Demystifying SAS Cloud Migration

You’ve probably considered running your SAS implementation in the cloud. If you’re still on the fence, let’s settle this once and for all. Don’t let fear of the unknown hold you back from harnessing innovative technology that could propel your organisation further, faster. This article discusses common misconceptions about cloudification for your SAS workload. Read on to set the record straight.


Myth 1: “It’s not broken now, so I don’t need to fix it.”

Truth 1: “Organisations should be future-ready.


If your company enjoys the benefits of a stable and highly performance SAS data analytics platform, that is an excellent place to be. However, success in the past does not guarantee success moving forward. With regular software updates, patches, and upgrades, the average lifespan of devices is between 3 to 6 years. For high-performance servers, the lifespan can increase to 10-15 years. Even so, it would serve your company well to think beyond that.

One of the emerging business management trends that have everyone talking is Big Data. Increasing data sources from innovative sensors, improved storage, and development in processing algorithms have fast-tracked this technology’s adoption.

Big Data may require more compute and storage capacities than your existing on-premise infrastructure may be able to handle. Cloud-based solutions fill this gap. Most companies have already relocated their data and have taken advantage of the cloud’s virtually unlimited storage. By moving their processing and analysis closer to their data, handling is made more efficient.

When your organisation upgrades to cloud-native SAS Viya 4, it gains the ability to analyse rich media and unstructured data including videos, audio files, sensor data, and many new formats. These capabilities, placed in the hands of your data scientists, give you an edge over your competitors.

Another emerging trend that your organisation should embrace is decentralised workspaces. Whether your staff are working from home or collaborating with different satellite offices, or collaborating with organisations worldwide, moving to the cloud creates more seamless interaction within your team.


Myth 2: “I don’t want to be locked into a subscription model.”

Truth 2: “The cloud is so scalable and elastic that you can move to a different cloud or back to an on-premise configuration in no time.


The cloud is very friendly to the cautious. If a solution does not suit your use case, it is relatively easy to find another platform that works for your requirements or to revert to an on-premise solution, if necessary. Because cloud providers offer highly scalable services, you can try a full-scale proof-of-concept before committing to it. In contrast, on-premise servers need to be purchased before the testing of a scalable proof-of-concept can start.

If you aren’t happy with your cloud implementation performance, it’s just as easy to quit as it is to begin.

In fact, cloud services are so scalable and so elastic that it’s possible to schedule additional resources only when your SAS batch process demands it. During off-peak usage, your server capacities can be reduced to minimum requirements. Ironically, it’s on-premise servers that lock you in with unused infrastructure capacity, as seen in the figure below.

Scripts can automate your SAS batch processes on selected days or hours. If you aren’t using the services, you need not be charged for it, unlike an on-premise server that can’t be so easily scaled.


Myth 3: “More cloudification projects fail than succeed.”

Truth 3: “Projects fail largely because of poor project management.”


In their 2019 Cloud Success Barometer, Unisys shares that 1 out of 3 cloud migration projects fail. Their findings reveal that companies which don’t integrate cloud adoption plans into their core business transformation strategy are likely to fail.

Infoworld reports the top causes for such a high failure rate. They are:

If you think about it, these are issues that good project management can solve. Many of the issues mentioned can be avoided by bringing in a partner with years of cloud migration experience.

Many stakeholders may not understand how a cloud migration affects their business units until they see the new solution. As such, proofs-of-concept by a partner can be very advantageous. Risks can be validated, and proper mitigation strategies can be put in place before it becomes a problem. Proofs-of-concept makes sure that everyone affected by the project are on the same page. It creates organisation-wide buy-in even before the project begins. With the users prepared for what to expect, business units can incorporate changes into their current operational procedures and increase user acceptance early-on.

Working with a partner that is certified both by your cloud service provider and analytics platform is a huge advantage. Partners like Analytium that are certified partners and resellers of AWS, Microsoft Azure and SAS, help you realise proofs-of-concept with minimal financial outlay. Analytium can arrange for free AWS and Azure credits as your organisation explores the cloudification of your SAS workload. In addition, Analytium can also identify the nuances of integrating your SAS platform into your cloud of choice.


Myth 4: “I’m wasting my current infrastructure investment.”

