AMANDA.CONF

Section: File formats and conventions (5) Updated: 10/18/2010 Local index Up

NAME

amanda.conf - Main configuration file for Amanda, the Advanced Maryland Automatic Network Disk Archiver

DESCRIPTION

amanda.conf(5) is the main configuration file for Amanda. This manpage lists the relevant sections and parameters of this file for quick reference.

The file <CONFIG_DIR>/<config>/amanda.conf is loaded.

SYNTAX

There are a number of configuration parameters that control the behavior of the Amanda programs. All have default values, so you need not specify the parameter in amanda.conf if the default is suitable.

COMMENTS

Lines starting with # are ignored, as are blank lines. Comments may be placed on a line with a directive by starting the comment with a #. The remainder of the line is ignored.

KEYWORDS AND IDENTIFIERS

Keywords are case insensitive, i.e. mailto and MailTo are treated the same. Also, the characters '-' and '_' are interchangeable in all predefined Amanda keywords: device_property and device-property have the same meaning. This manpage uses the dashed versions, but the underscored versions will be accepted for backward compatibility

Identifiers are names which are defined in the configuration itself, such as dumptypes or interfaces. Identifiers are are case-insensitive, but sensitive to '-' vs. '_'. Identifiers should be quoted in the configuration file, although For historical reasons, the quotes are optional.

Strings are always quoted with double quotes ("), and any double quotes or backslashes within the string are escaped with a backslash:

tapelist "/path/to/tapelist"
property "escaped-string" "escaping: \\ (backslash) and \" (double-quote)"

To summarize, then:

                          # QUOTES        CASE            -/_
logdir "logs"             # required      sensitive       sensitive
send-amreport-on strange  # prohibited    insensitive     insensitive
tapetype "EXABYTE"        # optional      insensitive     sensitive

define dumptype "dt" {    # optional      insensitive     sensitive
  "dumptype-common"       # optional      insensitive     sensitive
  strategy noinc          # prohibited    insensitive     insensitive
}

VALUE SUFFIXES

Integer arguments may have one of the following (case insensitive) suffixes, some of which have a multiplier effect:

b byte bytes

: Some number of bytes.

bps

: Some number of bytes per second.

k kb kbyte kbytes kilobyte kilobytes

: Some number of kilobytes (bytes*1024).

kps kbps

: Some number of kilobytes per second (bytes*1024).
It is the default multiplier for all size options.

m mb meg mbyte mbytes megabyte megabytes

: Some number of megabytes (bytes*1024*1024).

mps mbps

: Some number of megabytes per second (bytes*1024*1024).

g gb gbyte gbytes gigabyte gigabytes

: Some number of gigabytes (bytes*1024*1024*1024).

t tb tbyte tbytes terabyte terabytes

: Some number of terabytes (bytes*1024*1024*1024*1024).

tape tapes

: Some number of tapes.

day days

: Some number of days.

week weeks

Some number of weeks (days*7).

Note
The value inf may be used in most places where an integer is expected to mean an infinite amount.

Boolean arguments may have any of the values 1, y, yes, t, true or on to indicate a true state, or 0, n, no, f, false or off to indicate a false state. If no argument is given, true is assumed.

PARAMETER ORDER

In general, the order in which parameters occur in the configuration file does not matter, with the exception of subsection inheritance. For example, if dumptype "normal-encrypt" which inherits from dumptype "normal", then "normal" must appear first in the configuration file.

STRINGS

Quoted strings in Amanda follow a common, C-like syntax. Printable characters and whitespace are kept as-is, except that the backslash character (\) is used as an escape character, and a double-quote ends the string. The allowed escape sequences are

    ESCAPE SEQUENCE     BECOMES
    \\                  \
    \"                  "
    \n                  (newline)
    \t                  (tab)
    \r                  (carriage return)
    \f                  (form-feed)
    \1 - \7
    \01 - \77
    \001 - \377         (character specified in octal)

Illegally quoted strings are handled on a "best-effort" basis, which may lead to unexpected results.

Examples:

finserver "/data/finance/XYZ Corp's \"real\" finances" finance-high eth0 -1
property "syspath" "C:\\WINDOWS\\SYSTEM"

SUBSECTIONS AND INHERITANCE

Amanda configuration files may include various subsections, each defining a set of configuration directives. Each type of subsection is described below. Note that all types of subsections can inherit from other subsections of the same type by naming the "parent" section in the "child" subsection. For example:

define dumptype global {
    record yes
    index yes
}

define dumptype nocomp {
    global      # inherit the parameters in dumptype 'global'
    compress none
}

Note that multiple inheritance is also supported by simply naming multiple parent sections in a child. Parents are implicitly expanded in place in a child, and the last occurrence of each parameter takes precedence. For example,

define tapetype par1 {
    comment "Parent 1"
    filemark 8k
    speed 300bps
    length 200M
}
define tapetype par2 {
    comment "Parent 2"
    filemark 16k
    speed 400bps
}
define tapetype child {
    par1
    par2
    filemark 32k
}

In this example, 'child' will have a filemark of 32k, a speed of 400bps, and a length of 200M.

GLOBAL PARAMETERS

org string

: Default: "daily". A descriptive name for the configuration. This string appears in the Subject line of mail reports. Each Amanda configuration should have a different string to keep mail reports distinct.

mailer string

: Default found by configure. A mail program that can send mail with 'MAILER -s "subject" user < message_file'.

mailto string

: Default: none. A space separated list of recipients for mail reports. If not specified, amdump will not send any mail.

send-amreport-on [ all | strange | error | never ]

Default: all. Specify which types of messages will trigger an email from amreport. amreport is used by amdump and amflush.

all

: Send an email on any message.

strange

: Send an email on strange or error message. A strange message occurs when the dump succeeded, but returned one or more errors unknown to Amanda.

error

: Send an email only on error messages.

never

: Never send an email.

dumpcycle int

Default: 10 days. The number of days in the backup cycle. Each disk will get a full backup at least this often. Setting this to zero tries to do a full backup each run.

