Tip & Tricks for Snappy Dashboards
- 1.Create always light data single dashboards or group your visualizations into buckets of dashboards
Kibana sends all queries related to Visualizations in a Dashboard in a single
_msearchrequest, which then executes in parallel.
When a dashboard has multiple visualizations, requests are sent individually, causing some visualizations which are light on data to be rendered a bit quicker. If there are visualizations that consume large amounts of data, the entire dashboard as a whole will be rendered more slowly (from start to finish).
Thus either you can opt for a single dashboard with a few visualizations or use a group of visuals to create a bucket of dashboards to experience snappy performance on Kibana
2. Using panel filters
Most of the event sources publish more than one type of log or event, including access logs and error logs. Panel filters allow the user to reduce the data set loaded for visualization and pinpoint the exact data set where the data is to be utilized for building the visualizations in Kibana.
Using panel filters reduces the overall time for building the visualizations. Panel filters can be accessed from the Panel options while editing the TSVB visualizations under the “Panel option” as shown below
Panel Filter reduces size of dataset for snappy visualization
3. Use dataset filters in place of filters based on event module and metricset name
Under the panel filter option, you can directly pinpoint the event dataset instead of specifying separately the event module and the metric set name.
For example, if you are using visualization to monitor the CPU status of a docker. You can specify the data set in the following manner:
- 1.event.module:docker and metricset.name:cpu
- 2.event.dataset: docker.cpu
Using option 2 reduces time to point to the particular module compared to filtering based on the metricset name, which means that Kibana will instead be directly point to the event dataset of docker.cpu