druid/docs/development/extensions-core/parquet.md

223 lines
7.3 KiB
Markdown

---
id: parquet
title: "Apache Parquet Extension"
---
<!--
~ Licensed to the Apache Software Foundation (ASF) under one
~ or more contributor license agreements. See the NOTICE file
~ distributed with this work for additional information
~ regarding copyright ownership. The ASF licenses this file
~ to you under the Apache License, Version 2.0 (the
~ "License"); you may not use this file except in compliance
~ with the License. You may obtain a copy of the License at
~
~ http://www.apache.org/licenses/LICENSE-2.0
~
~ Unless required by applicable law or agreed to in writing,
~ software distributed under the License is distributed on an
~ "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
~ KIND, either express or implied. See the License for the
~ specific language governing permissions and limitations
~ under the License.
-->
This Apache Druid (incubating) module extends [Druid Hadoop based indexing](../../ingestion/hadoop.md) to ingest data directly from offline
Apache Parquet files.
Note: `druid-parquet-extensions` depends on the `druid-avro-extensions` module, so be sure to
[include both](../../development/extensions.md#loading-extensions).
## Parquet Hadoop Parser
This extension provides two ways to parse Parquet files:
* `parquet` - using a simple conversion contained within this extension
* `parquet-avro` - conversion to avro records with the `parquet-avro` library and using the `druid-avro-extensions`
module to parse the avro data
Selection of conversion method is controlled by parser type, and the correct hadoop input format must also be set in
the `ioConfig`:
* `org.apache.druid.data.input.parquet.DruidParquetInputFormat` for `parquet`
* `org.apache.druid.data.input.parquet.DruidParquetAvroInputFormat` for `parquet-avro`
Both parse options support auto field discovery and flattening if provided with a
[`flattenSpec`](../../ingestion/index.md#flattenspec) with `parquet` or `avro` as the format. Parquet nested list and map
[logical types](https://github.com/apache/parquet-format/blob/master/LogicalTypes.md) _should_ operate correctly with
JSON path expressions for all supported types. `parquet-avro` sets a hadoop job property
`parquet.avro.add-list-element-records` to `false` (which normally defaults to `true`), in order to 'unwrap' primitive
list elements into multi-value dimensions.
The `parquet` parser supports `int96` Parquet values, while `parquet-avro` does not. There may also be some subtle
differences in the behavior of JSON path expression evaluation of `flattenSpec`.
We suggest using `parquet` over `parquet-avro` to allow ingesting data beyond the schema constraints of Avro conversion.
However, `parquet-avro` was the original basis for this extension, and as such it is a bit more mature.
|Field | Type | Description | Required|
|----------|-------------|----------------------------------------------------------------------------------------|---------|
| type | String | Choose `parquet` or `parquet-avro` to determine how Parquet files are parsed | yes |
| parseSpec | JSON Object | Specifies the timestamp and dimensions of the data, and optionally, a flatten spec. Valid parseSpec formats are `timeAndDims`, `parquet`, `avro` (if used with avro conversion). | yes |
| binaryAsString | Boolean | Specifies if the bytes parquet column which is not logically marked as a string or enum type should be converted to strings anyway. | no(default == false) |
When the time dimension is a [DateType column](https://github.com/apache/parquet-format/blob/master/LogicalTypes.md), a format should not be supplied. When the format is UTF8 (String), either `auto` or a explicitly defined [format](http://www.joda.org/joda-time/apidocs/org/joda/time/format/DateTimeFormat.html) is required.
### Examples
#### `parquet` parser, `parquet` parseSpec
```json
{
"type": "index_hadoop",
"spec": {
"ioConfig": {
"type": "hadoop",
"inputSpec": {
"type": "static",
"inputFormat": "org.apache.druid.data.input.parquet.DruidParquetInputFormat",
"paths": "path/to/file.parquet"
},
...
},
"dataSchema": {
"dataSource": "example",
"parser": {
"type": "parquet",
"parseSpec": {
"format": "parquet",
"flattenSpec": {
"useFieldDiscovery": true,
"fields": [
{
"type": "path",
"name": "nestedDim",
"expr": "$.nestedData.dim1"
},
{
"type": "path",
"name": "listDimFirstItem",
"expr": "$.listDim[1]"
}
]
},
"timestampSpec": {
"column": "timestamp",
"format": "auto"
},
"dimensionsSpec": {
"dimensions": [],
"dimensionExclusions": [],
"spatialDimensions": []
}
}
},
...
},
"tuningConfig": <hadoop-tuning-config>
}
}
}
```
#### `parquet` parser, `timeAndDims` parseSpec
```json
{
"type": "index_hadoop",
"spec": {
"ioConfig": {
"type": "hadoop",
"inputSpec": {
"type": "static",
"inputFormat": "org.apache.druid.data.input.parquet.DruidParquetInputFormat",
"paths": "path/to/file.parquet"
},
...
},
"dataSchema": {
"dataSource": "example",
"parser": {
"type": "parquet",
"parseSpec": {
"format": "timeAndDims",
"timestampSpec": {
"column": "timestamp",
"format": "auto"
},
"dimensionsSpec": {
"dimensions": [
"dim1",
"dim2",
"dim3",
"listDim"
],
"dimensionExclusions": [],
"spatialDimensions": []
}
}
},
...
},
"tuningConfig": <hadoop-tuning-config>
}
}
```
#### `parquet-avro` parser, `avro` parseSpec
```json
{
"type": "index_hadoop",
"spec": {
"ioConfig": {
"type": "hadoop",
"inputSpec": {
"type": "static",
"inputFormat": "org.apache.druid.data.input.parquet.DruidParquetAvroInputFormat",
"paths": "path/to/file.parquet"
},
...
},
"dataSchema": {
"dataSource": "example",
"parser": {
"type": "parquet-avro",
"parseSpec": {
"format": "avro",
"flattenSpec": {
"useFieldDiscovery": true,
"fields": [
{
"type": "path",
"name": "nestedDim",
"expr": "$.nestedData.dim1"
},
{
"type": "path",
"name": "listDimFirstItem",
"expr": "$.listDim[1]"
}
]
},
"timestampSpec": {
"column": "timestamp",
"format": "auto"
},
"dimensionsSpec": {
"dimensions": [],
"dimensionExclusions": [],
"spatialDimensions": []
}
}
},
...
},
"tuningConfig": <hadoop-tuning-config>
}
}
}
```
For additional details see [Hadoop ingestion](../../ingestion/hadoop.md) and [general ingestion spec](../../ingestion/index.md) documentation.