RRDtool

1
RRDtool ??

• ???

2
Background

• RRD is the Acronym for Round Robin Database. RRD
  is a system to store and display time-series data
  . You can think of RRDtool as a reimplementation
  of MRTGs graphing and logging features.
• It stores the data in a very compact way that
  will not expand over time, and it presents useful
  graphs by processing the data.
• Developed by Tobias Oetiker ltoetiker_at_ee.ethz.chgt .

3
Basic ideas

• Round robin works with a fixed amount of data,
  and a pointer to the current element. Think of a
  circle with some dots plotted on the edge, these
  dots are the places where data can be stored.
  Draw an arrow from the center of the circle to
  one of the dots, this is the pointer. When the
  current data is read or written, the pointer
  moves to the next element. As we are on a circle
  there is no beginning nor an end, you can go on
  and on. After a while, all the available places
  will be used and the process automatically reuses
  old locations.

4
Functions(1)

• create Set up a new Round Robin Database (RRD).
• Update Store new data values into an RRD.
• Graph Create a graph from data stored in one or
  several RRD.
• Dump Dump the contents of an RRD in plain ASCII.
• Restore Restore an RRD in XML format to a binary
  rrd

5
Functions(2)

• Fetch Get data for a certain time period from a
  RRD. The graph function uses fetch to retrieve
  its data from an rrd.
• Tune Alter setup of an RRD.
• Last Find last update time of an RRD.
• Resize Change the size of individual RRAs
• Cgi This is a standalone tool for producing rrd
  graphs on the fly.

6
Look into rrd file format

• RRD file format is defined in rrd_format.h.
• Just as you might guess, a rrd file include a
  header section, followed by DATA STORAGE AREA
  which is made up from Consolidated Data Points
  organized in Round Robin Archives. In the
  following format.

7
RRD file heads structure(1)

• To properly describe the database structure lets
  define a few new words
• ds - Data Source (ds) providing input to the
  database. A Data Source (ds) can be a traffic
  counter, a temperature, the number of users
  logged into a system. The rrd database format can
  handle the input of several Data Sources (ds) in
  a singe database.
• dst - Data Source Type (dst). The Data Source
  Type (dst) defines the rules applied to Build
  Primary Data Points from the input provided by
  the data sources (ds).

8
RRD file heads structure(2)

• pdp - Primary Data Point (pdp). After the
  database has accepted the input from the data
  sources (ds). It starts building Primary Data
  Points (pdp) from the data. Primary Data Points
  (pdp) are evenly spaced along the time axis
  (pdp_step). The values of the Primary Data Points
  are calculated from the values of the data source
  (ds) and the exact time these values were
  provided by the data source (ds).
• pdp_st - PDP Start (pdp_st). The moments (pdp_st)
  in time where these steps occur are defined by
  the moments where the number of seconds since
  1970-jan-1 modulo pdp_step equals zero (pdp_st).

9
RRD file heads structure(3)

• cf - Consolidation Function (cf). An arbitrary
  Consolidation Function (cf) (averaging, min,
  max) is applied to the primary data points (pdp)
  to calculate the consolidated data point.
• cdp - Consolidated Data Point (cdp) is the long
  term storage format for data in the rrd database.
  Consolidated Data Points represent one or several
  primary data points collected along the time
  axis. The Consolidated Data Points (cdp) are
  stored in Round Robin Archives (rra).

10
RRD file heads structure(4)

• rra - Round Robin Archive (rra). This is the
  place where cdp get stored. The data is organized
  in rows and columns. The Round Robin Archive got
  its name from the method data is stored in there.
  An RRD database can contain several Round Robin
  Archives. Each Round Robin Archive can have a
  different row spacing along the time axis
  (pdp_cnt) and a different consolidation function
  (cf) used to build its consolidated data points
  (cdp).
• rra_st - RRA Start (rra_st). The moments (rra_st)
  in time where Consolidated Data Points (cdp) are
  added to an rra are defined by the moments where
  the number of seconds since1970-jan-1 modulo
  pdp_cntpdp_step equals zero (rra_st).
• row - Row (row). A row represent all
  consolidated data points (cdp) in a round robin
  archive who are of the same age.
• col - Column (col). A column (col) represent all
  consolidated data points (cdp) in a round robin
  archive (rra) who originated from the same data
  source (ds).