Back to List

On-Premises Data Solutions: Pros and Cons

Scott Hietpas Scott Hietpas  |  
Sep 24, 2019
 
In this blog series, Scott Hietpas, a Principal Consultant with Skyline Technologies’ data team, explores the pros and cons of different data storage architecture. For a full overview on this topic, check out the original Cloud vs On-Premises Architecture Webinar.
 
The goal of this blog series is to talk about how we decide between Cloud and On-Premises data solutions. To get started, let’s talk about an on-premises architecture.
 

Typical On-Premises Architecture

A typical organization would have started with some on-premises source systems (like an ERP or CRM), followed by the establishment of an on-premise data warehouse to facilitate reporting. That reporting might have also been on-premises with SQL Reporting Services.
 
That was the early architecture for most organizations. But in recent years we’ve started to see challenges around on-premises architecture from Big Data scenarios. Some of these challenges include:
 
  • Data volume and variety from marketing or social media
  • Data velocity due to the Internet of Things
  • New data science needs
  • Source systems moving to the cloud
 
on-premises architecture
 

Pros of the On-Premises Solution

Many organizations are still getting benefits from an on-premises data warehouse. We’ll cover some of the challenges these systems face, but first it is worth recognizing the advantages.
 

Cost Balance

The cost of an on-premises solution can be a pro or a con, depending on your perspective. It comes down to not so much comparing specific costs but balancing how those costs are paid for. On-premises is much more of an upfront capital expenditure. You have to buy the server and predict what your capacity is going to be, but from that point you can do some write-offs. You can leverage existing investments and maximize the lifespan of your solution.
 

Established Solution

An on-premises solution (SQL Server for example) is a widely available, mature toolset. It's easy to find additional team members that can work with on-prem technology and deliver proven best practices.
 

(The Perception of) Security

I hesitate to say that on-prem is “secure” and put that under a pro. The elephant in the room is that on-premises servers offer perceived data protection, but security really has several factors. The theory is that, if you have a box in your office that is unplugged from the internet, it would be highly secure from outside powers trying to get at your data. The reality is that, even if you are hosting your own on-premise solution, odds are good that it's connected to the Internet somewhere. Because of that, on-premises is not inherently more secure than a cloud-based solution. Security really comes down to the measures you take to secure it - whether those are firewall rules, multi-factor authentication, or other protections.
 

Data Movement

Bandwidth use can be a gray area. However, if you're working with core systems that are on-premises, then having your other systems be on-premises as well can mean less chatter across your Internet bandwidth.
 

Cons of On-Premises

While an on-premises solution has several advantages, there are also downsides.
 

Cost

While you can get a lot out of your up-front investment with an on-premises solution, that also has disadvantages. If you would rather pay each month based on what your actual need is, on-prem does not allow that. It’s all or nothing.
 

Scalability

Another disadvantage of on-prem is the scalability. You have to predict your future capacity upfront to make that initial hardware purchase. Generally, you have to purchase more than what you need right now so you're prepared for future growth. However, if that never happens, you might be scaled too high. Or, alternatively, the growth might be more than what you predicted and lead to an underperforming system.
 

Time to Market

You also don’t have a lot of speed when it comes increasing your capacity. Acquiring new hardware has a lead time to it. If you need to suddenly scale up, you have a delay before you can get additional hardware in place to meet that demand.
 

Update Frequency

Another downside to on-prem servers is the update frequency. Many cloud services have moved toward weekly or monthly updates. However, the on-prem versions of those software systems still tend to be on a quarterly, maybe even an annual, update schedule. The exact update frequency depends on the individual software, but on-prem system updates are less frequent because the manufacturers have less control over all the different factors.
 

Conclusion

Due to a combination of above factors, organizations typically avoid creation of ad-hoc environments to support additional testing or feature development scenarios. These scenarios can often benefit from a separate environment, but the up-front costs, delays in provisioning the environment, and inability to stop or scale down costs when done often makes them not viable.
 
You can see that an on-premises solution has several advantages, but also several sticking points. In my next blog, I’m going to walk through the pros and cons of the Cloud solution.
 
Analytics

 

Love our Blogs?

Sign up to get notified of new Skyline posts.

 


Related Content


Blog Article
Power BI’s Latest Features and How to Use Them
Marcus RadueMarcus Radue  |  
Apr 14, 2020
[Updated 05/11/20]  In this regularly updated blog, Marcus Radue (Data Analytics Engineer) highlights key features from his monthly Power BI Office Hours webinar series so you know how to capitalize on Power BI’s latest enhancements.   Power BI Features (April 2020 Update)
Blog Article
What is Microsoft’s Power Platform and How to Use It: The Guide
Skyline Technologies  |  
Jan 14, 2020
In this guide, Libby Fisette (Director of Skyline Modern Workplace team) and Marcus Radue (Data Analytics Engineer), dig into the functionality of the Microsoft Power Platform and how you can leverage this toolset to solve many business situations. From basics to key questions, you will find...
Blog Article
Realtime and Near-Realtime Data Sources and Data Integrity
Matt PlusterMatt Pluster  |  
Dec 17, 2019
In this blog series, Matt Pluster, Director of Skyline’s Data Analytics Consulting Practice, explores data sources and processing options. For a full overview on this topic, check out the Realtime vs Near-Realtime webinar.   In previous blogs in this series, I dug into advantages...
Blog Article
Mitigating the Risks of Realtime or Near-Realtime Data Processing
Matt PlusterMatt Pluster  |  
Dec 10, 2019
In this blog series, Matt Pluster, Director of Skyline’s Data Analytics Consulting Practice, explores data sources and processing options. For a full overview on this topic, check out the Realtime vs Near-Realtime webinar.   In previous blogs in this series, I’ve talked about...
Blog Article
“The Other Realtime”: Low-Latency Data Processing via DirectQuery
Matt PlusterMatt Pluster  |  
Dec 03, 2019
In this blog series, Matt Pluster, Director of Skyline’s Data Analytics Consulting Practice, explores data sources and processing options. For a full overview on this topic, check out the Realtime vs Near-Realtime webinar.    So far in this blog series, we have talked...