Using Grafana

Query and visualize AnyLog data in Grafana — increments/period/aggregations queries, blockchain visualizations, and importing sample dashboards.

This assumes Grafana is already deployed and connected to an AnyLog node as a JSON data source — see AnyLog & Grafana if you haven’t done that yet.

Grafana can display AnyLog data in two ways: Time Series (values over time) and Table (rows and columns). Queries are issued via the Additional JSON Data panel field, either using AnyLog’s two optimized query types (increments, period) or a plain SQL statement.

Grafana Page Layout

Query types reference

Field Description
type increments (default), period, info, map, aggregations
sql Custom SQL statement
details Any non-SQL AnyLog command
where Additional WHERE condition appended to the query
time_column Name of the timestamp column
value_column Name of the value column
functions List of aggregation functions to apply
include Treat additional tables as part of the queried table
extend Append node metadata to results (e.g. @table_name, @ip)
timezone utc (default) or local
time_range true/false — whether to apply the Grafana time range to the query
servers Override network-determined nodes with a specific IP:Port list
grafana.format_as timeseries or table
grafana.data_points Approximate number of data points — auto-tunes the increments interval

Increments query (time-series)

The default query type. Divides the selected time range into intervals and returns min/max/avg/count per interval.

{
  "type": "increments",
  "time_column": "timestamp",
  "value_column": "value",
  "grafana": {
    "format_as": "timeseries",
    "data_points": 1000
  }
}

Adding data_points lets AnyLog automatically calculate the optimal time interval/unit for the requested number of buckets — balancing performance, readability, and visual resolution. If omitted, Grafana’s own Interval setting is used instead. Grafana’s limit (Query Options) is also applied; if the result exceeds it, only a subset is returned.

With include and extend:

{
  "type": "increments",
  "time_column": "timestamp",
  "value_column": "value",
  "extend": ["@table_name"],
  "include": ["t98"],
  "grafana": { "format_as": "timeseries" }
}

include treats multiple tables as one logical source (querying t99 with include: ["t98"] pulls and merges data from both). extend appends source metadata to the result — @table_name groups results by their table of origin, preserving context.

With a WHERE filter:

{
  "type": "increments",
  "time_column": "timestamp",
  "value_column": "value",
  "where": "device_name='ADVA FSP3000R7'",
  "grafana": { "format_as": "timeseries" }
}

Increments graph

  1. Visualization: Time series
  2. Metric: select the table to query
  3. Payload: 
    {
  "type": "increments",
  "time_column": "timestamp",
  "value_column": "value",
  "grafana": { "format_as": "timeseries" }
}
  4. Under Query Options, set Max data points — otherwise min/max/avg collapse into what looks like a single line.

Increments Graph

Period query (latest value)

Returns the most recent value within the selected time range (or nearest to the end of it), then aggregates over a window ending at that point.

{
  "type": "period",
  "time_column": "timestamp",
  "value_column": "value",
  "grafana": { "format_as": "timeseries" }
}

Without a time range (all data, explicit functions):

{
  "type": "period",
  "time_column": "timestamp",
  "value_column": "value",
  "time_range": false,
  "functions": ["min", "max", "avg", "count"],
  "grafana": { "format_as": "timeseries" }
}

Period graph

  1. Visualization: Gauge
  2. Metric: select the table to query
  3. Payload: 
    {
  "type": "period",
  "time_column": "timestamp",
  "value_column": "value",
  "grafana": { "format_as": "timeseries" }
}
  4. Under Query Options, set Max data points — same reason as above.

Period Gauge

Aggregations query

Pulls rolling aggregations — configured via set aggregation on the AnyLog side — directly into Grafana:

{
  "servers": ["10.0.0.78:32149"],
  "type": "aggregations",
  "functions": ["min", "max", "avg", "count"],
  "table": "r_50",
  "timestamp_column": "timestamp",
  "value_column": ["filler_cyc_time", "run_hours"],
  "limit": 0
}

servers must name a single operator node for aggregations queries.

Network map (blockchain metadata)

Plot node locations on a world map.

  1. Visualization: Geomap
  2. Metric: any table (the map is populated from blockchain metadata, not table contents)
  3. Payload: 
    {
 "type" : "map",
 "member" : ["master", "query", "operator", "publisher"],
 "metric" : [0, 0, 0],
 "attribute" : ["name", "name", "name", "name"]
}

Network Map

Blockchain table

Display node metadata as a table.

  1. Visualization: Table
  2. Metric: any table
  3. Payload: 
    {
 "type": "info",
 "details": "blockchain get operator bring.json [*][cluster] [*][name] [*][company] [*][ip] [*][country] [*][state] [*][city]"
}

Blockchain Table

Importing the sample dashboards

Rather than building panels one at a time as above, AnyLog provides pre-built dashboard JSON files you can import wholesale:

  • Network Map — a map of all nodes in the network, a list of operator nodes, and a list of tables supported across the network.

    grafana_network_map.png

  • EdgeX Diagram — a line graph of min/avg/max plus gauges for total and per-node row counts, fed from the EdgeX MQTT sample connection.

    grafana_edgex_dashboard.png

Steps

  1. In a new dashboard, go to Settings:

    Empty Dashboard

  2. Go to JSON Model and paste in the desired model — the JSON object that defines the dashboard (e.g. the EdgeX Dashboard above):

    Empty JSON Model Filled JSON Model
    Empty JSON Model JSON Model

  3. Save changes.

  4. You should now see the new dashboard:

    Before After
    No Dashboards New Dashboard

  5. For each widget, update:

    • Data Source
    • Metric value (the AnyLog table name)
    View when accessing Dashboard Update Data Source Update Metric Value Outcome
    Edit Widget Update Data Source Update Metric Value Outcome

Note: the sample edgex_dashboard.json bundled with this doc set had two panels (“Total Rows - Server 1/2”) with real, non-placeholder IPs hardcoded into their query payloads. Anonymized before publishing — if you’re pulling a fresh copy of this dashboard from elsewhere, check the servers field in those two panels before sharing it further.

Exporting a dashboard

To share a dashboard you’ve built (or to save a copy of a customized sample dashboard):

  1. Open the dashboard, then go to its Settings (gear icon).
  2. Go to JSON Model.
  3. Either:
    • Copy the JSON directly from the editor, or
    • Use Export → Save to file (Grafana 9+) to download it as a .json file.

The exported file is the same format used for import above — it can be handed to someone else, checked into a repo as a versioned example (like network_summary.json / edgex_dashboard.json), or re-imported later via JSON Model on a fresh dashboard.

Before sharing an exported dashboard outside your team, check it for anything environment-specific — data source UIDs, hardcoded servers IPs in query payloads (see the note above), or table/database names — the same way you’d review any other exported config before publishing it.

Tips

  • Set Max data points in Query Options to control result density for time-series panels — without it, min/max/avg lines collapse into a single line.
  • Use format_as: timeseries for time-series panels (Time series, Gauge) and table for table panels.
  • See Querying Data for the full increments/period reference and query options like include/extend.