如何查询分区的 AWS Athena table

How to query on AWS Athena partitioned table

问题总结

When I attempts to SELECT query the partitioned table with WHERE clause, Athena produce an error.

有4种分区,在我的log table.

我在分区 table 上尝试 SELECT 查询。 但是收到以下错误消息。

错误信息

GENERIC_INTERNAL_ERROR: No value present
This query ran against the "default" database, unless qualified by the query.

SELECT 我试过的查询

SELECT *
FROM logs
WHERE year='2020'
  AND month='10'
  AND day ='05';

SELECT *
FROM "default"."logs"
WHERE year='2020'
  AND month='10'
  AND day ='05';

由于有关 No value present 的错误消息,我检查了分区结果。

SHOW PARTITIONS logs;

结果

year=2020/month=10/day=05/hour=17
year=2020/month=10/day=05/hour=11
year=2020/month=10/day=05/hour=19
year=2020/month=10/day=05/hour=04
year=2020/month=10/day=05/hour=18
year=2020/month=10/day=05/hour=15
year=2020/month=10/day=05/hour=14
year=2020/month=10/day=05/hour=16
year=2020/month=10/day=05/hour=13
year=2020/month=10/day=05/hour=21
year=2020/month=10/day=05/hour=05
year=2020/month=10/day=05/hour=08
year=2020/month=10/day=05/hour=20
year=2020/month=10/day=05/hour=12
year=2020/month=10/day=05/hour=03
year=2020/month=10/day=05/hour=01
year=2020/month=10/day=05/hour=10
year=2020/month=10/day=05/hour=02
year=2020/month=10/day=05/hour=09
year=2020/month=10/day=05/hour=22
year=2020/month=10/day=05/hour=23
year=2020/month=10/day=05/hour=06
year=2020/month=10/day=05/hour=07
year=2020/month=10/day=05/hour=00
year=2020/month=10/day=04/hour=00

非常感谢你的帮助。


更多信息

CREATE TABLE 我使用的命令

创建Table

CREATE EXTERNAL TABLE `logs`(
  `date` date, 
  `time` string, 
  `location` string, 
  `bytes` bigint, 
  `request_ip` string, 
  `method` string, 
  `host` string, 
  `uri` string, 
  `status` int, 
  `referrer` string, 
  `user_agent` string, 
  `query_string` string, 
  `cookie` string, 
  `result_type` string, 
  `request_id` string, 
  `host_header` string, 
  `request_protocol` string, 
  `request_bytes` bigint, 
  `time_taken` float, 
  `xforwarded_for` string, 
  `ssl_protocol` string, 
  `ssl_cipher` string, 
  `response_result_type` string, 
  `http_version` string, 
  `fle_status` string, 
  `fle_encrypted_fields` int)
PARTITIONED BY ( 
  `year` string, 
  `month` string, 
  `day` string, 
  `hour` string)
ROW FORMAT DELIMITED 
  FIELDS TERMINATED BY '\t' 
WITH SERDEPROPERTIES ( 
  'input.regex'='^(?!#)([^ \\t]+)\\s+([^ \\t]+)\\s+([^ \\t]+)\\s+([^ \\t]+)\\s+([^ \\t]+)\\s+([^ \\t]+)\\s+([^ \\t]+)\\s+([^ \\t]+)\\s+([^ \\t]+)\\s+([^ \\t]+)\\s+([^ \\t]+)\\s+([^ \\t]+)\\s+([^ \\t]+)\\s+([^ \\t]+)\\s+([^ \\t]+)\\s+([^ \\t]+)\\s+([^ \\t]+)\\s+([^ \\t]+)\\s+([^ \\t]+)\\s+([^ \\t]+)\\s+([^ \\t]+)\\s+([^ \\t]+)\\s+([^ \\t]+)\\s+([^ \\t]+)\\s+([^ \\t]+)\\s+([^ \\t]+)$') 
STORED AS INPUTFORMAT 
  'org.apache.hadoop.mapred.TextInputFormat' 
OUTPUTFORMAT 
  'org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat'
LOCATION
  's3://mybucket/path'
TBLPROPERTIES (
  'projection.date.format'='yyyy/MM/dd', 
  'projection.date.interval'='1', 
  'projection.date.interval.unit'='DAYS', 
  'projection.date.range'='2019/11/27, NOW-1DAYS', 
  'projection.date.type'='date', 
  'projection.day.type'='string', 
  'projection.enabled'='true', 
  'projection.hour.type'='string', 
  'projection.month.type'='string', 
  'projection.year.type'='string', 
  'skip.header.line.count'='2', 
  'storage.location.template'='s3://mybucket/path/distributionID/${year}/${month}/${day}/${hour}/', 
  'transient_lastDdlTime'='1575005094')

