IT Infrastructure and platform spending has always been tied to capital expenses until at the start of this decade. These are the sunk costs which every IT department must plan and justify the budget for those big-ticket purchases, and the IT department plans to do in the next calendar year.
With the prevalence of public cloud infrastructure, managing company’s IT costs needs a different perspective. Gone are the days of putting most of the company’s IT expenditures as capex but a new costing model must be considered for managing costs. With the use of cloud infrastructure and platform, budgets have shifted from capital expenses to operational expenses and the IT department must ensure that the running cloud infrastructure is optimally configured and well designed to avoid unnecessary added costs addition to operations.
Costs optimizations follow the same principles amongst public could providers. There are best practices to follow, tools to use, and other cost saving ways to consider. Microsoft Azure has no difference but aside from those common practices and recommendations, Microsoft Azure offers more options for further savings.
1. Know what you need, don’t pay for anything more
Application testing and profiling are the key to have an optimized Azure environment.
In the past, the tendency is always buying the more powerful servers and resources for your data center due to many factors such as cost difference and future resources needs. With Azure, you don’t need to pay for what you need. Save money by only buying the instance and the right resources that you need.
2. Running high traffic volume websites? Use a Content Delivery Network (CDN)
It may not be applicable to all but if you are running high volume websites with a lot of static contents such as files, images and pages, caching these contents not only make sense on the performance perspective but also on the monthly bill you get from Azure. Azure Egress traffic is one of the key variable cost contributors in Azure and using a CDN to handle your website traffic will help you minimize this cost.
3. Do you have tons of legacy data being archived? Consider the Azure Archive Storage Tier
Azure Storage can be created with the following three (3) Categories:
- General-Purpose V1 (GPv1)
- General-Purpose V2 (GPv2)
- Blob Storage
With GPv2 tiering, you can have flexibility on that storage to use, depending on how frequent you access data. For archived data that is rarely accessed, Azure Archive Storage Tier is the logical choice due to its huge savings benefits.
4. Use Burstable Virtual Machines
Microsoft Azure introduced the B-Series Virtual Machines for this purpose. For workloads like webservers, small or reporting databases and development environment, they are not necessary to keep the same level of CPU performance all through out. These workloads will run for long time idle or only a small fraction of the CPU then spike to full power due to a scheduled or incoming process.
The B-Series offers a cost-effective way to deploy these workloads that do not need the full performance of the CPU continuously and burst in their performance. While B-Series VMs are running in the low-points and not fully utilizing the baseline performance of the CPU, your VM instance builds up credits which in it can use to burst the usage up to 100% of the vCPU.
5. Auto-Scaling for your heavy fluctuating workloads
Don’t pay for machines you are not using. But what to have if you have a heavy fluctuating traffic?
Use auto-scaling to scale down some of your virtual machines at night or on the weekend when customers or users are not around, then have them ready to go on Monday morning when you open shop.
6. Monitor, Analyze, Optimize
Use the right tools to know how much you are spending and what you are spending them on. Use the Azure Billing Portal or Azure cost management (Cloudyn). Monitor your utilizations over time to see if you are using the right sizing.
Once all optimization options have been considered on the infrastructure and platform aspects, don’t stop there! Azure has a lot of options to optimize your applications through re-platforming and re-architecting initiatives. Take for example to use containers or serverless apps which will save you not only on the Virtual Machine costs but also on your VM infrastructure maintenance.
Azure functions, Azure Cosmos DB, Azure SQL Manage Instance and Logic Apps are just a few examples of Azure serverless computing.
Unique to Azure Options
7. Reserved Instance with hybrid benefits
Reserved Instances provide a way for organizations to pre-purchase specific Azure virtual machines, on a 1- or 3-year basis, for significant discounts. These work particularly well for long term, stable workloads where the size of the VM can remain static over time.
Get much more saving on your infrastructure with Hybrid Benefits. Hybrid benefits allow Volume Licenses with Software Assurance to be used to cover Windows Server virtual machines in Azure, saving up to 80% over regular Azure costs.
Fig. 1 Azure Reserved Instance and Azure Hybrid Benefit
Server Subscriptions
This is a game changing new offer from Microsoft and exclusive to Cloud Solutions Provider (CSP). One key inclusion of Server Subscription is the Azure Hybrid Benefits, which provides customers significant savings (up to 80% percent when combined with Azure RI). Previously, Hybrid benefits can only be enjoyed by customers with Software assurance but which not all small business or even mid-market customers have.
By bringing Azure Reservations and Server Subscriptions to the Cloud Solutions Provider (CSP) program, partners now will be able to address the cost concerns of moving to cloud especially with those predictable and persistent cloud workloads.
Moving a company’s infrastructure and workloads to any public cloud providers must be carefully planned and studied. The return of investment (ROI) comparison and scenarios related to the organizations services and applications must be analyzed to come up with the best solutions for your company needs. At the end of the day, decision makers must look at the merits of performance, stability, continuity and cost of operations before making the transition to cloud.