Truth 4: “There are ways to redirect your investment.”


Managers often face the challenge of justifying cloudification. Stakeholders often don’t appreciate the differences of a cloud-based data analytics platform versus an on-premise one. To many, it may seem redundant and unnecessary.  The question that is often asked is, “What do I do with my on-premise analytics infrastructure now?”

Moving your SAS workload to the cloud does not necessarily equate to wasting your prior investments. Much of the hardware can be repurposed to augment existing operational or transactional systems that may benefit from added resources.

Further, unneeded hardware can be disposed of in a secondary market. Buyers can supplement their on-premise systems with second-hand units. Of course, as the hardware ages, the value goes down. In which case, delays can reduce the amount that is recovered.

Again, working with a partner can help you with this as well. Your partner may have other clients or have the network to know which organisations are in the market for second-hand hardware.


Myth 5: “It costs more long term.”

Truth 5: “Your Total Cost of Ownership could be less than you think.”


To a Finance Manager, choosing between purchasing and leasing is straightforward. If the total leased amount for the expected lifetime of the purchase exceeds the purchased amount, it is more favourable to buy than to lease, from an accountant’s perspective at least. For example, if an on-premise server that is expected to last 36 months costs £36.000 and a cloud subscription costs £1.200 per month, or £43.200 for the same period, from the view of an accountant, an on-premise server makes more sense.

What the finance department does not see are the services that are implicitly purchased through a cloud subscription. By migrating to the cloud, you are sure that your servers are constantly upgraded. Every time that you start up an instance of your server, it’s as if you are using a new machine. When moving to the cloud, maintenance services are also procured as part of your subscription. Instead of hiring an administrator to maintain the hardware and network infrastructure, these tasks are effectively outsourced to your cloud service provider. Without you knowing it, you free up your organisation’s precious IT resources, allowing them to focus on more mission-critical tasks. It goes without saying that the staff needed to manage these resources are also removed.

By migrating to the cloud, you also procure resiliency that keep your data and compute capabilities available. When one data centre fails, another one can pick up where the failed one left off, as stipulated in the service level agreement with your cloud service provider.

In short, the subscription amount should be compared with the total cost of ownership. Often, with all the benefits tacked on to your cloud subscription, on-premise infrastructure costs the same, if not more.

If that isn’t enough, cloud service providers offer many options to reduce your cloud costs. Reserving your servers and storage can lead to an average of 35% or up to 64% savings. Moreover, selected cloud providers such as AWS have discounted prepaid schemes that allow you to bring down your cloud costs even further.

Carrying costs are also negligible because certain cloud providers such as AWS allow you to sell back your purchases to other tenants via an open marketplace. Azure, on the other hand, allows you to return unused services without any penalties for expenditures up to 50k USD.


Myth 6: “My employees don’t have the necessary skillset to adjust to something new.”

Truth 6: “Moving SAS to the cloud means minimal disruptions. In cases when you need to augment your team, help is never too far away.”


Cloudifying SAS means that your team has access to the same functions that they are familiar with. Your cloud implementation should feel like it’s on-premise.

If there are skill gaps that need to be filled, individuals from a large community of developers are always willing to step in.

Some organisations feel more secure with having a partner to guide them along their cloudification journey. These partners offer a variety of outsourcing modes to meet your requirements:


Myth 7: “It’s not safe.”

Truth 7: “Cloud service providers can secure your data and environment better than you can.”


Many organisations have justified concerns about the security of a public cloud.

For a public cloud that is directly accessible through an internet connection, cybersecurity firm, Checkpoint lists the most rampant threats here:

Misconfiguration and insecure interfaces should be covered by better platform design. Unauthorised access and hijacking of accounts can be controlled with other procedural measures implemented with shared responsibility between your organisation and your cloud service provider.

Other best practices include encryption of data, creating strong passwords, and implementing multi-factor authentication.

Having infrastructure physically on-site can give organisations a false sense of security. In fact, a survey conducted by the Digital Guardian reveals that inside security threats are more imminent than outsider threats. Removing physical access to malicious parties can reduce risks to a certain degree. According to RapidScale, 94% of businesses reported increased security after transferring data to the cloud.