: Note
This parameter may also be set in a specific dumptype (see below). This value sets the default for all dumptypes so must appear in amanda.conf before any dumptypes are defined.

runspercycle int

: Default: same as dumpcycle. The number of amdump runs in dumpcycle days. A value of 0 means the same value as dumpcycle. A value of -1 means guess the number of runs from the tapelist(5) file, which is the number of tapes used in the last dumpcycle days / runtapes.

tapecycle int

Default: 15 tapes. Specifies the number of "active" volumes - volumes that Amanda will not overwrite. While Amanda is always willing to write to a new volume, it refuses to overwrite a volume unless at least 'tapecycle -1' volumes have been written since.

It is considered good administrative practice to set the tapecycle parameter slightly lower than the actual number of tapes in use. This allows the administrator to more easily cope with damaged or misplaced tapes or schedule adjustments that call for slight adjustments in the rotation order.

Note: Amanda is commonly misconfigured with tapecycle equal to the number of tapes per dumpcycle. In this misconfiguration, amanda may erase a full dump before a new one is completed. Recovery is then impossible. The tapecycle must be at least one tape larger than the number of tapes per dumpcycle.

The number of tapes per dumpcycle is calculated by multiplying the number of amdump runs per dump cycle runspercycle (the number of amdump runs per dump cycle) and runtapes (the number of tapes used per run). Typically tapecycle is set to two or four times the tapes per dumpcycle.

usetimestamps bool

: Default: Yes. This option allows Amanda to track multiple runs per calendar day. The only reason one might disable it is that Amanda versions before 2.5.1 can't read logfiles written when this option was enabled.

label-new-tapes string

: Deprecated, use autolabel option with options volume-error empty to get equivalent behavior.
Default: not set. When set, this directive will cause Amanda to automatically write an Amanda tape label to any blank tape she encounters.

autolabel string [any] [other-config] [non-amanda] [volume-error] [empty]

Default: not set. When set, this directive will cause Amanda to automatically write an Amanda tape label to most volume she encounters. This option is DANGEROUS because when set, Amanda may erase near-failing tapes or tapes accidentally loaded in the wrong slot.

When using this directive, specify the template for new tape labels. The template should contain some number of contiguous '%' characters, which will be replaced with a generated number. Be sure to specify enough '%' characters that you do not run out of tape labels. Example: autolabel "DailySet1-%%%" empty

Note that many devices cannot distinguish an empty tape from an error condition, so it may is often necessary to include volume-error as an autolabel condition.

any

: equivalent to 'other-config non-amanda volume-error empty'

other-config

: Label volumes with a valid Amanda label that do not match our labelstr. Danger: this may erase volumes from other Amanda configurations without warning!

non-amanda

: Label volumes which do not start with data that resembles an Amanda header. Danger: this may erase volumes from other backup applications without warning!

volume-error

: Label volumes where an error occurs while trying to read the label. Danger: this may erase arbitrary volumes due to transient errors.

empty

: Label volumes where a read returns 0 bytes.

dumpuser string

: Default: "amanda". The login name Amanda uses to run the backups. The backup client hosts must allow access from the tape server host as this user via .rhosts or .amandahosts, depending on how the Amanda software was built.

printer string

: Printer to use when doing tape labels. See the lbl-templ tapetype option.

tapedev string

: Default: "null:". This parameter can either specify a device (explicitly or by referencing a device definition - see amanda-devices(7)) or a tape changer (explicitly or by referencing a device definition - see amanda-changers(7)).

device-property string string

These options can set various device properties. See amanda-devices(7) for more information on device properties and their syntax. Both strings are always quoted; the first string contains the name of the property to set, and the second contains its value. For example, to set a fixed block size of 128k, write:

device-property "BLOCK_SIZE" "128k"

property [append] string string+

: These options can set various properties, they can be used by third party software to store information in the configuration file. Both strings are quoted; the first string contains the name of the property to set, and the others contains its values. append keyword append the values to the list of values for that property.

tpchanger string

: Default: not set. The tape changer to use. In most cases, only one of tpchanger or tapedev is specified, although for backward compatibility both may be specified if tpchanger gives the name of an old changer script. See amanda-changers(7) for more information on configuring changers.

changerdev string

: Default: "dev/null". A tape changer configuration parameter. Usage depends on the particular changer defined with the tpchanger option.

changerfile string

: Default: "/usr/adm/amanda/log/changer-status". A tape changer configuration parameter. Usage depends on the particular changer defined with the tpchanger option.

runtapes int

Default: 1. The maximum number of tapes used in a single run. If a tape changer is not configured, this option is not used and should be commented out of the configuration file.

If a tape changer is configured, this may be set larger than one to let Amanda write to more than one tape.

Note that this is an upper bound on the number of tapes, and Amanda may use less.

maxdumpsize int

: Default: runtapes*tape-length. Maximum number of bytes the planner will schedule for a run.
The default unit is Kbytes if it is not specified.

Default: first. The algorithm used to choose which dump image to send to the taper.

first

: First in, first out.

firstfit

: The first dump image that will fit on the current tape.

largest

: The largest dump image.

largestfit

: The largest dump image that will fit on the current tape.

smallest

: The smallest dump image.

last

: Last in, first out.

taper-parallel-write int

: Default: 1. Amanda can write simultaneously up to that number of volume at any given time. The changer must have as many drives.

labelstr string

: Default: ".*". The tape label constraint regular expression. All tape labels generated (see amlabel(8)) and used by this configuration must match the regular expression. If multiple configurations are run from the same tape server host, it is helpful to set their labels to different strings (for example, "DAILY[0-9][0-9]*" vs. "ARCHIVE[0-9][0-9]*") to avoid overwriting each other's tapes.

tapetype string

: Default: no default. The type of tape drive associated with tapedev or tpchanger. This refers to one of the defined tapetypes in the config file (see below), which specify various tape parameters, like the length, filemark size, and speed of the tape media and device.

ctimeout int

: Default: 30 seconds. Maximum amount of time that amcheck will wait for each client host.

dtimeout int

: Default: 1800 seconds. Amount of idle time per disk on a given client that a dumper running from within amdump will wait before it fails with a data timeout error.

etimeout int

: Default: 300 seconds. Amount of time per estimate on a given client that the planner step of amdump will wait to get the dump size estimates (note: Amanda runs up to 3 estimates for each DLE). For instance, with the default of 300 seconds and four DLE's, each estimating level 0 and level 1 on client A, planner will wait up to 40 minutes for that machine. A negative value will be interpreted as a total amount of time to wait per client instead of per disk.

