SQL Query Management

Manage Top Applications Breakdown Graph Query

Endpoints

• Execute query to retrieve data from the database. The response is unique UUID which supposed to be used for requesting results of the query.

  • Path: /api/chf/graph/top_applications/breakdown

  • Method: POST

  • Request body:

    {
      [optional]  "filter_clause": "<filter bar expression> (e.g. application = 'HTTP' AND update_total_bytes > 10000000000)",
      [mandatory] "selected_columns": [
        "update_total_bytes"
      ],
      [mandatory] "entries_num": 300,
      [mandatory] "top_n_per_bucket": 10,
      [mandatory] "order_by": "update_total_bytes",
      [mandatory] "sort_order": "desc",
      [optional]  "from_timestamp": 1778001660,
      [optional]  "to_timestamp": 1778088060
    }

  • Response:

    {
      "message": "Success",
      "data": {
        "uuid": "c8144dc6-8aa0-ae89-1550-542cff5c3bab"
      }
    }

  • filter_clause is the part which is inserted after a WHERE keyword of an SQL query and qualifies the set of applications shown on the graph. If not provided or null, no filter is applied.

  • selected_columns are the metric columns aggregated per bucket. Only columns with is_chartable = true are accepted. The breakdown entity (application) is always returned and must not be listed here.

  • entries_num is the target number of time buckets, must not be 0.

  • top_n_per_bucket is the number of top entities returned per bucket, must not be 0. Everything outside the top-N is aggregated into the other field.

  • order_by is the chartable column used to rank entities within each bucket, must be present in selected_columns.

  • sort_order is desc or asc.

  • from_timestamp and to_timestamp are the range of requested data. If not provided or null, the minimum/maximum value is used.

• Get next page by UUID.

  • Path: /api/chf/graph/top_applications/breakdown/{uuid}

  • Method: GET

  • Path parameter:

    • uuid - received from the previous request

  • Response:

    {
      "message": "Success",
      "data": {
        "state": "no_page_no_processing",
        "custom_data": {
          "interval_sec": 300
        },
        "page": [
          {
            "timestamp": 1778001660,
            "result": [
              {
                "application": "HTTP",
                "update_total_bytes": 1234567890
              }
            ],
            "other": {
              "update_total_bytes": 98765432
            }
          }
        ]
      }
    }

  • state one of:

    • has_next - there is the next page, it can be asked for immediately
    • no_page_processing - there is no next page yet, but the query it still being processed, maybe it will appear later; sleep for a bit and try again
    • no_page_no_processing - the query processing has finished and there’s no next page; stop

  • custom_data - additional information that needs to be included with the result:

    • interval_sec - interval in seconds between timestamps in the page results.

  • page is a list of time buckets. Each bucket has:

    • timestamp - bucket start.
    • result - up to top_n_per_bucket top entities of that bucket, ranked by order_by. Empty for buckets with no data.
    • other - aggregate of all entities outside the top-N. Only additive metrics (bytes/packets) are present; non-summable metrics (avg/min/max/unique counts) are omitted. Absent for empty buckets.
• Interrupt the query.

  • Path: /api/chf/graph/top_applications/breakdown/{uuid}

  • Method: DELETE

  • Path parameter:

    • uuid - received from the previous request

  • Response:

    {
      "message": "Successfully interrupted processing the query (5f966c56-6acb-d5c5-7832-eb29b000f333)",
      "data": null
    }

Columns

selected_columns and order_by accept the same chartable columns as the Top Applications graph. This endpoint has no /columns route; fetch column metadata from /graph/top_applications/columns (see Top Applications Graph for the full list and properties).

Back to Clickhouse Facade