Chunked HTTP Responses

Warning! This page documents an old version of InfluxDB, which is no longer actively developed. InfluxDB v1.2 is the most recent stable version of InfluxDB.

If the request asks for a chunked response, JSON objects will get written to the HTTP response as they are ready. They will come in batches in the requested time order. That might look like this:

{
  "name": "a_series",
  "columns": ["time", "sequence_number", "column_a"],
  "points": [
    [1383059590062, 3, 27.3],
    [1383059590062, 4, 97.1]
  ]
}

Then followed by

{
  "name": "b_series",
  "columns": ["time", "sequence_number", "column_a"],
  "points": [
    [1383059590062, 2, 2232.1]
  ]
}

Then followed by

{
  "name": "a_series",
  "columns": ["time", "sequence_number", "column_a"],
  "points": [
    [1383059590000, 1, 291.7],
    [1383059590000, 2, 44.1]
  ]
}

The chunks for different series can be interleaved, but they will always come back in the correct time order. You should use chunked queries when pulling back a large number of data points. If you’re just pulling back data for a graph, which should generally have fewer than a few thousand points, non-chunked responses are easiest to work with.