connect-tries int

: Default: 3. How many times the server will try a connection.

req-tries int

: Default: 3. How many times the server will resend a REQ packet if it doesn't get the ACK packet.

netusage int

: Default: 8000 Kbps. The maximum network bandwidth allocated to Amanda, in Kbytes per second. See also the interface section.

inparallel int

: Default: 10. The maximum number of backups that Amanda will attempt to run in parallel. Amanda will stay within the constraints of network bandwidth and holding disk space available, so it doesn't hurt to set this number a bit high. Some contention can occur with larger numbers of backups, but this effect is relatively small on most systems.

displayunit "k|m|g|t"

: Default: "k". The unit used to print many numbers, k=kilo, m=mega, g=giga, t=tera.

dumporder string

Default: "tttTTTTTTT". The priority order of each dumper:

s: smallest size
S: largest size
t: smallest time
T: largest time
b: smallest bandwidth
B: largest bandwidth

maxdumps int

: Default: 1. The maximum number of backups from a single host that Amanda will attempt to run in parallel. See also the inparallel option.
Note that this parameter may also be set in a specific dumptype (see below). This value sets the default for all dumptypes so must appear in amanda.conf before any dumptypes are defined.

bumpsize int

Default: 10 Mbytes. The minimum savings required to trigger an automatic bump from one incremental level to the next, expressed as size. If Amanda determines that the next higher backup level will be this much smaller than the current level, it will do the next level. The value of this parameter is used only if the parameter bumppercent is set to 0.

The default unit is Kbytes if it is not specified.

The global setting of this parameter can be overwritten inside of a dumptype-definition.

See also the options bumppercent, bumpmult and bumpdays.

bumppercent int

Default: 0. The minimum savings required to trigger an automatic bump from one incremental level to the next, expressed as percentage of the current size of the DLE (size of current level 0). If Amanda determines that the next higher backup level will be this much smaller than the current level, it will do the next level.

If this parameter is set to 0, the value of the parameter bumpsize is used to trigger bumping.

The global setting of this parameter can be overwritten inside of a dumptype-definition.

See also the options bumpsize, bumpmult and bumpdays.

bumpmult float

: Default: 1.5. The bump size multiplier. Amanda multiplies bumpsize by this factor for each level. This prevents active filesystems from bumping too much by making it harder to bump to the next level. For example, with the default bumpsize and bumpmult set to 2.0, the bump threshold will be 10 Mbytes for level one, 20 Mbytes for level two, 40 Mbytes for level three, and so on.
The global setting of this parameter can be overwritten inside of a dumptype-definition.

bumpdays int

: Default: 2 days. To insure redundancy in the dumps, Amanda keeps filesystems at the same incremental level for at least bumpdays days, even if the other bump threshold criteria are met.
The global setting of this parameter can be overwritten inside of a dumptype-definition.

diskfile string

: Default: "disklist". The file name for the disklist file holding client hosts, disks and other client dumping information.

infofile string

: Default: "/usr/adm/amanda/curinfo". The file or directory name for the historical information database. If Amanda was configured to use DBM databases, this is the base file name for them. If it was configured to use text formated databases (the default), this is the base directory and within here will be a directory per client, then a directory per disk, then a text file of data.

logdir string

: Default: "/usr/adm/amanda". The directory for the amdump and log files.

indexdir string

: Default "/usr/adm/amanda/index". The directory where index files (backup image catalogues) are stored. Index files are only generated for filesystems whose dumptype has the index option enabled.

tapelist string

: Default: "tapelist". The file name for the active tapelist(5). Amanda maintains this file with information about the active set of tapes.

device-output-buffer-size int

: Default: 1280k. Controls the amount of memory used by Amanda to hold data as it is read from the network or disk before it is written to the output device. Higher values may be useful on fast tape drives and optical media.
The default unit is bytes if it is not specified.

tapebufs int

: Default: 20. This option is deprecated; use the device-output-buffer-size directive instead. tapebufs works the same way, but the number specified is multiplied by the device blocksize prior to use.

reserve int

: Default: 100. The part of holding-disk space that should be reserved for incremental backups if no tape is available, expressed as a percentage of the available holding-disk space (0-100). By default, when there is no tape to write to, degraded mode (incremental) backups will be performed to the holding disk. If full backups should also be allowed in this case, the amount of holding disk space reserved for incrementals should be lowered.

autoflush bool

: Default: off. Whether an amdump run will flush the dumps from holding disk to tape.

amrecover-do-fsf bool

: Deprecated; amrecover always uses fsf, and does not invoke amrestore.
Default: on. Amrecover will call amrestore with the -f flag for faster positioning of the tape.

amrecover-check-label bool

: Deprecated; amrecover always checks the label, and does not invoke amrestore.
Default: on. Amrecover will call amrestore with the -l flag to check the label.

amrecover-changer string

: Default: not set. Amrecover will use the changer if you use 'settape <string>' and that string is the same as the amrecover-changer setting.

columnspec string

default: "HostName=0:12:12,Disk=1:11:11,Level=1:1:1,OrigKB=1:-7:0,OutKB=1:-7:0,Compress=1:-6:1,DumpTime=1:-7:7,Dumprate=1:-6:1,TapeTime=1:-6:6,TapeRate=1:-6:1"

Defines the width of columns amreport should use. String is a comma (',') separated list of triples. Each triple consists of three parts which are separated by a equal sign ('=') and a colon (':') (see the example). These four parts specify:

1. the name of the column, which may be:

        Compress (compression ratio)
        Disk (client disk name)
        DumpRate (dump rate in KBytes/sec)
        DumpTime (total dump time in hours:minutes)
        HostName (client host name)
        Level (dump level)
        OrigKB (original image size in KBytes)
        OutKB (output image size in KBytes)
        TapeRate (tape writing rate in KBytes/sec)
        TapeTime (total tape time in hours:minutes)

: 2. the amount of space to display before the column (used to get whitespace between columns).

: 3. the width of the column itself. If set to a negative value, the width will be calculated on demand to fit the largest entry in this column.

