实践数据湖iceberg 第二课 iceberg基于hadoop的底层数据格式

 

前言

iceberg底层是如何管理数据，具体每次数据变更，在底层数据结构上发生哪些变化？
本文主要解答这个问题。
观察的方法是：
1.建立catalog, hdfs截图
2.创建表，hdfs截图
3.insert一条数据，hdfs截图，观察元数据和数据的变更
4.再insert一条数据，hdfs截图，观察元数据和数据的变更
5.结论
6.下一步，观察hive catalog

---

提示：以下是本篇文章正文内容，下面案例可供参考

1、hadoop catalog创建

创建脚本，warehouse的路径，它会自动创建
hdfs路径里面 ns是命名空间，但namenode的使用ip:port代替
在flink-sql client 中执行脚本

sql-client.sh embedded -j /opt/software/iceberg-flink-runtime-0.11.1.jar shell`

1.1 执行建立 catalog脚本

CREATE CATALOG hadoop_catalog2 WITH (
  'type'='iceberg',
  'catalog-type'='hadoop',
  'warehouse'='hdfs://ns/user/hive/warehouse/iceberg_hadoop_catalog2',
  'property-version'='1'
);

1.2 查看hdfs 目录

发现：建好warehouse到defalut的目录 /user/hive/warehouse/iceberg_hadoop_catalog2/default

2.建表

2.1 建表

建表语句：

CREATE TABLE `hadoop_catalog2`.`default`.`sample` (
    id BIGINT COMMENT 'unique id',
    data STRING
);

执行：

Flink SQL> CREATE TABLE `hadoop_catalog2`.`default`.`sample` (
>     id BIGINT COMMENT 'unique id',
>     data STRING
> );
[INFO] Table has been created.

2.2 查看建表后的目录结构

用命令行查看

[root@hadoop101 software]# hadoop fs -ls -R /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/
drwxr-xr-x   - root supergroup          0 2022-01-13 14:29 /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/metadata
-rw-r--r--   2 root supergroup        826 2022-01-13 14:29 /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/metadata/v1.metadata.json
-rw-r--r--   2 root supergroup          1 2022-01-13 14:29 /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/metadata/version-hint.text

浏览器查看：

看看这个两个文件的内容：

[root@hadoop101 software]# hadoop fs -cat /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/metadata/v1.metadata.json
{
  "format-version" : 1,
  "table-uuid" : "956e7d6f-7184-4147-990e-90c923c43c2f",
  "location" : "hdfs://ns/user/hive/warehouse/iceberg_hadoop_catalog2/default/sample",
  "last-updated-ms" : 1642055374103,
  "last-column-id" : 2,
  "schema" : {
    "type" : "struct",
    "fields" : [ {
      "id" : 1,
      "name" : "id",
      "required" : false,
      "type" : "long"
    }, {
      "id" : 2,
      "name" : "data",
      "required" : false,
      "type" : "string"
    } ]
  },
  "partition-spec" : [ ],
  "default-spec-id" : 0,
  "partition-specs" : [ {
    "spec-id" : 0,
    "fields" : [ ]
  } ],
  "default-sort-order-id" : 0,
  "sort-orders" : [ {
    "order-id" : 0,
    "fields" : [ ]
  } ],
  "properties" : { },
  "current-snapshot-id" : -1,
  "snapshots" : [ ],
  "snapshot-log" : [ ],
  "metadata-log" : [ ]
}[root@hadoop101 software]# hadoop fs -cat /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/metadata/version-hint.text
1[root@hadoop101 software]#

1.写入数据

1.1 写入1条数据

执行写入sql

Flink SQL> INSERT INTO `hadoop_catalog2`.`default`.`sample` VALUES (1, 'a');
[INFO] Submitting SQL update statement to the cluster...
[INFO] Table update statement has been successfully submitted to the cluster:
Job ID: 794a9c88a06bd2e889968ee11d213a93

