host IP address
This command is only used for DHCP hosts: given an IP address, this command should try to find its NetBios name.
$nmbLookupPath path to nmblookup ($Conf{NmbLookupPath}) $host NetBios name
In some cases you might need to change the broadcast address, for example if nmblookup uses 192.168.255.255 by default and you find that doesn't work, try 192.168.1.255 (or your equivalent class C address) using the -B option:
$Conf{NmbLookupFindHostCmd} = '$nmbLookupPath -B 192.168.1.255 $host';
If you use a WINS server and your machines don't respond to multicast NetBios requests you can use this (replace 1.2.3.4 with the IP address of your WINS server):
$Conf{NmbLookupFindHostCmd} = '$nmbLookupPath -R -U 1.2.3.4 $host';
This is preferred over multicast since it minimizes network traffic.
Experiment manually for your site to see what form of nmblookup command works.
If you want to disable ping checking, set this to some program that exits with 0 status, eg:
$Conf{PingPath} = '/bin/echo';
$pingPath path to ping ($Conf{PingPath}) $host host name
Wade Brown reports that on solaris 2.6 and 2.7 ping -s returns the wrong exit status (0 even on failure). Replace with ``ping $host 1'', which gets the correct exit status but we don't get the round-trip time.
$sshPath path to ssh ($Conf{SshPath}) $serverHost same as $Conf{ServerHost} $serverInitdPath path to init.d script ($Conf{ServerInitdPath})
Example:
$Conf{ServerInitdPath} = '/etc/init.d/backuppc'; $Conf{ServerInitdStartCmd} = '$sshPath -q -x -l root $serverHost' . ' $serverInitdPath start' . ' < /dev/null >& /dev/null';
Changing compression on or off after backups have already been done will require both compressed and uncompressed pool files to be stored. This will increase the pool storage requirements, at least until all the old backups expire and are deleted.
It is ok to change the compression value (from one non-zero value to another non-zero value) after dumps are already done. Since BackupPC matches pool files by comparing the uncompressed versions, it will still correctly match new incoming files against existing pool files. The new compression level will take effect only for new files that are newly compressed and added to the pool.
If compression was off and you are enabling compression for the first time you can use the BackupPC_compressPool utility to compress the pool. This avoids having the pool grow to accommodate both compressed and uncompressed backups. See the documentation for more information.
Note: compression needs the Compress::Zlib perl library. If the Compress::Zlib library can't be found then $Conf{CompressLevel} is forced to 0 (compression off).
Note that stdout buffering combined with huge files being backed up could cause longish delays in the output from smbclient that BackupPC_dump sees, so in rare cases you might want to increase this value.
Despite the name, this parameter sets the timeout for all transport methods (tar, smb etc).
If you decrease this number after BackupPC has been running for a while you will have to manually remove the older log files.
$Conf{DumpPreUserCmd} = '$sshPath -q -x -l root $host /usr/bin/dumpMysql';
The following variable substitutions are made at run time for $Conf{DumpPreUserCmd} and $Conf{DumpPostUserCmd}:
$type type of dump (incr or full) $xferOK 1 if the dump succeeded, 0 if it didn't $client client name being backed up $host host name (could be different from client name if $Conf{ClientNameAlias} is set) $hostIP IP address of host $user user name from the hosts file $moreUsers list of additional users from the hosts file $share the first share name $shares list of all the share names $XferMethod value of $Conf{XferMethod} (eg: tar, rsync, smb) $sshPath value of $Conf{SshPath}, $cmdType set to DumpPreUserCmd or DumpPostUserCmd
The following variable substitutions are made at run time for $Conf{RestorePreUserCmd} and $Conf{RestorePostUserCmd}:
$client client name being backed up $xferOK 1 if the restore succeeded, 0 if it didn't $host host name (could be different from client name if $Conf{ClientNameAlias} is set) $hostIP IP address of host $user user name from the hosts file $moreUsers list of additional users from the hosts file $share the first share name $XferMethod value of $Conf{XferMethod} (eg: tar, rsync, smb) $sshPath value of $Conf{SshPath}, $type set to "restore" $bkupSrcHost host name of the restore source $bkupSrcShare share name of the restore source $bkupSrcNum backup number of the restore source $pathHdrSrc common starting path of restore source $pathHdrDest common starting path of destination $fileList list of files being restored $cmdType set to RestorePreUserCmd or RestorePostUserCmd
The following variable substitutions are made at run time for $Conf{ArchivePreUserCmd} and $Conf{ArchivePostUserCmd}:
$client client name being backed up $xferOK 1 if the archive succeeded, 0 if it didn't $host Name of the archive host $user user name from the hosts file $share the first share name $XferMethod value of $Conf{XferMethod} (eg: tar, rsync, smb) $HostList list of hosts being archived $BackupList list of backup numbers for the hosts being archived $archiveloc location where the archive is sent to $parfile amount of parity data being generated (percentage) $compression compression program being used (eg: cat, gzip, bzip2) $compext extension used for compression type (eg: raw, gz, bz2) $splitsize size of the files that the archive creates $sshPath value of $Conf{SshPath}, $type set to "archive" $cmdType set to ArchivePreUserCmd or ArchivePostUserCmd
$Conf{ClientNameAlias} = 'realHostName'; $Conf{ClientNameAlias} = '192.1.1.15';
will cause the relevant smb/tar/rsync backup/restore commands to be directed to realHostName, not the client name.