: 4. the precision of the column, number of digit after the decimal point for number.

Here is an example:

columnspec "Disk=1:18,HostName=0:10,OrigKB=::2,OutKB=1:7"

The above will display the disk information in 18 characters and put one space before it. The hostname column will be 10 characters wide with no space to the left. The Original KBytes print 2 decimal digit. The output KBytes column is seven characters wide with one space before it.

includefile string

: Default: no default. The name of an Amanda configuration file to include within the current file. Useful for sharing dumptypes, tapetypes and interface definitions among several configurations. Relative pathnames are relative to the configuration directory.

debug-days int

: Default: 3. The number of days the debug files are kept.

debug-auth int

: Default: 0. Debug level of the auth module

debug-event int

: Default: 0. Debug level of the event module

debug-holding int

: Default: 0. Debug level of the holdingdisk module

debug-protocol int

: Default: 0. Debug level of the protocol module

debug-planner int

: Default: 0. Debug level of the planner process

debug-driver int

: Default: 0. Debug level of the driver process

debug-dumper int

: Default: 0. Debug level of the dumper process

debug-chunker int

: Default: 0. Debug level of the chunker process

debug-taper int

: Default: 0. Debug level of the taper process

flush-threshold-dumped int

Default: 0. Amanda will not begin writing data to a new volume until the amount of data on the holding disk is at least this percentage of the volume size. In other words, Amanda will not begin until the amount of data on the holding disk is greater than the tape length times this parameter. This parameter may be larger than 100%, for example to keep more recent dumps on the holding disk for faster recovery.

Needless to say, your holding disk must be big enough that this criterion could be satisfied. If the holding disk cannot be used for a particular dump (because, for example, there is no remaining holding space) then Amanda will disregard the constraint specified by this setting and start a new volume anyway. Once writing to a volume has begun, this constraint is not applied unless and until a new volume is needed.

The value of this parameter may not exceed than that of the flush-threshold-scheduled parameter.

flush-threshold-scheduled int

Default: 0. Amanda will not begin writing data to a new volume until the sum of the amount of data on the holding disk and the estimated amount of data remaining to be dumped during this run is at least this percentage of the volume size. In other words, Amanda will not begin until the inequality h + s > t × d is satisfied, where h is the amount of data on the holding disk, s is the total amount of data scheduled for this run but not dumped yet, t is the capacity of a volume, and d is this parameter, expressed as a percentage. This parameter may be larger than 100%.

The value of this parameter may not be less than that of the flush-threshold-dumped or taperflush parameters.

taperflush int

: Default: 0. At the end of a run, Amanda will start a new tape to flush remaining data if there is more data on the holding disk at the end of a run than this setting allows; the amount is specified as a percentage of the capacity of a single volume. In other words, at the end of a run, Amanda will begin a new tape if the inequality h > t × f is satisfied, where h is the amount of data remaining on the holding disk from this or previous runs, t is the capacity of a volume, and f is this parameter, expressed as a percentage. This parameter may be greater than 100%.
The value of this parameter may not exceed that of the flush-threshold-scheduled parameter.; autoflush must be set to 'yes' if taperflush is greater than 0.

reserved-udp-port int,int

: Default: --with-udpportrange or 512,1023. Reserved udp port that will be used (bsd, bsdudp). Range is inclusive.

reserved-tcp-port int,int

: Default: --with-low-tcpportrange or 512,1023. Reserved tcp port that will be used (bsdtcp). Range is inclusive.

unreserved-tcp-port int,int

: Default: --with-tcpportrange or 1024,65535. Unreserved tcp port that will be used (bsd, bsdudp). Range is inclusive.

recovery-limit [ string | same-host ]

Default: none (no limitations). This parameter limits the hosts that may do remote recoveries. Hosts are identified by their authenticated peer name, as described in amanda-auth(7); if this is not available and the recovery-limit parameter is present, recovery will be denied. The arguments to the parameter are strings giving host match expressions (see amanda-match(7)) or the special keyword same-host, which requires an exact match to the hostname of the DLE being recovered. Specifying no arguments at all will disable all recoveries from any host.

Note that match expressions can be constructed to be forgiving of e.g., fully-qualified vs. unqualified hostnames, but same-host requires an exact match.

The error messages that appear in amrecover are intentionally vague to avoid information leakage. Consult the amindexd debug log for more details on the reasons a recovery was rejected.

Recovery limits can be refined on a per-DLE basis using the dumptype parameter of the same name. Note that the default value will apply to any dumpfiles for disks which no longer appear in the disklist; thus leaving the global parameter at its default value but setting it for all DLEs is not sufficient to maintain secure backups.

HOLDINGDISK SECTION

The amanda.conf file may define one or more holding disks used as buffers to hold backup images before they are written to tape. The syntax is:

define holdingdisk name {
    holdingdisk-option holdingdisk-value
    ...
}

The { must appear at the end of a line, and the } on its own line.

Name is a logical name for this holding disk.

The options and values are:

comment string

: Default: not set. A comment string describing this holding disk.

directory string

: Default: "/dumps/amanda". The path to this holding area.

use int

: Default: 0 Gb. Amount of space that can be used in this holding disk area. If the value is zero, all available space on the file system is used. If the value is negative, Amanda will use all available space minus that value.

chunksize int

Default: 1 Gb. Holding disk chunk size. Dumps larger than the specified size will be stored in multiple holding disk files. The size of each chunk will not exceed the specified value. However, even though dump images are split in the holding disk, they are concatenated as they are written to tape, so each dump image still corresponds to a single continuous tape section.

The default unit is Kbytes if it is not specified.

If 0 is specified, Amanda will create holding disk chunks as large as ((INT_MAX/1024)-64) Kbytes.

Each holding disk chunk includes a 32 Kbyte header, so the minimum chunk size is 64 Kbytes (but that would be really silly).

Operating systems that are limited to a maximum file size of 2 Gbytes actually cannot handle files that large. They must be at least one byte less than 2 Gbytes. Since Amanda works with 32 Kbyte blocks, and to handle the final read at the end of the chunk, the chunk size should be at least 64 Kbytes (2 * 32 Kbytes) smaller than the maximum file size, e.g. 2047 Mbytes.

