谷歌云bigtable列版本不会被删除

本文介绍了谷歌云bigtable列版本不会被删除的处理方法，对大家解决问题具有一定的参考价值，需要的朋友们下面随着小编来一起学习吧！ 问题描述

我们创建了一个包含两列系列的云大表格。一列有30个版本，另一个有1个版本。但是，当我们查询表时，我们得到了多个版本的列，我们已经将最大版本数设置为1。

表创建语句：

创建'myTable'，{NAME => 'cf1'，VERSIONS => '30'}，{NAME => 'cf2'，VERSIONS => '1'} 描述'myTable'： {NAME => 'cf2'，BLOOMFILTER => 'ROW'，版本=> '** 1 **'，IN_MEMORY => 'false'，KEEP_DELETED_CELLS => 'FALSE'，DATA_BLOCK_ENCODING => 'NONE'，TTL => 'FOREVER'，COMPRESSION => 'NONE'，MIN_VERSIONS => '0'，BLOCKCACHE => 'true'，BLOCKSIZE => '65536'，REPLICATION_SCOPE => '0'} {NAME => 'cf1'，BLOOMFILTER => 'ROW'，版本=> '** 30 **'，IN_MEMORY => 'false'，KEEP_DELETED_CELLS => 'FALSE'，DATA_BLOCK_ENCODING => 'NONE'，TTL => 'FOREVER'，COMPRESSION => 'NONE'，MIN_VERSIONS => '0'，BLOCKCACHE => 'true'，BLOCKSIZE => '65536'，REPLICATION_SCOPE => '0'}

bigtable垃圾回收是如何工作的？它多久删除旧版本？或者我们是否在创建表时丢失了一些东西？ 解决方案

从Bigtable Docs中删除： ，因此您可能仍然能够在几天之后读取数据。

链接到文档

We have created a table in cloud bigtable with two column families. One column family with 30 versions and the other with 1 version. However, when we query the table we are getting multiple versions of the columns for which we have set max number of versions to 1.

Table create statement:

create 'myTable', {NAME => 'cf1', VERSIONS => '30'}, {NAME => 'cf2', VERSIONS => '1'} Describe 'myTable': {NAME => ‘cf2’, BLOOMFILTER => ‘ROW’, VERSIONS => ‘**1**’, IN_MEMORY => ‘false’, KEEP_DELETED_CELLS => ‘FALSE’, DATA_BLOCK_ENCODING => ‘NONE’, TTL => ‘FOREVER’, COMPRESSION => ‘NONE’, MIN_VERSIONS => ‘0’, BLOCKCACHE => ‘true’, BLOCKSIZE => ‘65536’, REPLICATION_SCOPE => ‘0’} {NAME => ‘cf1’, BLOOMFILTER => ‘ROW’, VERSIONS => ‘**30**’, IN_MEMORY => ‘false’, KEEP_DELETED_CELLS => ‘FALSE’, DATA_BLOCK_ENCODING => ‘NONE’, TTL => ‘FOREVER’, COMPRESSION => ‘NONE’, MIN_VERSIONS => ‘0’, BLOCKCACHE => ‘true’, BLOCKSIZE => ‘65536’, REPLICATION_SCOPE => ‘0’}

How does the bigtable garbage collection work? How frequently does it delete the older versions? or are we missing something while creating the table ?

解决方案

From Bigtable Docs: Deletion of values happens opportunistically in the background, so you might still be able to read the data for several days after it has expired.

Link to docs