On-Premise vs. Cloud: How to Know Which Deployment Model is Right for You
So you’ve found a great video analytics software and now you’re looking to deploy it. You’ve got a few options here – cloud, hybrid, or on-premise.
So which one is best for you?
The answer to this question depends on your particular situation, but here are the basic concepts to start:
- Cloud: A fully cloud-based deployment means that all of your software (i.e.: analytics processing and management interface) is hosted online through a cloud-service provider which can be accessed through your web browser.
- Hybrid: A middle ground between a cloud deployment and on-premise deployment. In this scenario, your analytics processing is done locally (on-premise) and your management interface is hosted in the cloud.
- On-premise: Your software is hosted locally (on your local network) and installed on your own hardware (i.e.: data servers).
Each model can adapt to your existing infrastructure and scale with your business. Contrary to popular belief, all three models are also extremely secure against cyber threats.
Before you go further! There are a few key questions you should keep top of mind when deciding which model is best for you:
- How much responsibility do I want to have over my IT (ie: ensuring compliance with data protection laws)?
- How quickly do I need my system up and running?
- How quick and reliable is your internet connection?
Let’s get into the key differences between cloud, hybrid, and on-premise deployments.
One of the best features of a fully cloud-based deployment is that you don’t need to set up, maintain, fix, or monitor your IT – your provider will do it for you.
Cloud infrastructures are designed to comply with data protection laws and to be secure against cyber attacks, so you don’t need to worry about data breaches and fines.
Is it important that your system is always up to date with the newest features? With a cloud system, you never need to worry. Automatic, 24/7 updates will be dealt with by your provider.
Here, IT updates and maintenance are typically covered by the provider that hosts your core infrastructure. This includes updates for your on-premise analytics processing devices (“edge devices”) as they are connected.
You will, however, be responsible for fixing and potentially replacing your edge machines over time.
On-premise deployments come with considerably more IT responsibility. Because the system is hosted locally, you will be responsible for setup and maintenance. Since you are storing your data locally, you will also need to ensure that you are handling data in compliance with local data protection laws.
But with more responsibility comes more control. If you operate in a unique environment, you can tailor your system exactly how you want. For example, you may not want your system to be updated before your team can analyze how the changes could affect your operations.
You can set up a cloud-based system very quickly – a matter of days – and it doesn’t require the procurement of new hardware to work.
Cloud-based systems are also extremely easy to scale. After the initial setup, you can easily add or subtract users, cameras, servers, and clients with a click of a button.
For companies with smaller infrastructures, a hybrid deployment is nearly as quick as cloud-based – typically installing one preconfigured machine on-premise will do.
For companies with larger infrastructures – think 100+ cameras – much more processing power is required, which means more data centers and longer installation times.
Typically takes longer to implement. The main reason being that you may need to buy and install new hardware, like on-site data centers, in order for the system to work.
Data transfer from your cameras to your management interface will require more bandwidth than operating solely on a local network.
If you operate in a remote location and your connection is unstable, your analytics stream could be interrupted if you lose connection. We suggest that you discuss internally and with your vendor about setting up a “backup and disaster recovery” plan should your connection go down.
On the other hand, if you have a rock-solid connection, hosting your analytics in the cloud means you can access your system anytime from anywhere.
As long as your local network is up, analytics will continue to work. Alarms will be sent to local receivers, if any, such as a local VMS installation in a guard booth.
Alarms will also be buffered and sent to cloud receivers when the internet is up again. So, a 30-second internet outage will result in a 30-second delay in alarm receiving, but they will not be missed.
Here is where an on-premise installation shines. Your system is much more resilient to connection problems. So long as your local network is up and running, even if your internet connection goes down, your system won’t miss a beat.
A quick note on costs
A final consideration is whether a capital expenditure (CapEx) or operational expenditure (OpEx) payment method is best for your business.
On-premise systems typically require larger up-front payments (CapEx), whereas cloud-based systems tend to be monthly recurring (OpEx).
CapEx are very high in the short term, but OpEx can become more expensive depending on how long pay for your subscription.
But a recurring subscription is more predictable. You always know what you will pay each month and there shouldn’t be any hidden, extra costs.
You are responsible for maintaining your on-premise system. This means that you could run into unpredictable maintenance costs.
In a hybrid scenario, you will have a mixed payment model – CapEx for your the edge devices you need on-premise and an OpEx recurring payment for your cloud hosting.
Written by Christopher Jacklin
With the ever-growing power of AI, countless hours of video footage are now a wealth of actionable data. However, if put in the wrong hands, this data can also be used for a variety of purposes that extend far beyond safety, security, or efficiency.
Camera-based fire and smoke detection can enable early fire notifications in environments where traditional sensors struggle. By sending alarms as soon as flames or smoke is detected this type of technology provides an extra layer of protection for people and valuable assets.
By letting IRIS™ tirelessly sift through thousands of hours of streaming video in real time operators can focus their attention on the mere seconds of truly important video. Each suspicious activity is ...