DUMPTYPE SECTION

The amanda.conf(5) file may define multiple sets of backup options and refer to them by name from the disklist(5) file. For instance, one set of options might be defined for file systems that can benefit from high compression, another set that does not compress well, another set for file systems that should always get a full backup and so on.

A set of backup options are entered in a dumptype section, which looks like this:

define dumptype "name" {
    dumptype-option dumptype-value
    ...
}

The { must appear at the end of a line, and the } on its own line.

Name is the name of this set of backup options. It is referenced from the disklist(5) file.

Some of the options in a dumptype section are the same as those in the main part of amanda.conf(5). The main option value is used to set the default for all dumptype sections. For instance, setting dumpcycle to 50 in the main part of the config file causes all following dumptype sections to start with that value, but the value may be changed on a section by section basis. Changes to variables in the main part of the config file must be done before (earlier in the file) any dumptypes are defined.

The dumptype options and values are:

auth string

: Default: "bsd". Type of authorization to perform between tape server and backup client hosts. See amanda-auth(7) for more detail.

amandad-path string

: Default: "$libexec/amandad". Specify the amandad path of the client, only use with rsh/ssh authentification.

client-username string

: Default: CLIENT_LOGIN. Specify the username to connect on the client, only use with rsh/ssh authentification.

client-port [ int | string ]

: Default: "amanda". Specifies the port to connect to on the client. It can be a service name or a numeric port number.

bumpsize int

The default unit is Kbytes if it is not specified.

See also the options bumppercent, bumpmult and bumpdays.

bumppercent int

If this parameter is set to 0, the value of the parameter bumpsize is used to trigger bumping.

See also the options bumpsize, bumpmult and bumpdays.

bumpmult float

: Default: 1.5. The bump size multiplier. Amanda multiplies bumpsize by this factor for each level. This prevents active filesystems from bumping too much by making it harder to bump to the next level. For example, with the default bumpsize and bumpmult set to 2.0, the bump threshold will be 10 Mbytes for level one, 20 Mbytes for level two, 40 Mbytes for level three, and so on.

bumpdays int

: Default: 2 days. To insure redundancy in the dumps, Amanda keeps filesystems at the same incremental level for at least bumpdays days, even if the other bump threshold criteria are met.

comment string

: Default: not set. A comment string describing this set of backup options.

comprate float [, float ]

: Default: 0.50, 0.50. The expected full and incremental compression factor for dumps. It is only used if Amanda does not have any history information on compression rates for a filesystem, so should not usually need to be set. However, it may be useful for the first time a very large filesystem that compresses very little is backed up.

compress [ none | client | server ] [ best | fast | custom ]

Default: client fast. If Amanda does compression of the backup images, it can do so either on the backup client host before it crosses the network or on the tape server host as it goes from the network into the holding disk or to tape. Which place to do compression (if at all) depends on how well the dump image usually compresses, the speed and load on the client or server, network capacity, holding disk capacity, availability of tape hardware compression, etc.

For either type of compression, Amanda also allows the selection of three styles of compression. best is the best compression available, often at the expense of CPU overhead. fast is often not as good a compression as best, but usually less CPU overhead. Or to specify custom to use your own compression method. (See dumptype custom-compress in example/amanda.conf for reference)

So the compress options line may be one of:

compress none

compress client fast

compress client best

compress client custom

: Specify client-custom-compress "PROG"
PROG must not contain white space and it must accept -d for uncompress.

compress server fast

compress server best

compress server custom

: Specify server-custom-compress "PROG"
PROG must not contain white space and it must accept -d for uncompress.

Note that some tape devices do compression and this option has nothing to do with whether that is used. If hardware compression is used (usually via a particular tape device name or mt option), Amanda (software) compression should be disabled.

client-custom-compress string

: Default: none. The program to use to perform compression/decompression on the client; used with "compress client custom". Must not contain whitespace. Must accept -d to uncompress.

server-custom-compress string

: Default: none. The program to use to perform compression/decompression on the server; used with "compress server custom". Must not contain whitespace. Must accept -d to uncompress.

dumpcycle int

: Default: 10 days. The number of days in the backup cycle. Each disk using this set of options will get a full backup at least this of ten. Setting this to zero tries to do a full backup each run.

encrypt [ none | client | server ]

Default: not set. To encrypt backup images, it can do so either on the backup client host before it crosses the network or on the tape server host as it goes from the network into the holding disk or to tape.

So the encrypt options line may be one of:

encrypt none

encrypt client

Specify client-encrypt "PROG"

PROG must not contain white space.

Specify client-decrypt-option "decryption-parameter" Default: "-d"

decryption-parameter must not contain white space.

(See dumptype client-encrypt-nocomp in example/amanda.conf for reference)

encrypt server

Specify server-encrypt "PROG"

PROG must not contain white space.

Specify server-decrypt-option "decryption-parameter" Default: "-d"

decryption-parameter must not contain white space.

(See dumptype server-encrypt-fast in example/amanda.conf for reference)

Note that current logic assumes compression then encryption during backup(thus decrypt then uncompress during restore). So specifying client-encryption AND server-compression is not supported. amcrypt which is a wrapper of aespipe is provided as a reference symmetric encryption program.

client-encrypt string

: Default: none. The program to use to perform encryption/decryption on the client; used with "encrypt client". Must not contain whitespace.

client-decrypt-option string

: Default: -d. The option that can be passed to client-encrypt to make it decrypt instead. Must not contain whitespace.

server-encrypt string

: Default: none. The program to use to perform encryption/decryption on the server; used with "encrypt server". Must not contain whitespace.

server-decrypt-option string

: Default: -d. The option that can be passed to server-encrypt to make it decrypt instead. Must not contain whitespace.

estimate [ client | calcsize | server ]+

Default: client. Determine the way Amanda estimates the size of each DLE before beginning a backup. This is a list of acceptable estimate methods, and Amanda applies the first method supported by the application. The methods are:

client

: Use the same program as the dumping program. This is the most accurate method to do estimates, but it can take a long time.

calcsize

: Use a faster program to do estimates, but the result is less accurate.

server

: Use only statistics from the previous few runs to give an estimate. This very quick, but the result is not accurate if your disk usage changes from day to day. If this method is specified, but the server does not have enough data to make an estimate, then the option is internally moved to the end of the list, thereby preferring 'client' or 'calcsize' in this case.

exclude [ list | file ][[optional][append][ string ]+]

Default: file. Exclude is the opposite of include and specifies files that will be excluded from the backup. The format of the exclude expressions depends on the application, and some applications do not support excluding files at all.

