cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Constance Eustace (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (CASSANDRA-7062) Extension of static columns for compound cluster keys
Date Mon, 21 Apr 2014 15:16:15 GMT

     [ https://issues.apache.org/jira/browse/CASSANDRA-7062?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Constance Eustace updated CASSANDRA-7062:
-----------------------------------------

    Description: 
CASSANDRA-6561 implemented static columns for a given partition key.

What this is proposing for a compound cluster key is a static column that is static at intermediate
parts of a compound cluster key. This example shows a table modelling a moderately complex
EAV pattern  :

CREATE TABLE t (
   entityID text,
   propertyName text,
   valueIndex text,
   entityName text static (entityID),
   propertyType text static (propertyName),
   propertyRelations List<text> static (propertyName),
   PRIMARY KEY (entityID, (propertyName,valueIndex))
)

So in this example has the following static columns:
- the entityName column behaves exactly as CASSANDRA-6561 details, so all cluster rows have
the same value
- the propertyName column is static with respect to the remaining parts of the cluster key
(that is, across all valueIndex values for a given propertyName)

Is this a relatively simple extension of the same mechanism in -6561, or is this a "whoa,
you have no idea what you are proposing"?


  was:
CASSANDRA-6561 implemented static columns for a given partition key.

What this is proposing for a compound cluster key is a static column that is static at intermediate
parts of a compound cluster key. This example shows a table modelling a moderately complex
EAV pattern  :

CREATE TABLE t (
   entityID text,
   propertyName text,
   valueIndex text,
   entityName text static (entityID),
   propertyType text static (propertyName),
   propertyRelations List<text> static (propertyName),
   PRIMARY KEY (entityID, (propertyName,valueIndex)
)

So in this example has the following static columns:
- the entityName column behaves exactly as CASSANDRA-6561 details, so all cluster rows have
the same value
- the propertyName column is static with respect to the remaining parts of the cluster key
(that is, across all valueIndex values for a given propertyName)

Is this a relatively simple extension of the same mechanism in -6561, or is this a "whoa,
you have no idea what you are proposing"?



> Extension of static columns for compound cluster keys
> -----------------------------------------------------
>
>                 Key: CASSANDRA-7062
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-7062
>             Project: Cassandra
>          Issue Type: New Feature
>            Reporter: Constance Eustace
>            Priority: Minor
>
> CASSANDRA-6561 implemented static columns for a given partition key.
> What this is proposing for a compound cluster key is a static column that is static at
intermediate parts of a compound cluster key. This example shows a table modelling a moderately
complex EAV pattern  :
> CREATE TABLE t (
>    entityID text,
>    propertyName text,
>    valueIndex text,
>    entityName text static (entityID),
>    propertyType text static (propertyName),
>    propertyRelations List<text> static (propertyName),
>    PRIMARY KEY (entityID, (propertyName,valueIndex))
> )
> So in this example has the following static columns:
> - the entityName column behaves exactly as CASSANDRA-6561 details, so all cluster rows
have the same value
> - the propertyName column is static with respect to the remaining parts of the cluster
key (that is, across all valueIndex values for a given propertyName)
> Is this a relatively simple extension of the same mechanism in -6561, or is this a "whoa,
you have no idea what you are proposing"?



--
This message was sent by Atlassian JIRA
(v6.2#6252)

Mime
View raw message