Published: November 10, 2010 - 01:00

Three Configurations for Dynamic Scalability and Deployment

In enterprise search, quality, usability and style are as important as relevancy of results and performance to engage your users right from the start.
Let's take a look at typical scale-out scenarios that become relevant when implementing enterprise environments with Fabasoft Mindbreeze.


Starting with Entry-Level

The first configuration consists of a single Fabasoft Mindbreeze Appliance. It allows you to crawl all your data sources, analyze and process all your documents and records and provides  comprehensive unified information access to all your users.
Don't get me wrong, with “entry-level”: you can go all up to at least 6 million documents with one entry-level appliance. However, in large-scale enterprise environments, your data volume can easily blow this limit. So we go to the next level.

Entry-Level Configuration

Taming huge data volume by automatic data partitioning

So what can you do if your data sources hold so many documents and records, that one server is not enough?
The scale-out configuration aims at large-scale enterprise infrastructures with terabytes of data from millions of documents and records.  To support such scenarios, you only need to add additional appliance nodes.
Fabasoft Mindbreeze will automatically and uniformly distribute the documents over all available appliance nodes, thus scaling-out to match the data volume while maintaining excellent query performance.

You need highest availability? Index Replication is the way to go

Another axis of scale-out is targeted in the third configuration: High Availability End users demand their systems to be available 24/7 and to guarantee shortest query times.
Fabasoft Mindbreeze achieves that via index replication. By scaling-out an existing appliance infrastructure, the replica nodes inherit the data sources crawling capabilities as well as the queryable index of the original nodes.

Ultimate query performance through designated query nodes

Complex document analysis and document processing demands resources.
On the other hand, Web 2.0 applications and Cloud Computing demand high
availability with exponential growing numbers of concurrent users as the mission critical requirement.
For ultimate query performance, the fourth configuration decouples data source crawler (producer) nodes completely from query (consumer) nodes:

  1. Data sources can be crawled independently of the query nodes in terms of time and resource consumption.
  2. The indices produced by the crawler nodes can be replicated to the query nodes and made searchable without any downtime, using our Hot Reopen technology.
  3. Combined with a load-balancer, the appliances hosting the query nodes can be
    scaled independently of the crawler nodes, to keep up with the growing number of concurrent users.



Daniel Fallmann

Stephen
Arnold referenced this article in his blog. Read their opinion on our appliance
here: http://arnoldit.com/wordpress/2010/11/19/fabasoft-mindbreeze-search-appliance/

Submitted: February 3rd, 2012 06:14


SharePoint 2010, Windows Azure, and the Issue ofScalability : Beyond Search

[...]
Fabasoft Mindbreeze is one excellent third-party option for an organization of
any size looking for an enterprise solution. 
Their scalable offerings are customized based on the needs of the
customer, and expansion is an easy option whenever needed.  Read more about “Three Configurations for
Dynamic Scalability and Deployment.” [...]

Submitted: January 27th, 2012 06:14 


The Good, the Bad, and the Ugly of SharePointHosting Options : Beyond Search

[...]
different scalable options depending on an organization’s size and needs.  Read more in, “Three Configurations for
Dynamic Scalability and Deployment.” In enterprise search, quality, usability
and style are as important as relevancy of results and [...]

Submitted: January 17th, 2012 06:14