Azure Answered – 3 Azure tips to help you to reduce costs, increase performance and reduce risk
Being Australia’s #1 managed cloud business and having an 18-year relationship with Microsoft means having a Gold Cloud Platform status and an escalated support contract. We’ve learnt and delivered some on-ground valuable lessons, thanks to the many customers who shared their growing pains with us.
Azure Optimise, to quote the blurb, is a service that Macquarie Cloud Services offers to help customers overcome common operational challenges in Azure that inevitably come with scale. It’s particularly relevant given the current situation, where business are streamlining operations where possible.
Picture this – you’ve launched your Azure environment, it’s going great, and then you realise one (or more) of the following:
- There are 7 days to go to the end of the month and you’ve already blown your budget.
- Your reservations are not aligning with right-sizing recommendations.
- Your best practices are ranking way lower than your industry peers.
- Your developers have mysteriously set up PaaS services in Singapore, and you can’t really migrate across regions.
If this sounds like a predicament you’ve been in before, we’re not surprised. All of the above were real situations we were presented with during the previous Azure Optimise engagements with organisations.
Here are 3 simple takeaways that we found repeating itself across customers of all shapes and sizes.
- Keep an eye on utilisation of your Azure VM reservations, as you can’t accrue your reservation hours. If there are no VMs running in that instance family in hour 1 of the month, you lose that hour for good. Yes, you can change, cancel or increase scope, but you can never get that hour back. Pulling and monitoring under-utilisation data can make your committed payments worthwhile.
- A right-sizing strategy must be appropriate for the situation as well as the infrastructure – there is no one size fits all. Recommending right-sizing based on average CPU or DTU patterns alone may not give you accurate predictions. What’s right for Citrix won’t be right for web servers or for databases, and we haven’t even started on environment and lifecycle parameters yet. We’re very proud of the complex conditions that we can automate based on state, application, resource group name as well as the permutations and combinations of the raw metrics of CPU, memory, disk and network patterns. By the way, we did not find many over-utilised resources in the thousands that we went over – bearing out Flexera’s estimate that 35% of public cloud spend is wasted.
- Azure Policy is your friend. Governance methodologies are best started in Stage 3 of the Cloud Adoption Framework, and maintaining frameworks is an iterative process, and one that is never done. Many mature environments also benefit from establishing the practice in-flight, as the corporate exercise could prevent risks from becoming real threats. Changing or setting up corporate governance policies may result in having to remediate many non-compliant resources, and Azure Policy can not only help fix existing problems, it can help prevent new ones. It is the weapon of choice when we speak to the implementation of a pragmatic Cloud Center of Excellence.
Save money, reduce risk and drive accountability through Azure Optimise. Click here and leave your details to express interest in the consultation offered.