1.2 写入数据本质上触发一个flink jiob

看日志,flink中生成一个job，到管理页面看看，发现是生成了

1.3 查看hdfs目录

查看hdfs的目录，发现生成了v2.metadata.json 和data目录的数据文件

[root@hadoop101 software]# hadoop fs -ls -R /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample
drwxr-xr-x   - root supergroup          0 2022-01-13 14:55 /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/data
-rw-r--r--   2 root supergroup        637 2022-01-13 14:55 /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/data/00000-0-2f98c753-78a8-434f-ba52-60aa9b9a95bf-00001.parquet
drwxr-xr-x   - root supergroup          0 2022-01-13 14:55 /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/metadata
-rw-r--r--   2 root supergroup       5591 2022-01-13 14:55 /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/metadata/47517408-7822-4345-940b-a4ee956ab29f-m0.avro
-rw-r--r--   2 root supergroup       3522 2022-01-13 14:55 /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/metadata/snap-7677920539998136227-1-47517408-7822-4345-940b-a4ee956ab29f.avro
-rw-r--r--   2 root supergroup        826 2022-01-13 14:29 /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/metadata/v1.metadata.json
-rw-r--r--   2 root supergroup       1825 2022-01-13 14:55 /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/metadata/v2.metadata.json
-rw-r--r--   2 root supergroup          1 2022-01-13 14:55 /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/metadata/version-hint.text

1.4 再写一条数据，并观察

Flink SQL> INSERT INTO `hadoop_catalog2`.`default`.`sample` VALUES (2, 'b');
[INFO] Submitting SQL update statement to the cluster...
[INFO] Table update statement has been successfully submitted to the cluster:
Job ID: 4ce2ab5fc0f12ccc196a25da8c307028

查看hdfs文件

[root@hadoop101 software]# hadoop fs -ls -R /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample
drwxr-xr-x   - root supergroup          0 2022-01-13 15:03 /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/data
-rw-r--r--   2 root supergroup        637 2022-01-13 14:55 /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/data/00000-0-2f98c753-78a8-434f-ba52-60aa9b9a95bf-00001.parquet
-rw-r--r--   2 root supergroup        636 2022-01-13 15:03 /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/data/00000-0-d753dd95-bfae-4bd7-be87-d6f6cc53ae60-00001.parquet
drwxr-xr-x   - root supergroup          0 2022-01-13 15:03 /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/metadata
-rw-r--r--   2 root supergroup       5591 2022-01-13 14:55 /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/metadata/47517408-7822-4345-940b-a4ee956ab29f-m0.avro
-rw-r--r--   2 root supergroup       5591 2022-01-13 15:03 /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/metadata/7277ffe1-6e1e-4959-9bbd-16e583835dee-m0.avro
-rw-r--r--   2 root supergroup       3588 2022-01-13 15:03 /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/metadata/snap-4489783265506344084-1-7277ffe1-6e1e-4959-9bbd-16e583835dee.avro
-rw-r--r--   2 root supergroup       3522 2022-01-13 14:55 /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/metadata/snap-7677920539998136227-1-47517408-7822-4345-940b-a4ee956ab29f.avro
-rw-r--r--   2 root supergroup        826 2022-01-13 14:29 /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/metadata/v1.metadata.json
-rw-r--r--   2 root supergroup       1825 2022-01-13 14:55 /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/metadata/v2.metadata.json
-rw-r--r--   2 root supergroup       2858 2022-01-13 15:03 /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/metadata/v3.metadata.json
-rw-r--r--   2 root supergroup          1 2022-01-13 15:03 /user/hive/warehouse/iceberg_hadoop_catalog2/default/sample/metadata/version-hint.text

1.5 结论

对追加写入数据管理方式：生成新的medata和snapshot快照文件

现在好奇，对update语句，对文件的组织有什么影响？
接下来把id=1的数据进行修改

发现：不支持update语句

总结