1RRDTUNE(1) rrdtool RRDTUNE(1)
2
6 rrdtune - Modify some basic properties of a Round Robin Database
7
9 rrdtool tune filename [--heartbeat|-h ds-name:heartbeat]
10 [--minimum|-i ds-name:min] [--maximum|-a ds-name:max]
11 [--data-source-type|-d ds-name:DST] [--data-source-rename|-r old-
12 name:new-name] [--deltapos scale-value] [--deltaneg scale-value]
13 [--failure-threshold failure-threshold] [--window-length window-length]
14 [--alpha adaption-parameter] [--beta adaption-parameter]
15 [--gamma adaption-parameter] [--gamma-deviation adaption-parameter]
16 [--smoothing-window fraction-of-season]
17 [--smoothing-window-deviation fraction-of-season] [--aberrant-reset ds-
18 name]
19
21 The tune option allows you to alter some of the basic configuration
22 values stored in the header area of a Round Robin Database (RRD).
23 One application of the tune function is to relax the validation rules
25 on an RRD. This allows to fill a new RRD with data available in larger
26 intervals than what you would normally want to permit. Be very careful
27 with tune operations for COMPUTE data sources. Setting the min, max,
28 and heartbeat for a COMPUTE data source without changing the data
29 source type to a non-COMPUTE DST WILL corrupt the data source header in
30 the RRD.
31 A second application of the tune function is to set or alter parameters
33 used by the specialized function RRAs for aberrant behavior detection.
34 filename
36 The name of the RRD you want to tune.
37 --heartbeat|-h ds-name:heartbeat
39 modify the heartbeat of a data source. By setting this to a
40 high value the RRD will accept things like one value per day.
41 --minimum|-i ds-name:min
43 alter the minimum value acceptable as input from the data
44 source. Setting min to 'U' will disable this limit.
45 --maximum|-a ds-name:max
47 alter the maximum value acceptable as input from the data
48 source. Setting max to 'U' will disable this limit.
49 --data-source-type|-d ds-name:DST
51 alter the type DST of a data source.
52 --data-source-rename|-r old-name:new-name
54 rename a data source.
55 --deltapos scale-value
57 Alter the deviation scaling factor for the upper bound of the
58 confidence band used internally to calculate violations for the
59 FAILURES RRA. The default value is 2. Note that this parameter
60 is not related to graphing confidence bounds which must be
61 specified as a CDEF argument to generate a graph with
62 confidence bounds. The graph scale factor need not to agree
63 with the value used internally by the FAILURES RRA.
64 --deltaneg scale-value
66 Alter the deviation scaling factor for the lower bound of the
67 confidence band used internally to calculate violations for the
68 FAILURES RRA. The default value is 2. As with --deltapos, this
69 argument is unrelated to the scale factor chosen when graphing
70 confidence bounds.
71 --failure-threshold failure-threshold
73 Alter the number of confidence bound violations that constitute
74 a failure for purposes of the FAILURES RRA. This must be an
75 integer less than or equal to the window length of the FAILURES
76 RRA. This restriction is not verified by the tune option, so
77 one can reset failure-threshold and window-length
78 simultaneously. Setting this option will reset the count of
79 violations to 0.
80 --window-length window-length
82 Alter the number of time points in the temporal window for
83 determining failures. This must be an integer greater than or
84 equal to the window length of the FAILURES RRA and less than or
85 equal to 28. Setting this option will reset the count of
86 violations to 0.
87 --alpha adaption-parameter
89 Alter the intercept adaptation parameter for the Holt-Winters
90 forecasting algorithm. This parameter must be between 0 and 1.
91 --beta adaption-parameter
93 Alter the slope adaptation parameter for the Holt-Winters
94 forecasting algorithm. This parameter must be between 0 and 1.
95 --gamma adaption-parameter
97 Alter the seasonal coefficient adaptation parameter for the
98 SEASONAL RRA. This parameter must be between 0 and 1.
99 --gamma-deviation adaption-parameter
101 Alter the seasonal deviation adaptation parameter for the
102 DEVSEASONAL RRA. This parameter must be between 0 and 1.
103 --smoothing-window fraction-of-season
105 Alter the size of the smoothing window for the SEASONAL RRA.
106 This must be between 0 and 1.
107 --smoothing-window-deviation fraction-of-season
109 Alter the size of the smoothing window for the DEVSEASONAL RRA.
110 This must be between 0 and 1.
111 --aberrant-reset ds-name
113 This option causes the aberrant behavior detection algorithm to
114 reset for the specified data source; that is, forget all it is
115 has learnt so far. Specifically, for the HWPREDICT or
116 MHWPREDICT RRA, it sets the intercept and slope coefficients to
117 unknown. For the SEASONAL RRA, it sets all seasonal
118 coefficients to unknown. For the DEVSEASONAL RRA, it sets all
119 seasonal deviation coefficients to unknown. For the FAILURES
120 RRA, it erases the violation history. Note that reset does not
121 erase past predictions (the values of the HWPREDICT or
122 MHWPREDICT RRA), predicted deviations (the values of the
123 DEVPREDICT RRA), or failure history (the values of the FAILURES
124 RRA). This option will function even if not all the listed
125 RRAs are present.
126 Due to the implementation of this option, there is an indirect
128 impact on other data sources in the RRD. A smoothing algorithm
129 is applied to SEASONAL and DEVSEASONAL values on a periodic
130 basis. During bootstrap initialization this smoothing is
131 deferred. For efficiency, the implementation of smoothing is
132 not data source specific. This means that utilizing reset for
133 one data source will delay running the smoothing algorithm for
134 all data sources in the file. This is unlikely to have serious
135 consequences, unless the data being collected for the non-reset
136 data sources is unusually volatile during the reinitialization
137 period of the reset data source.
138 Use of this tuning option is advised when the behavior of the
140 data source time series changes in a drastic and permanent
141 manner.
142
144 "rrdtool tune data.rrd -h in:100000 -h out:100000 -h through:100000"
145 Set the minimum required heartbeat for data sources 'in', 'out' and
147 'through' to 10'000 seconds which is a little over one day in data.rrd.
148 This would allow to feed old data from MRTG-2.0 right into RRDtool
149 without generating *UNKNOWN* entries.
150
152 "rrdtool tune monitor.rrd --window-length 5 --failure-threshold 3"
153 If the FAILURES RRA is implicitly created, the default window-length is
155 9 and the default failure-threshold is 7. This command now defines a
156 failure as 3 or more violations in a temporal window of 5 time points.
157
159 Tobias Oetiker <tobi@oetiker.ch>
1601.3.8 2008-03-15 RRDTUNE(1)