您的 table 使用分区投影,但您的配置与分区不匹配。分区投影是一个相当新的功能,文档仍然有一些不足之处,所以我完全理解它令人困惑。我想我明白你想做什么了。

分区投影配置必须与 table 的分区键完全匹配。在您的情况下,table 有四个分区键,分区投影配置提到了五个。除了四个的类型不对之外,没有string分区投影类型

您可以通过进行两处更改来解决问题。首先像这样更改分区键:

PARTITIONED BY (
  `date` string,
  `hour` string
)

这会删除“年”、“月”和“日”分区键,取而代之的是“日期”键。仅仅因为它们是单独的“目录”就拥有单独的日期组件是没有必要的,仅仅拥有一个“日期”键将使查询更容易编写。

然后将 table 属性更改为:

TBLPROPERTIES (
  'projection.date.format' = 'yyyy/MM/dd', 
  'projection.date.interval' = '1', 
  'projection.date.interval.unit' = 'DAYS', 
  'projection.date.range' = '2019/11/27, NOW-1DAYS', 
  'projection.date.type' = 'date', 
  'projection.hour.type' = 'integer',
  'projection.hour.range' = '0-23',
  'projection.hour.digits' = '2'
  'projection.enabled' = 'true',
  'storage.location.template'='s3://mybucket/path/distributionID/${date}/${hour}/', 
  'skip.header.line.count' = '2'
)

这告诉 Athena,“日期”分区键的类型为 date,并且它的格式为“YYYY/MM/DD”(对应于 S3 URI 中的格式,这很重要).它还告诉 Athena,“小时”分区键是范围为 0-23 的 integer,格式为两位数(即 zero-filled)。最后,它指定了这些分区键如何映射到 S3 上的分区位置。当查询中的日期为“2020/10/06”时,该字符串将逐字插入到位置模板中。

通过这些更改,您应该能够 运行 查询如下(“date”是保留字,当它是列名时必须用引号引起来):

SELECT *
FROM logs
WHERE "date" = '2020/10/06'
SELECT *
FROM logs
WHERE "date" BETWEEN '2020/10/01' AND '2020/10/06'
AND hour BETWEEN 9 AND 21

请注意,日期格式必须与分区投影配置中的格式完全相同,即YYYY/MM/DD

Theo 的回答特别有助于解决小时和天的位数问题,因为我的 S3 分区格式为 YYYY/MM/DD:

  'projection.hour.digits' = '2'

这是让它在这里发挥作用的关键。谢谢@theo

就我而言,我使用的是镶木地板文件:

    CREATE EXTERNAL TABLE `table_name`(
  `id` string, 
   -- more columns..
)
PARTITIONED BY ( 
  `year` string, 
  `month` string, 
  `day` string)
ROW FORMAT SERDE 
  'org.apache.hadoop.hive.ql.io.parquet.serde.ParquetHiveSerDe' 
STORED AS INPUTFORMAT 
  'org.apache.hadoop.hive.ql.io.parquet.MapredParquetInputFormat' 
OUTPUTFORMAT 
  'org.apache.hadoop.hive.ql.io.parquet.MapredParquetOutputFormat'
LOCATION
  's3://fullbucketname/full_prefix_dir'
TBLPROPERTIES (
  'classification'='parquet', 
  'compressionType'='none', 
  'has_encrypted_data'='false', 
  'projection.day.range'='1,31', 
  'projection.day.type'='integer', 
  'projection.day.digits' = '2',
  'projection.enabled'='true', 
  'projection.month.range'='1,12', 
  'projection.month.type'='integer',
  'projection.month.digits' = '2',
  'projection.year.range'='2020,2051', 
  'projection.year.type'='integer',
'storage.location.template'='s3://s3://fullbucketname/full_prefix_dir/year=${year}/month=${month}/day=${day}', 
  'typeOfData'='file')