1 NFVbench Kibana visualization: overview
2 =======================================
4 The fluentd integration offers the possibility to use elasticsearch and kibana as a visualization chain.
8 .. image:: images/nfvbench-kibana.png
10 Example of NFVbench visualizations
11 ----------------------------------
13 Kibana offers a lot of visualization type (line and bar charts, pie, time series chart, data table ...) and also provide a plugin to develop graph using Vega.
14 In the below examples, visualizations are based on an NDR result and are developed using `Vega-lite <https://vega.github.io/vega-lite-v2>`_.
15 Data are aggregated using ``user_label`` and ``flow_count`` properties.
17 In ``kibana/visualizations/`` pre-created graphs are available into json files.
19 For NDR capacity in Gbps using line chart, the offered load in Gbps (``offered_tx_rate_bps``) is used and only the maximum value of the aggregation is kept.
20 For NDR capacity in Mpps using line chart, the actual TX rate is used (``rate_pps``) and only the maximum value of the aggregation is kept.
22 Scatter plot graphs use the same values but keep all values instead of keeping maximum.
24 Example of a line chart:
26 .. image:: images/nfvbench-kibana-gbps-line.png
28 Example of a scatter plot chart:
30 .. image:: images/nfvbench-kibana-pps-scatter.png
32 Vega offers the possibility to add another graph as a new layer of current graph.
33 This solution is used to combine NFVbench results and theoretical line rate.
34 Using ``extra_encapsulation_bytes`` in --user-info property (see `User info data section <https://opnfv-nfvbench.readthedocs.io/en/latest/testing/user/userguide/advanced.html#user-info-data>`_),
35 the theoretical max value (for bps and pps) will be calculated and can be used in graph through ``theoretical_tx_rate_bps`` and ``theoretical_tx_rate_pps`` properties.
37 Example of chart with theoretical value (red line):
39 .. image:: images/nfvbench-kibana-pps-theoretical.png
41 Each Vega graph can be moved, zoomed (using mouse scroll) and one set of data can be selected.
45 .. image:: images/nfvbench-kibana-zoom-selection.png
47 These visualizations are included into Kibana dashboard for a synthesis of one set of result (i.e. same ``user_label`` value) or for comparison (i.e. a selection of ``user_label`` values).
48 See :ref:`filterkibana` for more details about ``user_label`` selection.
50 All these visualizations and dashboards are saved into the ``export.ndjson`` file and can be imported in an existing Kibana. See :ref:`importkibana`.
54 Import Kibana dashboards and visualization
55 ------------------------------------------
57 To import Kibana dashboard and visualization:
61 curl -X POST localhost:5601/api/saved_objects/_import -H "kbn-xsrf: true" --form file=@export.ndjson
63 .. note:: ``.kibana`` index must exists in elasticsearch.
64 .. note:: ``.kibana`` index is created automatically after a first deployment and configuration of Kibana.
68 Kibana user guide: Filter dashboards and visualizations
69 =======================================================
71 Filter Kibana dashboard or visualization using Kibana query language (KQL)
72 --------------------------------------------------------------------------
76 user_label:*demo* and (flow_count: 128 or flow_count:130000 or flow_count:1000000)
78 .. note:: This query will filter all user label which contains ``demo`` in the value and filter 3 flow count (128, 130k, 1M).
79 .. note:: ``flow_count`` is a number so KQL query can not contain formatted string.
83 .. image:: images/nfvbench-kibana-filter-kql.png
86 Filter Kibana dashboard or visualization using Kibana filters
87 -------------------------------------------------------------
89 Kibana offers the possibility to add filter by selecting field and operator (is, is not, is one of, is not one of, exists, does not exist).
93 .. image:: images/nfvbench-kibana-filter.png