Goutham Upadhyay, Microsoft Cloud Solution Architect

Are you a Microsoft customer running your workload on Azure?

Holiday seasons like Black Friday, Double 11, and Christmas Sale will test your app’s limits, and so it’s time for your Infrastructure and Application teams to ensure that your platforms deliver when it is needed the most. Be it shopping applications on the web and mobile or payment gateways or banking systems supporting payments or inventory systems or billing systems – anything and everything associated with the shopping season should be prepared to face the load.

Application resilience is vital to keep up with the higher customer demand during the holiday season. Below are the top 10 considerations to ensure your App is resilient on Azure and to make sure your Technology team can handle your platform with grace.

Here are my top ten considerations for running your workload on Azure to handle this holiday season:

• Multi Region: Consider running in multiple Azure regions for resilience check our mission critical reference architecture. If running Active-Passive, then Perform DR drill in paired Azure region. If running in single region for whatever reason please consider running your workload across multiple availability zones for better resilience.

• Load Test: Perform load testing preferably use fully managed service like Azure load test it is easy to generate high-scale load and identify app performance bottlenecks with Azure load tests.

• Right Sizing: Ensure you have tested if your current instance size/count of different Azure components in your solution can handle the load, if not, please re-size well ahead of time. Baseline your Azure resources to support peak. Don’t just fully rely on Auto scale as scaling your infrastructure to extreme spikes will still take some time.

• Test for reliability: using Azure chaos studio a method of experimenting with controlled fault injection against your applications. Review this reliability checklist for more exhaustive set of reliability guidance under well architected framework (WAF) and best practices.

• Azure Advisor:  Advisor is a personalized cloud consultant that helps you follow best practices to optimize your Azure deployments.

• Quota: Review the Quota for the resources you are using in your subscriptions. Quotas can be adjustable or non-adjustable. Adjustable quotas for which you can request quota increases fall into this category. Each subscription has a default quota value for each quota. You can request an increase for an adjustable quota from the Azure Home My quotas page, providing an amount or usage percentage and submitting it directly. This is the quickest way to increase quotas.

• Supported version: Be on supported version of the service ex: If you are running Azure Kubernetes Service upgrade to supported AKS versions ideally 1.26 or even higher. This will make sure even if you run into issues Microsoft support teams can help you.

• Service Health: Configure service health advisory and service issues alerts. This is the only way Azure communicating back and most often missing to act on key alerts/notifications can be costlier.

• Monitoring for reliability: Get an overall picture of application health. If something fails, you need to know that it failed, when it failed, and why. Azure Monitor is a comprehensive monitoring solution for collecting, analyzing, and responding to monitoring data from your cloud environments.

• Review Reliability: Across all the services on Azure via Reliability Workbook key tabs to look at is Availability zones and capacity for each services that you care about. This is the newest and most cool feature that I personally found useful to review against some common mistakes.

• DDOS: Holiday season is DDOS Season as well. Ensure all key security measures are also taken care along with DDOS

Obviously, this list is not fully exhaustive list, but I’ve tried to capture few top ones to make sure you are setup for success running on Azure this holiday season. Please let me know what you think.

Source from Microsoft