EnterpriseDB: The Enterprise Postgres Company Postgres Plus Forums: The PostgreSQL Open Source Database from EnterpriseDB
  [Search] Search   [Recent Topics] Recent Topics   [Hottest Topics] Hottest Topics   [Members]  Member Listing   [Groups] Back to home page 

Replicated Tables  XML

Forum Index » GridSQL - Developers
Author Message
[Post New]27 Jan 2010 21:02:15
    Subject: Replicated Tables
[Up]
John_Smith_3

New member

Joined: 27 Jan 2010 20:53:13
Messages: 1
Offline
What is the execution path for replicated tables? I realize it's not optimal, but if you had a large dimension replicated across the nodes and you wanted to execute a count(*) against that table (no filter and no join), would the engine attempt to resolve it through one node or would all nodes contribute?
[Post New]29 Jan 2010 10:20:33
    Subject: L
[Up]
Mason_S

Senior member

Joined: 1 Apr 2008 09:03:08
Messages: 380
Offline
Hi John,

In the case of replicated tables, it will pick one node and just execute the statement on one.

Note that since GridSQL uses PostgreSQL, count may be a little slow. PostgreSQL is known for executing counts a little bit slowly due to its Multi-version Concurrency Control; it verifies the visibility of each tuple for the current transaction context (snapshot).

Regards,

Mason
[WWW]
 
Forum Index » GridSQL - Developers
Go to:   
Powered by JForum 2.1.8 © JForum Team