Page History

...

            Note: the "state" could be "pending", "started", "downloaded", "parsed", "persisted", "completed"completed, "failed"

Pre-defined format for import files

...

{
  "rows":[
    {"uid": 101, "vector": [1.1, 1.2, 1.3, 1.4]},
    {"uid": 102, "vector": [2.1, 2.2, 2.3, 2.4]},
    {"uid": 103, "vector": [3.1, 3.2, 3.3, 3.4]},
    {"uid": 104, "vector": [4.1, 4.2, 4.3, 4.4]},
    {"uid": 105, "vector": [5.1, 5.2, 5.3, 5.4]},
  ]
}

Call import() to import the file:

...

import(collection_name="test", row_based=false, files=["file_1.json", "vector.npy"])

Note: for column-based, we don't support multiple json files, all columns should be stored in one json file. If user use a numpy file to store vector field, then other scalar fileds should be stored in one json file.

Error handling

The Import():

...

For column-based request, all files will be regarded as one ImportTask.

Image RemovedImage Added

5. Datanode interfaces

...