There are two exclude parameters, exclude file and exclude list. With exclude file, the string is an exclude expression. With exclude list , the string is a file name on the client containing GNU-tar exclude expressions. The path to the specified exclude list file, if present (see description of 'optional' below), must be readable by the Amanda user.

All exclude expressions are concatenated in one file and passed to the application as an --exclude-from argument.

For GNU-tar, exclude expressions must always be specified as relative to the top-level directory of the DLE, and must start with "./". See the manpages for individual applications for more information on supported exclude expressions.

With the append keyword, the string is appended to the current list, without it, the string overwrites the list.

If optional is specified for exclude list, then amcheck will not complain if the file doesn't exist or is not readable.

For exclude list, if the file name is relative, the disk name being backed up is prepended. So if this is entered:

    exclude list ".amanda.excludes"

the actual file used would be /var/.amanda.excludes for a backup of /var, /usr/local/.amanda.excludes for a backup of /usr/local, and so on.

holdingdisk [ never | auto | required ]

Default: auto. Whether a holding disk should be used for these backups or whether they should go directly to tape. If the holding disk is a portion of another file system that Amanda is backing up, that file system should refer to a dumptype with holdingdisk set to never to avoid backing up the holding disk into itself.

never|no|false|off

: Never use a holdingdisk, the dump will always go directly to tape. There will be no dump if you have a tape error.

auto|yes|true|on

: Use the holding disk, unless there is a problem with the holding disk, the dump won't fit there or the medium doesn't require spooling (e.g., VFS device)

required

: Always dump to holdingdisk, never directly to tape. There will be no dump if it doesn't fit on holdingdisk

ignore boolean

: Default: no. Whether disks associated with this backup type should be backed up or not. This option is useful when the disklist file is shared among several configurations, some of which should not back up all the listed file systems.

include [ list | file ][[optional][append][ string ]+]

Default: file ".". There are two include lists, include file and include list. With include file , the string is a glob expression. With include list , the string is a file name on the client containing glob expressions.

All include expressions are expanded by Amanda, concatenated in one file and passed to GNU-tar as a --files-from argument. They must start with "./" and contain no other "/".

Include expressions must always be specified as relative to the head directory of the DLE.

: Note
For globbing to work at all, even the limited single level, the top level directory of the DLE must be readable by the Amanda user.

With the append keyword, the string is appended to the current list, without it, the string overwrites the list.

If optional is specified for include list, then amcheck will not complain if the file doesn't exist or is not readable.

For include list, If the file name is relative, the disk name being backed up is prepended.

index boolean

: Default: no. Whether an index (catalogue) of the backup should be generated and saved in indexdir. These catalogues are used by the amrecover utility.

kencrypt boolean

: Default: no. Whether the backup image should be encrypted by Kerberos as it is sent across the network from the backup client host to the tape server host.

maxdumps int

: Default: 1. The maximum number of backups from a single host that Amanda will attempt to run in parallel. See also the main section parameter inparallel.

maxpromoteday int

: Default: 10000. The maximum number of day for a promotion, set it 0 if you don't want promotion, set it to 1 or 2 if your disks get overpromoted.

priority [ low | medium | high ]

: Default: medium. When there is no tape to write to, Amanda will do incremental backups in priority order to the holding disk. The priority may be high (2), medium (1), low (0) or a number of your choice.

program [ "DUMP" | "GNUTAR" | "APPLICATION" ]

Default: "DUMP". The type of backup to perform. Valid values are:

"DUMP"

: The native operating system backup program.

"GNUTAR"

: To use GNU-tar or to do PC backups using Samba.

"APPLICATION"

: To use an application, see the application option.

application string

: No default. Must be the name of an application if program is set to APPLICATION. See APPLICATION SECTION below.

script string

: No default. Must be the name of a script. You can have many script. See SCRIPT SECTION below.

property [append] string string+

: These options can set various properties, they can be used by third party software to store information in the configuration file. Both strings are quoted; the first string contains the name of the property to set, and the others contains its values. append keyword append the values to the list of values for that property.

record boolean

: Default: yes. Whether to ask the backup program to update its database (e.g. /var/lib/dumpdates for DUMP or /usr/local/var/amanda/gnutar-lists for GNUTAR) of time stamps. This is normally enabled for daily backups and turned off for periodic archival runs.

skip-full boolean

: Default: no. If true and planner has scheduled a full backup, these disks will be skipped, and full backups should be run off-line on these days. It was reported that Amanda only schedules level 1 incrementals in this configuration; this is probably a bug.

skip-incr boolean

: Default: no. If true and planner has scheduled an incremental backup, these disks will be skipped.

ssh-keys string

: Default: not set. The key file the ssh auth will use, it must be the private key. If this parameter is not specified, then the default ssh key will be used.

starttime int

: Default: not set. Backup of these disks will not start until after this time of day. The value should be hh*100+mm, e.g. 6:30PM (18:30) would be entered as 1830.

strategy [ standard | nofull | noinc | skip | incronly ]

Default: standard. Strategy to use when planning what level of backup to run next. Values are:

standard

: The standard Amanda schedule.

nofull

: Never do full backups, only level 1 incrementals.

noinc

: Never do incremental backups, only full dumps.

skip

: Treat this DLE as if it doesn't exist (useful to disable DLEs when sharing the disklist file between multiple configurations). Skipped DLEs will not be checked or dumped, and will not be matched by disklist expressions.

incronly

: Only do incremental dumps. amadmin force should be used to tell Amanda that a full dump has been performed off-line, so that it resets to level 1.

allow-split bool

: Default: true. If true, then dumps with this dumptype can be split on the storage media. If false, then the dump will be written in a single file on the media. See "Dump Splitting Configuration" below.

tape-splitsize int

: Deprecated. See "Dump Splitting Configuration" below.
Default: not set. Split dump file on tape into pieces of a specified size. The default unit is Kbytes if it is not specified.

split-diskbuffer string

: Deprecated. See "Dump Splitting Configuration" below. Default: not set. When dumping a split dump in PORT-WRITE mode (usually meaning "no holding disk"), buffer the split chunks to a file in the directory specified by this option.

fallback-splitsize int

: Deprecated. See "Dump Splitting Configuration" below.
Default: 10M. This specifies the part size used when no split-diskbuffer is specified, or when it is too small or does not exist, and thus the maximum amount of memory consumed for in-memory splitting. The default unit is Kbytes if it is not specified.

recovery-limit [ same-host | string ]*

: Default: global value. This parameter overrides the global recovery-limit parameter for DLEs of this dumptype.

