Signals APIFusion Job REST Server APIs

Table of Contents

Examples

The Signals API accepts a set of signals, encoded as JSON objects, for indexing into a signals collection.

For more information, view the API specification.

Normally, signals are indexed just like ordinary documents, through a configured datasource and index pipeline. This API is provided for cases where it is more convenient to index signals directly.

To aggregate signals, use a SQL aggregation job, which is a kind of Spark job.

You can manage aggregation jobs in the Jobs manager and Scheduler in the Admin UI, or with the Spark jobs API. We recommend using the Admin UI.

See Signal types and structure to learn how to send well-formed signals to this API.

Examples

Send two signal events to record user clicks:

REQUEST

This example uses the commit=true setting, which commits after every signal. The setting is not recommended in production and performance testing because it negatively impacts performance. Instead of sending the commit parameter, consider using no commit parameter and let the defaults of the Solr collection dictate the commit frequency. Alternatively, send the commitWithin parameter when shorter read visibility is needed, for example, commitWithin=3000.

curl \
 -u USERNAME:PASSWORD -X POST -H 'Content-type:application/json' -d @- \
https://FUSION_HOST:FUSION_PORT/api/apps/APP_NAME/signals/COLLECTION_NAME?async=false&commit=true \
<<EOF
[
 {"params": {
      "query": "Televisiones Panasonic  50 pulgadas",
      "filterQueries": ["cat00000","abcat0100000", "abcat0101000", "abcat0101001"],
      "docId": "2125233" },
 "type":"click",
 "timestamp": "2011-09-01T23:44:52.53Z"
 },
 {"params": {
       "query": "Sharp",
       "filterQueries": ["cat00000", "abcat0100000", "abcat0101000", "abcat0101001"],
       "docId": "2009324" },
  "type":"click",
  "timestamp": "2011-09-05T12:25:37.42Z"
 }
]
EOF

A successful request results in events being added to the signals collection. For the above example, the events will be represented as follows:

 {
  "responseHeader":{
    "status":0,
    "QTime":1,
    "params":{
      "indent":"true",
      "q":"docId: 2125233",
      "wt":"json"}
 },
  "response":{"numFound":1198,"start":0,"docs":[
      {
        "id": "7aee7b1f-5cde-4957-b73c-c15881f559ec",
        "filters_s": "abcat0100000 $ abcat0101000 $ abcat0101001 $ cat00000",
        "query_orig_s": "Televisiones Panasonic  50 pulgadas",
        "params.user_s": "000000df17cd56a5df4a94074e133c9d4739fae3",
        "docId": "2125233",
        "params.query_time__s": "2011-09-01T23:43:59.75Z",
        "query_t": "televisiones panasonic 50 pulgadas",
        "query_s": "televisiones panasonic 50 pulgadas",
        "filters_orig_ss": [
          "abcat0100000",
          "abcat0101000",
          "abcat0101001",
          "cat00000"
        ],
        "flag_s": "EVENT",
        "type_s": "click",
        "attr_params.filterQueries_": [
          "cat00000",
          "abcat0100000",
          "abcat0101000",
          "abcat0101001"
        ],
        "timestamp_dt": "2011-09-01T23:44:52.53Z",
        "_version_": 1478892846857584600
      },
      {
        "id": "6789a209-f5b5-457e-9df6-8033b8f7f317",
        "filters_s": "abcat0100000 $ abcat0101000 $ abcat0101001 $ cat00000",
        "query_orig_s": "Sharp",
        "params.user_s": "000001928162247ffaf63185cd8b2a244c78e7c6",
        "docId": "2009324",
        "params.query_time__s": "2011-09-05T12:25:01.18Z",
        "query_t": "sharp",
        "query_s": "sharp",
        "filters_orig_ss": [
          "abcat0100000",
          "abcat0101000",
          "abcat0101001",
          "cat00000"
        ],
        "flag_s": "EVENT",
        "type_s": "click",
        "attr_params.filterQueries_": [
          "cat00000",
          "abcat0100000",
          "abcat0101000",
          "abcat0101001"
        ],
        "timestamp_dt": "2011-09-05T12:25:37.42Z",
        "_version_": 1478892846859681800
      }
    ]
  }
}