Hdridder at 16:37, 17 January 2021

2021-01-17T16:37:27Z

← Older revision		Revision as of 16:37, 17 January 2021
Line 3:		Line 3:
	In the example below:		In the example below:
	;start		;start
	:First date in the database in epoch (seconds since January 1, 1970). date +%s gives current epoch.		:First date in the database in [https://www.epochconverter.com/ epoch] (seconds since January 1, 1970). date +%s gives current epoch.
	;step		;step
	:Expected rate data comes in in seconds		:Expected rate data comes in in seconds

Hdridder: Created page with "Round Robin Databases by Toby Oetker [https://oss.oetiker.ch/rrdtool/] In the example below: ;start :First date in the database in epoch (seconds since January 1, 1970). date..."

2018-02-05T21:24:33Z

Created page with "Round Robin Databases by Toby Oetker [https://oss.oetiker.ch/rrdtool/] In the example below: ;start :First date in the database in epoch (seconds since January 1, 1970). date..."

New page

Round Robin Databases by Toby Oetker [https://oss.oetiker.ch/rrdtool/]

In the example below:
;start
:First date in the database in epoch (seconds since January 1, 1970). date +%s gives current epoch.
;step
:Expected rate data comes in in seconds
;<nowiki>DS:name:type:heartbeat:min:max</nowiki>
:Datasource, the raw data to store
:type; Most used are GAUGE (set of data) or COUNTER (always increasing data like from a smartmeter)
:heartbeat; Maximum time between two enty's. After this time a record with *UNKNOWN* values will be put in.
:min, max; Values outside this range will be recorded as *UNKNOWN*
;<nowiki>RRA:consolidatefunction:missing:steps:rows</nowiki>
:Round Robin Archive, how the raw data will be consolidated
:consolidatefunction; <nowiki>[AVERAGE,MIN,MAX,LAST]</nowiki> what to take from the raw data to store
:missing; The amount of allowed *UNKNOWN* records that still give a valid result (0,5 means half the data may be missing)
:steps; The number of raw data rows to consolidate
:rows; The amount of records to keep in the RRA
:What timespan is saved depends on the stepsize, the number of steps and the number of rows.

Example:
<syntaxhighlight lang=bash>
rrdtool create database.rrd \
--start 1 \
--step 10 \
DS:count1:COUNTER:60:0:999999999 \
DS:count2:COUNTER:60:0:999999999 \
DS:gauge1:GAUGE:60:0:1 \
DS:gauge2:GAUGE:60:0:99999 \
DS:count3:COUNTER:60:0:99999999 \
RRA:AVERAGE:0.5:1:60480 \ # One week of 10 seconds data
RRA:AVERAGE:0.5:6:44640 \ # 31 days of 1 minute data (consolidate 6 10sec records)
RRA:AVERAGE:0.5:8640:365 # 365 days of 1 day data (consolidate 8640 10sec records)
</syntaxhighlight>

RRD's are used to generate graphs.
;start
:At what time the X-axis starts
;<nowiki>DEF:graphname=database.rrd:counter:consolidate</nowiki>
:What DS to display
:graphname; Name to refer to in other graph commands (like AREA)
:database.rrd:counter; DS to display
:consolidate; How to handle data if it needs to be consolidated before displaying it.
;<nowiki>type:graphname#008000:legend:STACK</nowiki>
:How to display the graph
:type; Common are LINE, AREA
:graphname#000000; Counter and color to display it in
:STACK; For subsequent AREA's, stack them instead of displaying over eachother.

Example:
<syntaxhighlight lang=bash>
rrdtool graph imagename.png --title "This is a nice graph" \
--width 800 \
--vertical-label aLabel \
--x-grid HOUR:1:HOUR:4:HOUR:1:0:%H \
--start now-1d \
--upper-limit 1 -r\
DEF:graphname1=database.rrd:counter1:AVERAGE \
DEF:graphname2=database.rrd:counter2:AVERAGE \
AREA:graphname1#FF0000:legend1 \
AREA:graphname2#008000:legend2:STACK
</syntaxhighlight>

RRD - Revision history

Hdridder at 16:37, 17 January 2021

Hdridder: Created page with "Round Robin Databases by Toby Oetker [https://oss.oetiker.ch/rrdtool/] In the example below: ;start :First date in the database in epoch (seconds since January 1, 1970). date..."