Note: this setting doesn't work for hosts with DHCP set to 1.
$Cong{EMailUserDestDomain} = '@mydomain.com';
With this setting user email will be set to 'user@mydomain.com'.
These values are language-dependent. The default versions can be found in the language file (eg: lib/BackupPC/Lang/en.pm). If you need to change the message, copy it here and edit it, eg:
$Conf{EMailNoBackupEverMesg} = <<'EOF'; To: $user$domain cc: Subject: $subj
Dear $userName,
This is a site-specific email message. EOF
These values are language-dependent. The default versions can be found in the language file (eg: lib/BackupPC/Lang/en.pm). If you need to change the message, copy it here and edit it, eg:
$Conf{EMailNoBackupRecentMesg} = <<'EOF'; To: $user$domain cc: Subject: $subj
Dear $userName,
This is a site-specific email message. EOF
These values are language-dependent. The default versions can be found in the language file (eg: lib/BackupPC/Lang/en.pm). If you need to change the message, copy it here and edit it, eg:
$Conf{EMailOutlookBackupMesg} = <<'EOF'; To: $user$domain cc: Subject: $subj
Dear $userName,
This is a site-specific email message. EOF
Administrative users have full access to all hosts, plus overall status and log information.
The administrative users are the union of the unix/linux group $Conf{CgiAdminUserGroup} and the manual list of users, separated by spaces, in $Conf{CgiAdminUsers}. If you don't want a group or manual list of users set the corresponding configuration setting to undef or an empty string.
If you want every user to have admin privileges (careful!), set $Conf{CgiAdminUsers} = '*'.
Examples:
$Conf{CgiAdminUserGroup} = 'admin'; $Conf{CgiAdminUsers} = 'craig celia'; --> administrative users are the union of group admin, plus craig and celia.
$Conf{CgiAdminUserGroup} = ''; $Conf{CgiAdminUsers} = 'craig celia'; --> administrative users are only craig and celia'.
Currently the Language setting applies to the CGI interface and email messages sent to users. Log files and other text are still in English.
sprintf()
strings,
that each contain a single '%s' that will be replaced by the user
name. The default is a mailto: link.
$Conf{CgiUserHomePageCheck} should be an absolute file path that is used to check (via ``-f'') that the user has a valid home page. Set this to undef or an empty string to turn off this check.
$Conf{CgiUserUrlCreate} should be a full URL that points to the user's home page. Set this to undef or an empty string to turn off generation of URLs for user names.
Example:
$Conf{CgiUserHomePageCheck} = '/var/www/html/users/%s.html'; $Conf{CgiUserUrlCreate} = 'http://myhost/users/%s.html'; --> if /var/www/html/users/craig.html exists, then 'craig' will be rendered as a link to http://myhost/users/craig.html.
Example:
$Conf{CgiImageDir} = '/usr/local/apache/htdocs/BackupPC';
$Conf{CgiExt2ContentType} = { 'pl' => 'text/plain', };
Example:
$Conf{CgiImageDirURL} = '/BackupPC';
Starting with v1.4.0 BackupPC uses a X.Y.Z version numbering system, instead of X.0Y. The first digit is for major new releases, the middle digit is for significant feature releases and improvements (most of the releases have been in this category), and the last digit is for bug fixes. You should think of the old 1.00, 1.01, 1.02 and 1.03 as 1..0, 1.1.0, 1.2.0 and 1.3.0.
Additionally, patches might be made available. A patched version number is of the form X.Y.ZplN (eg: 2.1.0pl2), where N is the patch level.
Craig Barratt <cbarratt@users.sourceforge.net>
See http://backuppc.sourceforge.net.
Copyright (C) 2001-2004 Craig Barratt
Ryan Kucera contributed the directory navigation code and images for v1.5.0. He contributed the first skeleton of BackupPC_restore. He also added a significant revision to the CGI interface, including CSS tags, in v2.1.0, and designed the BackupPC logo.
Xavier Nicollet, with additions from Guillaume Filion, added the internationalization (i18n) support to the CGI interface for v2.0.0. Xavier provided the French translation fr.pm, with additions from Guillaume.
Guillaume Filion wrote BackupPC_zipCreate and added the CGI support for zip download, in addition to some CGI cleanup, for v1.5.0. Guillaume continues to support fr.pm updates for each new version.
Josh Marshall implemented the Archive feature in v2.1.0.
Ludovic Drolez supports the BackupPC Debian package.
Javier Gonzalez provided the Spanish translation, es.pm for v2.0.0.
Manfred Herrmann provided the German translation, de.pm for v2.0.0. Manfred continues to support de.pm updates for each new version, together with some help frmo Ralph Paßgang.
Lorenzo Cappelletti provided the Italian translation, it.pm for v2.1.0.
Lieven Bridts provided the Dutch translation, nl.pm, for v2.1.0, with some tweaks from Guus Houtzager.
Many people have reported bugs, made useful suggestions and helped with testing; see the ChangeLog and the mail lists.
Your name could appear here in the next version!
This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version.
This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License in the LICENSE file along with this program; if not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA.
BackupPC |