The following dumptype entries are predefined by Amanda:

define dumptype "no-compress" {
    compress none
}
define dumptype "compress-fast" {
    compress client fast
}
define dumptype "compress-best" {
    compress client best
}
define dumptype "srvcompress" {
    compress server fast
}
define dumptype "bsd-auth" {
    auth bsd
}
define dumptype "no-record" {
    record no
}
define dumptype "no-hold" {
    holdingdisk no
}
define dumptype "no-full" {
    skip-full yes
}

In addition to options in a dumptype section, one or more other dumptype names may be supplied as identifiers, which make this dumptype inherit options from other previously defined dumptypes. For instance, two sections might be the same except for the record option:

define dumptype "normal" {
    comment "Normal backup, no compression, do indexing"
    no-compress
    index yes
    maxdumps 2
}
define dumptype "testing" {
    comment "Test backup, no compression, do indexing, no recording"
    "normal"
    record no
}

Amanda provides a dumptype named global in the sample amanda.conf file that all dumptypes should reference. This provides an easy place to make changes that will affect every dumptype, although you must be careful that every dumptype explicitly inherits from the global dumptype - Amanda does not do so automatically.

TAPETYPE SECTION

The amanda.conf file may define multiple types of tape media and devices. The information is entered in a tapetype section, which looks like this in the config file:

define tapetype "name" {
    tapetype-option tapetype-value
    ...
}

The { must appear at the end of a line, and the } on its own line.

Name is the name of this type of tape medium/device. It is referenced from the tapetype option in the main part of the config file.

The tapetype options and values are:

comment string

: Default: not set. A comment string describing this set of tape information.

filemark int

: Default: 1 kbytes. How large a file mark (tape mark) is, measured in kbytes. If the size is only known in some linear measurement (e.g. inches), convert it to kbytes using the device density.

length int

: Default: 2000 kbytes. How much data will fit on a tape, expressed in kbytes.
Note that this value is only used by Amanda to schedule which backups will be run. Once the backups start, Amanda will continue to write to a tape until it gets an error, regardless of what value is entered for length (but see amanda-devices(7) for exceptions).

blocksize int

: Default: 32 kbytes. How much data will be written in each tape record, expressed in kbytes. This is similar to the BLOCK_SIZE device property, but if the blocksize is not a multiple of 1024 bytes, then this parameter cannot be used to specify it, and the property must be used instead.

readblocksize int

: Default: 32 kytes How much data will be read in each tape record. This can be used to override a device's block size for reads only. This may be useful, for example, in reading a tape written with a 256k block size when Amanda is configured to use 128k blocks. This unusual feature is not supported by all operating systems and tape devices.
The default unit is Kbytes if it is not specified.

part-size int

: Default: none. This is the size (in KB if no units are specified) of each split part written to the volume. It is reduced to part-cache-max-size when part caching is required. If this is set to zero, then no splitting will take place; in this case, some devices can span dumps from volume to volume, while others will cause the entire dump to fail if they encounter end-of-medium before the dump is complete. See "Dump Splitting Configuration" below.

part-cache-type [ none | disk | memory ]

: Default: none. When part caching is required, this parameter specifies the type of caching that will be used. The options include no caching (none), in which case a failed part will cause the entire dump to fail; on-disk caching (disk), for which part-cache-dir must be set properly; and in-memory caching (memory), which on most systems severely restrains the size of the part that can be written. See "Dump Splitting Configuration" below.

part-cache-dir string

: Default: none. The directory in which part-cache files can be written when caching on disk. See "Dump Splitting Configuration" below.

part-cache-max-size int

: Default: none. The maximum part size to use when caching is in effect. This is used to limit the part size when disk or memory space for caching is constrained. This value must be greater than zero.

speed int

: Default: 200 bps. How fast the drive will accept data, in bytes per second. This parameter is NOT currently used by Amanda.

lbl-templ string

: Default: not set. A PostScript template file used by amreport to generate labels. Several sample files are provided with the Amanda sources in the example directory. See the amreport(8) man page for more information.

In addition to options, another tapetype name may be supplied as an identifier, which makes this tapetype inherit options from another tapetype. For instance, the only difference between a DLT4000 tape drive using Compact-III tapes and one using Compact-IV tapes is the length of the tape. So they could be entered as:

define tapetype "DLT4000-III" {
    comment "DLT4000 tape drives with Compact-III tapes"
    length 12500 mbytes         # 10 Gig tapes with some compression
    filemark 2000 kbytes
    speed 1536 kps
}
define tapetype "DLT4000-IV" {
    "DLT4000-III"
    comment "DLT4000 tape drives with Compact-IV tapes"
    length 25000 mbytes         # 20 Gig tapes with some compression
}

INTERFACE SECTION

The amanda.conf file may define multiple types of network interfaces. The information is entered in an interface section, which looks like this:

define interface "name" {
    interface-option interface-value
    ...
}

The { must appear at the end of a line, and the } on its own line.

name is the name of this type of network interface. It is referenced from the disklist file.

Note that these sections define network interface characteristics, not the actual interface that will be used. Nor do they impose limits on the bandwidth that will actually be taken up by Amanda. Amanda computes the estimated bandwidth each file system backup will take based on the estimated size and time, then compares that plus any other running backups with the limit as another of the criteria when deciding whether to start the backup. Once a backup starts, Amanda will use as much of the network as it can leaving throttling up to the operating system and network hardware.

The interface options and values are:

comment string

: Default: not set. A comment string describing this set of network information.

use int

: Default: 8000 Kbps. The speed of the interface in Kbytes per second.

