[
https://issues.apache.org/jira/browse/PHOENIX-7458?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17897105#comment-17897105
]
Viraj Jasani commented on PHOENIX-7458:
---------------------------------------
We can do that too, but separate table is considered to not worry about
maintaining 1:n relationships in future (sort of, learnings from
SYSTEM.CHILD_LINK use case, though this is not expected to be as gigantic as
num of views).
> Create new SYSTEM tables for tracking CDC Stream metadata
> ---------------------------------------------------------
>
> Key: PHOENIX-7458
> URL: https://issues.apache.org/jira/browse/PHOENIX-7458
> Project: Phoenix
> Issue Type: Sub-task
> Reporter: Palash Chauhan
> Assignee: Palash Chauhan
> Priority: Major
>
> # *SYSTEM.CDC_STREAM_STATUS :* Store the mapping from Table Name to a Stream
> Name (and its status). CDC can be disabled on a table and then enabled again
> - we will create a new Stream Name in that case. There can only be one Active
> stream per table - Stream Status can be part of the primary key for efficient
> lookup of the Stream Name, given a Table Name.
> ## TABLE_NAME : VARCHAR
> ## STREAM_STATUS : VARCHAR
> ## STREAM_NAME : VARCHAR
> ## _PK - (TABLE_NAME, STREAM_STATUS)_
> # *SYSTEM.CDC_STREAM :* Store metadata related to all open/closed partitions
> of a stream. Partitions represent hbase regions of the data table for which a
> CDC stream was enabled/created.
> ## TABLE_NAME : VARCHAR
> ## STREAM_NAME : VARCHAR
> ## PARTITION_ID : VARCHAR
> ## PARENT_PARTITION_ID : VARCHAR
> ## PARTITION_START_TIME : DATE/BIGINT
> ## PARTITION_END_TIME : DATE/BIGINT
> ## PARTITION_START_KEY : VARBINARY_ENCODED
> ## PARTITION_END_KEY : VARBINARY_ENCODED
> ## _PK - (TABLE_NAME, STREAM_NAME, PARTITION_ID)_
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
