[
https://issues.jboss.org/browse/TEIID-3050?page=com.atlassian.jira.plugin...
]
Steven Hawkins commented on TEIID-3050:
---------------------------------------
Looking at this more, we do little in the way of parallel read/write in the current code
anyway. As requests are sent to be executed we block on the results - without performing
a prefetch. So I'll take a look at improving that as well.
allow for more incremental insert with iterator
-----------------------------------------------
Key: TEIID-3050
URL:
https://issues.jboss.org/browse/TEIID-3050
Project: Teiid
Issue Type: Enhancement
Security Level: Public(Everyone can see)
Components: Misc. Connectors, Query Engine
Reporter: Steven Hawkins
Assignee: Steven Hawkins
Fix For: 8.9
The current bulk insert logic assumes that it is avoiding a global transactions by fully
reading the results before sending the tuplebuffer iterator to the source for processing,
which can then just use a local transaction. However if we are already operating under a
transaction or the user doesn't want a transaction (such as with a staging
materialized view load) then the insert should happen more incrementally.
--
This message was sent by Atlassian JIRA
(v6.2.6#6264)