In addition to options, another interface name may be supplied as an identifier, which makes this interface inherit options from another interface. At the moment, this is of little use.

APPLICATION SECTION

The amanda.conf file may define multiple types of application. The information is entered in a application section, which looks like this:

define application "name" {
    application-option application-value
    ...
}

The { must appear at the end of a line, and the } on its own line.

name is the name of this type of application. It is referenced from the dumptype

The application options and values are:

comment string

: Default: not set. A comment string describing this application.

plugin string

: No default. Must be set to the name of the program. This program must be in the $libexecdir/amanda/application directory on the client.

property [append] [priority] string string+

: No default. You can set property for the application, each application have a different set of property. Both strings are quoted; the first string contains the name of the property to set, and the others contains its values. append keyword append the values to the list of values for that property. priority keyword disallow the setting of that property on the client.

SCRIPT SECTION

The amanda.conf file may define multiple types of script. The information is entered in a script section, which looks like this:

define script "name" {
    script-option script-value
    ...
}

The { must appear at the end of a line, and the } on its own line.

name is the name of this type of script. It is referenced from the dumptype

The script options and values are:

comment string

: Default: not set. A comment string describing this script.

plugin string

: No default. Must be set to the name of the program. This program must be in the $libexecdir/amanda/application directory on the client and/or server.

order int

: Default: 5000. Scripts are executed in that order, it is useful if you have many scripts and they must be executed in a spefific order.

execute-where [ client | server ]

: Default: client. Where the script must be executed, on the client or server.

execute-on execute_on [,execute_on]*

No default. When the script must be executed, you can specify many of them:

pre-dle-amcheck

: Execute before the amcheck command for the dle.

pre-host-amcheck

: Execute before the amcheck command for all dle for the client.

post-dle-amcheck

: Execute after the amcheck command for the dle.

post-host-amcheck

: Execute after the amcheck command for all dle for the client.

pre-dle-estimate

: Execute before the estimate command for the dle.

pre-host-estimate

: Execute before the estimate command for all dle for the client.

post-dle-estimate

: Execute after the estimate command for the dle.

post-host-estimate

: Execute after the estimate command for all dle for the client.

pre-dle-backup

: Execute before the backup command for the dle.

pre-host-backup

: Execute before the backup command for all dle for the client. It can't be run on client, it must be run on server

post-dle-backup

: Execute after the backup command for the dle.

post-host-backup

: Execute after the backup command for all dle for the client. It can't be run on client, it must be run on server

pre-recover

: Execute before any level is recovered.

post-recover

: Execute after all levels are recovered.

pre-level-recover

: Execute before each level recovery.

post-level-recover

: Execute after each level recovery.

inter-level-recover

: Execute between two levels of recovery.

If you recover level 0 and 2 of the disk /usr with amrecover, it will execute:

script --pre-recover
script --pre-level-recover --level 0
#recovering level 0
script --post-level-recover --level 0
script --inter-level-recover --level 0 --level 2
script --pre-level-recover --level 2
#recovering level 2
script --post-level-recover --level 2
script --post-recover

property [append] [priority] string string+

: No default. You can set property for the script, each script have a different set of property. Both strings are quoted; the first string contains the name of the property to set, and the others contains its values. append keyword append the values to the list of values for that property. priority keyword disallow the setting of that property on the client.

DEVICE SECTION

Backend storage devices are specified in amanda.conf in the form of "device" sections, which look like this:

define device name {
    commend "comment (optional)"
    tapedev "device-specifier"
    device-property "prop-name" "prop-value"
    ...
}

The { must appear at the end of a line, and the } on its own line.

name is the user-specified name of this device. It is referenced from the global tapedev parameter. The device-specifier specifies the device name to use; see amanda-devices(7). As with most sections, the comment parmeter is optional and only for the user's convenience.

An arbitrary number of device-property parameters can be specified. Again, see amanda-devices(7) for information on device properties.

CHANGER SECTION

Changers are described in amanda.conf in the form of "changer" sections, which look like this:

define changer name {
    comment "comment (optional)"
    tpchanger "changer-spec"
    changerdev "device-name"
    changerfile "state-file"
    ...
}

The { must appear at the end of a line, and the } on its own line.

name is the user-specified name of this device. The remaining parameters are specific to the changer type selected.

See amanda-changers(7) for more information on configuring changers.

DUMP SPLITTING CONFIGURATION

Amanda can "split" dumps into parts while writing them to storage media. This allows Amanda to recover gracefully from a failure while writing a part to a volume, by simply selecting a new volume and re-writing the dump from the beginning of the failed part. Parts also allow Amanda to seek directly to the required data, although this functionality is not yet used.

In order to support re-writing from the beginning of a failed part, Amanda must have access to the contents of the part after it has been partially written. If the dump is being read from holding disk, then the part contents are availble there. Otherwise, the part must be cached, and this can be done memory or on disk. In either of the latter cases, the cache must have enough space to hold an entire part.

Because it is common for a single Amanda configuration to use both holding-disk (FILE-WRITE) and direct (known as PORT-WRITE) dumps, Amanda allows the configuration of different split sizes for the two cases. This allows, for example, for a part size appropriate to large tapes when performing FILE-WRITE dumps, with a part size limited by available disk or memory when performing PORT-WRITE dumps.

Selecting a proper split size is a delicate matter. If the parts are too large, substantial storage space may be wasted in failed parts. If too small, large dumps will be split into innumerable tiny dumpfiles, adding to restoration complexity; furthermore, an excess of filemarks will cause slower tape drive operation and reduce the usable space on tape. A good rule of thumb is 1/10 of the size of a volume of storage media.

In versions of Amanda through 3.1.*, splitting was controlled by the dumptype parameters tape-splitsize, split-diskbuffer, and fallback-splitsize. These keywords had confusing and non-intuitive interactions, and have since been deprecated.

If the deprecated keywords are not present, subsequent versions of Amanda use the dumptype parameter allow-split to control whether a DLE can be split, and the tapetype parameters part-size, part-cache-type, part-cache-dir, and part-cache-max-size. The part-size specifies the "normal" part size, while the part-cache-* parameters describe how to behave when caching is required (on PORT-WRITE). Full details on these parameters are given above.