NAME
rrdpoller - Retrieve RRD file data and apply some threshold algorithm
SYNOPSIS
rrdpoller list|ls filename
rrdpoller get filename datasource [--offset time] [--cf func-name]
rrdpoller exact filename datasource value
rrdpoller boundaries filename datasource [--min number] [--max number]
rrdpoller relation filename datasource [<|>]threshold[%] [--target filename] [--compare-ds datasource] [--offset time]
rrdpoller quotient filename datasource [<|>]threshold[%] [--target filename] [--target-ds datasource] [--offset time]
rrdpoller hunt filename datasource roll [--parent filename] [--parent-ds datasource]
DESCRIPTION
rrdpoller allows you to poll RRD files and get current value of a given datasource. Additionally it implement some advanced checks able to use past values to decide if the current one is out of threshold or not.
The main purpose of this tool is to do the interface between an existing monitoring system like BigBrother, Nagios or Mon and a set of RRD graphs generated by some specialized data collectors like rrdcollect, Cricket and such. Thus you can build a very modular monitoring/trend architecture without the need of double polling.
EXAMPLES
Those examples are fictive, they doesn't represent real numbers. It's just to let you figure out what you can do with each algorithm.
- rrdpoller ls host1-net.rrd
-
Returns the whole list of datasources contained in the file.
- rrdpoller get host1-load.rrd 1min_avg
-
Get the current value of the 1min_avg datasource of the host1-load.rrd file.
- rrdpoller exact host1-service.rrd http 1
-
Get the current value of the http datasource of the host1-load.rrd file and return false if not equal to 1.
- rrdpoller boundaries host1-load.rrd 5min_avg --max 15
-
Get the current value of the 5min_avg datasource of the host1-load.rrd file and return false if greater than 15.
- rrdpoller relation host1-disk.rrd usage 100000 --offset 1min
-
Compare the current usage datasource value of the host1-disk.rrd file with the same datasource 1 minute ago and return false if the delta is greater than 100000.
- rrdpoller quotient host1-mem.rrd usage 90% --target-ds total
-
Compare the current usage datasource of the host1-mem.rrd file and divide it by the total datasource of the same file. The command returns false if the obtained percentage is greater than 90%.
- rrdpoller hunt pop2.rrd users 40 --target pop1.rrd
-
If the users datasource of the pop2.rrd file is non-zero and the same datasource of the pop1.rrd hasn't yet reach 40, the command will return false.
SUB-COMMANDS
All sub-commands needs at least the too first arguments filename and datasource (except for list command). Except for the get and list commands (which doesn't perform a test), all sub-command will change the rrdpoller return code to false (1) if the test fails. All sub-commands will print the given datasource value of the given filename on the standard output.
list|ls filename
This command list all datasources of the given file.
get filename datasource [OPTIONS]
This sub-command the last inserted value in the given datasource of the given RRD filename. You can get an older value by giving an offset that will be expressed as: <last_inserted_time> - <offset>.
options:
- --offset time
-
Makes the get command to retrieve an earlier value. It substract the value of time to the timestamp of the last value inserted in the datasource. The time can be expressed in second (ie 60) or in RRD time reference specification format (see "TIME REFERENCE SPECIFICATION" in rrdfetch).
- --cf func-name
-
Default consolidation function is AVERAGE. This option allow you to change it to whichever you want.
exact filename datasource exact
This threshold allows you to monitor the datasource for an exact match. This is useful in cases where an enumerated (or boolean) SNMP object instruments a condition where a transition to a specific state requires attention. For example, a datasource might return either true(1) or false(2), depending on whether or not a power supply has failed.
boundaries filename datasource [OPTIONS]
This threshold takes too optional values, a minimum and a maximum value. If the datasource strays outside of this interval, the test fail.
options:
- --min number
-
If current value is lower than number, the test fail
- --max number
-
If current value is greater than number, the test fail
relation filename datasource [<|>]threshold[%] [OPTIONS]
A relation threshold considers the difference between two datasources (possibly from different RRD files), or alternatively, the difference between two temporally distinct values for the same datasource. The difference can be expressed as absolute value, or as a percentage of the second datasource (comparison) value. This difference is compared to a threshold argument with either the greater than (>) or lesser than (<) operator. The criteria fails when the expression (<absolute or relative difference> <either greater-than or less-than> <threshold>) evaluates to false.
