What is the Difference Between Partitions and Clusters?

If you’re new to pricing software having spent most of the last generation with your company’s Pricing Department doing all their work in Excel, the chances are you’re now faced with a new range of terms you have never encountered before.  For example, using the power of the innovative technology at your fingertips, you want to differentiate between the prices you charge across different regions because the distribution costs vary wildly between California and New York, or Germany and Spain. But what the heck is the point of the differentiation called? Is it a partition, a cluster, a tenant, production (or PROD as it is commonly referred to for short) or different instances? We feel your pain. 

At Pricefx, we’ve spent more than a decade assisting our clients come to terms with their new pricing software technology and helping them to make the most of it and become more successful and profitable along the way. From our real-life experiences with our clients, we have put together this article as a guide to what are the differences between partitions, clusters, production (PROD) and instances. 

So, let’s dive straight in and go through each of the terms mentioned considering what they are, what they do, examples of each and how they apply in your everyday pricing software use.

What Are Clusters and Instances in Pricing Software & What Do They Do?

Okay, straight out of the box, if you hear or read the two terms, ‘clusters’ or ‘instances’, they are precisely the same thing.  

 

The terms are 100% interchangeable. 

 

Clusters = Instances 

 

Pricing software companies generally use one or the other of the two terms (At Pricefx we prefer the term ‘instance’), but if you are used to the term ‘cluster’, rest assured, they are the same thing.

 

As many of you are asking about clusters, we will retain that term throughout this article for consistency.

Cluster-based (or instance-based) architectures help pricing software companies (like us) to deploy and manage their applications in the cloud. What that means is that several computers (two or more), simply run in parallel to achieve a common goal. The shared workload allows a high number of parallel tasks to be distributed among the different clusters of systems.

As a result, these individual tasks can leverage the combined memory and processing power of each computer to increase overall performance of your entire pricing software system.

 

Clustering your data allows you to configure your pricing software as a grouping, so that multiple systems can work together to maximize the access to your data in the way THAT YOU WANT IT to achieve the business objectives you’re seeking, and to prevent data loss.

When your pricing software updates, the clusters are overarching architecture of your pricing software that updates and powers the functionality of the system. Filtering down, each of the clusters can have a number of partitions that can be dedicated to other purposes of your software like integration, price regionality, etc.

What Are Examples of Clusters and What Are They Used For?

Clusters are used for different purposes in your pricing software. For example, your organization may have a Production (PROD) Cluster, a QA Cluster, a DEV cluster etc. Each of those Clusters is a separate entity. That means, when your pricing software provider runs their update and upgrade to a new software version, everything in the entire cluster and everything filtered down below (e.g. including all partitions) will be updated. 

Some examples of different clusters (and their uses) in pricing software include; 

• QA Cluster – Usually used for testing, development, integration and Quality Assurance purposes.  
• PROD Cluster – A product cluster will be your company’s everyday pricing software working environment. The production datasets are the real-world data regularly collected in the process of running your business, as opposed to sample and test data used in the QA cluster environment.  

This is what is sometimes simply referred to as a ‘Product’, ‘PROD’ or ‘Production’ in pricing software terminology. 

 

Do not get it confused, it is simply your everyday working environment.

• DEV Cluster – A DEV Cluster is another cluster environment used for development and testing. Some pricing software companies choose not to have DEV Clusters as they are often considered an additional piece of wasted infrastructure – as most development and testing can be done in a QA Cluster environment. 

Benefits of Using Clusters in Pricing Software 

Using clusters helps create what are called multi-tenant systems which enable your pricing software provider to service a theoretical infinite number of customers. But what do you get out of it?  

There are a range of benefits to your business including: 

• Availability of Your Pricing Software – the accessibility of a system experiences little downtime due the solid ‘load-spreading’ design of the architecture. 
• Resilience of Your Pricing Software – In the unlikely event of a system failure, you can be back up again and running quickly. 
• Fault tolerance – The ability of your pricing software to continue providing a service in the event of a failure. 
• Reliability – Your pricing software will function as expected over time.
• Redundancy – Critical resources that apply to your unique pricing software set-up improve overall system reliability. 

What Are Partitions in Pricing Software & What Do They Do? 

As discussed above, each of the clusters of your pricing software can have a number of nodes that can be dedicated to different purposes of your pricing software like integration, price regionality, countries, currencies etc. While you may share your cluster space with many other companies, the partition environment is where the pricing software can drill down into your data exclusively to achieve what your business specifically wants to achieve.  

The actual number of partitions in your pricing is usually driven by your company’s own business model and organizational objectives governing what you require the pricing software to do uniquely for your business. For example, each of the Business Units of your organization may have separate partitions, and within that, those partitions can have identical or separate modules, configuration and logic elements. 

What Are Examples of Partitions and What Are They Used For? 

The number of different types of partitions are (more or less) endless. However, to streamline your pricing software experience, think strategically about the points of differentiation you wish to consider in your pricing strategy and discuss it with your pricing software provider when considering how you want your unique system configured.  

The easiest way to think about partitions and how they work is with a practical real-life user model. 

For example, you may have a business selling electrical goods across different countries.  

As part of your company’s business strategy, you might want to have a different pricing strategy for each country as the regions are distinct entities and have different demographics, currencies and customer behaviors plus other unique market forces. In this case, each country’s data can be a partition. 

You may want to break it down even further and differentiate your pricing strategy for your retail electrical goods between your online shops and your bricks-and-mortar stores.  

Each data breakdown between online stores and bricks-and-mortar stores can be an additional partition. 

Benefits of Using Partitions in Pricing Software 

The major positives to come from partitioning your data for use in pricing software:

• Improve performance – Each partition breaks down your data into smaller units. Correctly done, partitioning can make your system more efficient. Functions that affect more than one partition can run in parallel. 
• Improve security – If you need to, you may be able to separate sensitive and non-sensitive data into different partitions. 
• Provide operational flexibility – Fine-tune your pricing operations, maximize administrative efficiency, and minimize cost. What’s more, as mentioned above, partitioning can enable you to run different pricing strategies across different sectors or regions of your business. 
• Improve availability – Separating and spreading your data through partitioning avoids a single point of failure. If one partition fails, only the data in that partition is unavailable. Operations on other partitions can continue to run. 

How Many Users Can Simultaneously Connect to the Pricefx Partition & Does the Number Affect the Performance?

It depends.

 

 

There is no limit set on the number of simultaneous connections to the Pricefx partition. It depends on the number of concurrent operations that users are doing in the partition, and also depends on the “size” or complexity of the operations that users are performing.

 

 

Pricefx does have customers with approximately 800 users using one partition within the same hour for example.

 

 

It also depends on the size of the partition and the cluster (shared vs dedicated).

 

 

So, the answer in a nutshell is: depends on the complexity of the concurrent operations users are performing at the same time, the size of the cluster, the usage of shared cluster vs private cluster and the complexity of the calculations.

 

 

However, that being said, we do have some customers operating Pricefx in a standard configuration, with over 1k of end users (including salespeople using the system without any performance degradation.

 

What Support Can I Expect to Implement Pricing Software?

Now you know the differences in the terms partitions, clusters, production (product) and instances plus how they apply in the pricing software environment. What’s more, you now also know how many users can simultaneously connect to the Pricefx partition and if it will affect your system’s performance. Now you have all that information, you may be wondering what support you can receive once you have purchased your pricing software to make maximum use of its functionality.

To help you understand the software support you will receive from Pricefx after implementation of your pricing project and remind you that we are with you for the long haul or your business partner, check out this handy article to learn more: