You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by "ASF GitHub Bot (JIRA)" <ji...@apache.org> on 2018/11/02 04:40:00 UTC
[jira] [Commented] (FLINK-10697) Create an in-memory catalog that
stores Flink's meta objects
[ https://issues.apache.org/jira/browse/FLINK-10697?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16672563#comment-16672563 ]
ASF GitHub Bot commented on FLINK-10697:
----------------------------------------
bowenli86 opened a new pull request #6997: [FLINK-10697] Create an in-memory catalog that stores Flink's meta objects
URL: https://github.com/apache/flink/pull/6997
## What is the purpose of the change
Currently all Flink meta objects (currently tables only) are temporarily stored in memory as part of Calcite catalog. As we introduce catalog for those objects (tables, views, and UDFs), it makes sense to organize them neatly. Further, having a catalog implementation that store those objects in memory is to retain the currently behavior, which can be configured by user.
`FlinkInMemoryCatalog` is similar to existing `InMemoryExternalCatalog` class, which exists only for testing purpose. `FlinkInMemoryCatalog` requires a minimum of implementation of `CrudExternalCatalog` interface. For instance, it doesn’t need subcatalogs. While it’s possible to extract some common code from `InMemoryExternalCatalog` and let `InMemoryExternalCatalog` extend from `FlinkInMemoryCatalog`, it’s very likely the two will divert at some point. Thus, it’s acceptable to have two independent classes from the very beginning.
## Brief change log
- added an in-memory impl of `CrudExternalCatalog` as `FlinkInMemoryCatalog` for production use
- renamed `InMemoryExternalCatalog` to `TestingInMemoryCatalog` to clarify its usage scope
## Verifying this change
This change is already covered by existing tests, such as *InMemoryExternalCatalogTests*.
## Does this pull request potentially affect one of the following parts:
none
## Documentation
- Does this pull request introduce a new feature? (yes)
- If yes, how is the feature documented? (JavaDocs)
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
> Create an in-memory catalog that stores Flink's meta objects
> ------------------------------------------------------------
>
> Key: FLINK-10697
> URL: https://issues.apache.org/jira/browse/FLINK-10697
> Project: Flink
> Issue Type: Sub-task
> Components: Table API & SQL
> Affects Versions: 1.6.1
> Reporter: Xuefu Zhang
> Assignee: Bowen Li
> Priority: Major
> Labels: pull-request-available
> Fix For: 1.8.0
>
>
> Currently all Flink meta objects (currently tables only) are stored in memory as part of Calcite catalog. Those objects are temporary (such as inline tables), others are meant to live beyond user session. As we introduce catalog for those objects (tables, views, and UDFs), it makes sense to organize them neatly. Further, having a catalog implementation that store those objects in memory is to retain the currently behavior, which can be configured by user.
> Please note that this implementation is different from the current {{InMemoryExternalCatalog}, which is used mainly for testing and doesn't reflect what's actually needed for Flink meta objects.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)