In reality, a public cloud can be more secure than on-premise infrastructure. Cloud service providers have the ability to maintain Tier IV Data Centres (armed security patrols, 24/7 CCTV monitoring, biometric access to control entry points). In 2017, Microsoft pledged an annual investment of USD 1B in cybersecurity for its data centres. This figure is significantly higher than what most of other businesses can invest in cyber security on their own.


Myth 8: “My industry is highly-regulated, and I will lose control of certain components.”

Truth 8: “When done correctly, a public cloud can be secured to the equivalent of a private cloud.”


Selected industries such as healthcare, legal, manufacturing or financial services require more stringent security measures. The impact of data breaches in these industries can be catastrophic, which is why certain regulators demand the use of private clouds or data centres.

Public cloud service providers offer added security controls that allow a public cloud to appear as a private cloud. From the point of view of some regulators, public clouds secured with customer-managed encryption keys where hardware has been dedicated to a single customer are considered the equivalent of a private cloud.

Analytium has successfully migrated customers with the most sensitive P4 level data to the cloud using this strategy. Upon careful architecture reviews with the relevant regulators, the cloudification projects have been deemed successful.

Additionally, certain cloud providers offer security certifications or attestations of compliance that are recognised around the world by a wide array of industries.

If you think you are in a heavily regulated industry and the public cloud is not for you, think again.


Myth 9: “I won’t be able to access my data and system if the internet connection is down.”

Truth 9: “Your cloud is portable yet secure.”


Moving to the cloud will require a robust and fast connection to the internet. Depending on how your system is designed, your organisation will consume higher bandwidth. This is particularly true for systems where the data storage is separate from compute capabilities. In which case, data has to move up and down the Internet for it to be processed. If your SAS processing is located on the cloud along with your data store, you can expect a smaller bandwidth requirement and faster processing times.

Practically speaking, if the internet connection at your office fails, your team can move to a coffeeshop and access your cloud services from there. Mobile data connections may be sufficient since the connection you need to initiate a batch process or access results is not as large. Business can continue as usual, even in extreme circumstances.


Myth 10: “I don’t have the time to move to the cloud.”

Truth 10: “Cloud Migration does not mean you need to cease operations.”


One of the fears of organisations interested in cloudification is the timeline involved. According to TechRepublic, 73% of cloud migration projects take a year or longer.

Aside from cost implications, lengthy timelines of cloud migration projects often mean disruption in SAS – related business processes that many customers simply cannot afford. With a good change management policy, migration should have minimal effects on your existing data analytics workload.

No-freeze options are available and should be explored. Clients are free to access their data and tools with no downtime at all. During scheduled off-peak hours, metadata is compared, and only the items found to be different are copied over to the cloud. This process is repeated until the parallel environments are identical and the stakeholders accept the new system.

Experienced partners such as Analytium, who have completed numerous migration projects, have distilled best practices into workflows and tools that optimise the migration process. Proprietary software and solutions such as SAS9API, SAS9 HealthCheck and SAS Monitor have been developed by Analytium to partially automate the cloudification process. These tools have massively reduced migration timelines and chances of failure.  Analytium has delivered projects in as fast as two weeks, while the biggest to date project only took eight months.




As early as 2016, Forbes predicted that moving to the cloud is inevitable. In 2020, the IDG Cloud Computing Survey demonstrated how companies are continuing in the cloud migration trend. The survey reveals that cloud migration accounts for one-third of IT spending, and companies anticipate that more than half of their organisation will use cloud-based solutions.

Source: IDG Cloud Computing Survey.

Regardless of your company’s industry, scale or growth stage, you should be preparing for your move to the cloud, albeit with proper caution. Cloudification is not a matter of if but a matter of when and how. Moving to the cloud should bring value to your company. With the right partner giving you the right solution, the possibilities are endless.

If you’d like to hear about our customer successes or avail of a no-obligation discovery assessment, feel free to leave your contact details, and we’ll contact you.

The post Demystifying SAS Cloud Migration appeared first on MySASteam.

Let’s Get the Conversation Going

Leave your details below so we can help find the best solution for your organisation.

    Let’s Get the Conversation Going

    At MySASteam, we believe that every SMB deserves the Power to Know. Please leave your details below so we can find the best solution for your organisation. We’ll call you.