[teiid-issues] [JBoss JIRA] (TEIID-3509) Consolidate internal and external materialization management

[Kylin Soong (JIRA)]

    [ https://issues.jboss.org/browse/TEIID-3509?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13239105#comment-13239105 ] 

Kylin Soong edited comment on TEIID-3509 at 5/18/16 6:19 AM:
-------------------------------------------------------------

> 1) The execution looks correct, what is the error you are seeing?
I was gotten confused by the document, but if look at the source code it's easy to make sense 
{code}
CREATE FOREIGN PROCEDURE refreshMatViewRow(OUT RowsUpdated integer NOT NULL RESULT, IN ViewName string NOT NULL, IN Key object NOT NULL, VARIADIC KeyOther object) OPTIONS (UPDATECOUNT 1)
{code}
Note that Key is the PK column, but PK may contain more columns, VARIADIC KeyOther is for others, so if the mat view looks(only 1 column id is the PK)
{code}
+-----+----+----+----+
| id  | a  | b  | c  |
+-----+----+----+----+
| 100 | a0 | b0 | c0 |
| 101 | a1 | b1 | c1 |
| 102 | a2 | b2 | c2 |
+-----+----+----+----+
{code}
The SQL
{code}
EXEC SYSADMIN.refreshMatViewRows('TestInterMat.SAMPLEMATVIEW', ('101', '102'))
{code}
is not  make sense. If want update two rows use
{code}
EXEC SYSADMIN.refreshMatViewRows('TestInterMat.SAMPLEMATVIEW', ('101',), ('102',))
{code}
If PK have more columns, for example, id, a and b compose of the PK, then
{code}
EXEC SYSADMIN.refreshMatViewRows('TestInterMat.SAMPLEMATVIEW', ('101', 'a1', 'b1'), ('102', 'a2', 'b2'));
{code}
will update 2 rows, I will update the document.

By the way, as per Steven's comment in yesterday's meeting, refresh via PK is related with cluster, but current external update not use primary key, does this fine?





was (Author: kylin):
> 1) The execution looks correct, what is the error you are seeing?
I was gotten confused by the document, but if look at the source code it's easy to make sense 
{code}
CREATE FOREIGN PROCEDURE refreshMatViewRow(OUT RowsUpdated integer NOT NULL RESULT, IN ViewName string NOT NULL, IN Key object NOT NULL, VARIADIC KeyOther object) OPTIONS (UPDATECOUNT 1)
{code}
Note that Key is the PK column, but PK may contain more columns, VARIADIC KeyOther is for others, so if the mat view looks(only 1 column id is the PK)
{code}
+-----+----+----+----+
| id  | a  | b  | c  |
+-----+----+----+----+
| 100 | a0 | b0 | c0 |
| 101 | a1 | b1 | c1 |
| 102 | a2 | b2 | c2 |
+-----+----+----+----+
{code}
The SQL
{code}
EXEC SYSADMIN.refreshMatViewRows('TestInterMat.SAMPLEMATVIEW', ('101', '102'))
{code}
is not  make sense. If PK have more columns, for example, id, a and b compose of the PK, then
{code}
EXEC SYSADMIN.refreshMatViewRows('TestInterMat.SAMPLEMATVIEW', ('101', 'a1', 'b1'));
EXEC SYSADMIN.refreshMatViewRows('TestInterMat.SAMPLEMATVIEW', ('102', 'a2', 'b2'))
{code}
Will work, each SQL only update one row. I will update the document.

By the way, as per Steven's comment in yesterday meeting, refresh via PK is related with cluster, but current external update not use primary key, does this fine?




> Consolidate internal and external materialization management
> ------------------------------------------------------------
>
>                 Key: TEIID-3509
>                 URL: https://issues.jboss.org/browse/TEIID-3509
>             Project: Teiid
>          Issue Type: Enhancement
>          Components: Server
>    Affects Versions: 8.7
>            Reporter: Ramesh Reddy
>            Assignee: Kylin Soong
>             Fix For: 9.0
>
>
> - Currently there separate SYSADMIN procedures for internal and external materialization to refresh them, we need to provide single set
> - The management of the (usage of status table) is not fully utilized by the internal materialization this needs to addressed such that both external and internal use this table to keep track of load status



--
This message was sent by Atlassian JIRA
(v6.4.11#64026)