Skip Ribbon Commands
Skip to main content
Navigate Up
Sign In

Quick Launch

Average Rating:

facebook Twitter
Email
Print Bookmark Alert me when this article is updated

Feedback

The distributed search is seen to be slower than the non-distributed search in Relate 360
Problem Description
In Relate 360, the distributed search is a search model that distributes a search request to multiple region servers of the repository and performs the search directly on the region servers.

Distributed search optimally uses the available resources and should result in improved search performance.

However, in earlier versions of Relate 360 (prior to 10.1), it is seen that, instead of increasing performance, enabling the distributed search actually decreases the performance.
Cause
This issue occurs due to a design flaw in previous incarnations of the distributed search (DS), which can perform a lot more matches than the non-distributed search (NDS) since a particular candidate record can fall into multiple ranges for the same match-rule/input record.

In NDS, since all the candidates are retrieved at the client-side, the records that have been already matched against the input record can be tracked. 

However, in case of DS, since the matching takes place at the region level, the matches that have already been done cannot be tracked, hence the same candidate may be matched multiple times.
Solution
From Relate 360 10.1 onwards, the distributed search has been redesigned and additional concurrency features have been added. To resolve the issue of a slowdown in the DS performance, upgrade to Relate 360 10.1.
More Information
Applies To
Product: Relate 360
Problem Type: Product Feature
User Type: Architect
Project Phase: Implement
Product Version:
Database:
Operating System:
Other Software:

Reference
Attachments
Last Modified Date:2/13/2020 2:15 AMID:594284
People who viewed this also viewed

Feedback

Did this KB document help you?



What can we do to improve this information (2000 or fewer characters)