How do you optimize cloud, mobile and distributed systems?

It's crucial to begin by business goals and immediate requirements identification. Collecting some aspects such as cost, performance, and potential expansions. This deeper understanding will serve as the foundation upon which you can build your technical strategies. To gain insight into business objectives, conducting interviews and engaging with key stakeholders and get users feedback is imperative, also aligning it with larger company mission. Surveys or design thinking sessions can provide invaluable insights Moving to technical evaluation, it becomes essential to have observability solution enables the extraction of critical information such as system topology, performance metrics, and main issues empowering with data-driven decisions

Many IT leaders approach this task by trying to fix what they already have, or creating a map of high value targets to focus on without ever putting anything down in writing as it relates to high level goals or objectives. This is due in part to fear that they might not get it quite right the first time through or the strategy may not achieve what we set out. But as long as you recognize that your strategy is fluid you'll be fine and may make some mods!!

Secure your cloud so it doesn't rain. When using mobile apps, don't rely on geoip as everything is based somewhere. If you run your application from a bottle then the bottleneck is obvious.

optimize depends on need , such as vm on azure .we optimized such thing depend on usage, performance and cost, for that we can use azure advisor as example , it is one of the free tool which help organization to optimize their infrastructure. it help to identify which resouces is under utilization or over utlization, according to that we can opimze the resource.

Again it is important to understand the business and, on this case, evaluating the workload types for all solutions while also categorizing their criticality We can use Observability to help us comprehend these aspects when the workload already exist and correlate with business data to determine criticality This enables us to establish the required resilience levels, performance criteria, SLAs, business continuity, and security for each solution With these requirements, we can design the architecture. This involves selecting hyperscale providers, determining the need for multi-cloud or hybrid cloud solutions, and determine the most suitable approach for each workload, whether it's traditional VMs, microservices, serverless, and so forth

Best practices are a whole school on their own. Here are a few more: Scalability: Design systems to be scalable so they can automatically handle increased workloads and traffic as the user base grows. High Availability: Distributed systems should be designed to fail without service disruptions. Microservices Architecture: Use microservices for better manageability, scalability, and flexibility. API Design: Create well-defined and versioned secure APIs to ensure seamless communication between different components of your system. Monitoring and Logging: Implement comprehensive monitoring and logging solutions to gain visibility into system performance and troubleshoot issues quickly. The list goes one and on!

Proper monitoring is essential. Ongoing optimization beyond initial is absolutely KEY. Routinely analyze your metrics and define/tweak sane baselines for your detections/rules. Be sure to tailor and tweak alert thresholds to your specific workloads. This will greatly cut down on false positives and increase the accuracy of metrics collected. Metrics can then be used to train ML to watch for "anomalies", subtle differences in the values of metrics that deviate from "normal" pattern. Extremely useful in troubleshooting. For instance, services that aren't fully failed, so not triggering an alarm, but no longer responding. Anomalies are likely going to be triggered by traffic level fluctuations, concurrent connection drop, etc.

Continuous Improvement needs to be in your DNA. If it's not there, make a plan to build it in as soon as you can. Technology moves too fast and IT has too many variables to get comfortable on a job well done. Once you're done celebrating your wins, get back to improving. Monitoring is a wonderful tool for finding things that are broken -- but it's also a great way to improve services before a breakage occurs.

Fostering a culture of collaboration and feedback among development teams is paramount. Encouraging developers, operations teams, and other stakeholders to share insights and ideas for improvement can lead to innovative solutions. Agile and DevOps practices can be precious in this regard, facilitating the rapid deployment of enhancements and the ability to respond to changing requirements. Continuous improvement should not be limited to the initial design phase, nor should it be bolted on at the end; it should be an ongoing, integral part of the system's lifecycle.