The threshold number, optionally preceded by the greater than (>) or less than (<) symbol, and optionally followed by the symbol percent (%). If omitted, greater than is used by default and the expression: difference > threshold, is evaluated. "<10%", ">1000", "50%", and "500" are all examples of valid thresholds.
options:
- --target filename
-
The path of the comparison RRD file. This argument is optional and if omitted the first RRD file is also taken as the comparison target.
- --target-ds datasource
-
The name of the comparison datasource. This datasource must belong to the comparison RRD file. This argument is optional and if omitted the first datasource name is also taken as the comparison datasource name. If the value is a number, the value is considered as a fix value and is taked for the comparison.
- --offset time
-
The temporal offset to go back from the first value to fetch the comparison datasource value. Note that a data source value must exist in the RRD file for that exact offset. If This argument is optional and if omitted, it is set to 0. The time value can be expressed in second (ie 60) or in RRD time reference specification format (see "TIME REFERENCE SPECIFICATION" in rrdfetch).
quotient filename datasource [<|>]threshold[%] [OPTIONS]
Quotient thresholds are similar to relation thresholds, except that they consider the quotient of two data sources, or alternatively, the same data source at two different time points. For a quotient monitor threshold, the value of the first data source is computed as a percentage of the second data source value (such as 10 (first datasource) is 50% of 20 (second datasource)). This percentage is then compared to a threshold argument with either the greater than (>) or less than (<) operator. The criteria fails when the expression (<percentage> <either greater-than or less-than> <threshold>) evaluates to false.
The threshold number, optionally preceded by the greater than (>) or less than (<) symbol and followed by the symbol percent (%). If omitted, greater than is used by default and the expression: difference > threshold, is evaluated. "<10%" and "50%" are all examples of valid thresholds.
options:
- --target filename
-
The path of the comparison RRD file. This argument is optional and if omitted the first RRD file is also taken as the comparison target.
- --target-ds datasource
-
The name of the comparison datasource. This datasource must belong to the comparison RRD file. This argument is optional and if omitted the first datasource name is also taken as the comparison datasource name. If the value is a number, the value is considered as a fix value and is taked for the comparison.
- --offset time
-
The temporal offset to go back from the first value to fetch the comparison datasource value. Note that a data source value must exist in the RRD file for that exact offset. If This argument is optional and if omitted, it is set to 0. The time value can be expressed in second (ie 60) or in RRD time reference specification format (see "TIME REFERENCE SPECIFICATION" in rrdfetch).
hunt roll --parent <filename> --parent-ds <datasource> [OPTIONS]
The hunt threshold is designed for the situation where the data source serves as an overflow for another data source; that is, if one data source (the parent) is at or near capacity, then traffic will begin to appear on this (the monitored) data source. One application of hunt monitor thresholds is to identify premature rollover in a set of modem banks configured to hunt from one to the next. Specifically, the criteria of the hunt monitor threshold fails if the value of the monitored data source is non-zero and the current value of the parent data source falls below a specified capacity threshold.
The roll argument is the threshold of the parent data source. Generally this should be slightly less than the maximum capacity of the target.
options:
- --parent filename
-
The path of the comparison RRD file. This argument is optional and if omitted the first RRD file is also taken as the comparison target.
- --parent-ds datasource
-
The name of the comparison datasource. This datasource must belong to the comparison RRD file. This argument is optional and if omitted the first datasource name is also taken as the comparison datasource name. If the value is a number, the value is considered as a fix value and is taked for the comparison.
EXIT STATUS
AUTHOR
Olivier Poitrey <rs@rhapsodyk.net>
LICENCE
rrdpoller retrieves RRD file data and apply some threshold algorithm. Copyright (C) 2004 Olivier Poitrey
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 along with this program; if not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
SEE ALSO
RRD::Query, RRD::Threshold, rrdtool, RRDs
1 POD Error
The following errors were encountered while parsing the POD:
- Around line 408:
Non-ASCII character seen before =encoding in 'source. One'. Assuming CP1252