If the cloud is not already part of your current IT initiatives, it is definitively coming your way. According to Gartner, the public cloud services market is forecast to grow 18.5% in 2013 and total a whooping $131 billion worldwide. As you get ready for the cloud, or already running there in full speed, here are 5 things you should do to ensure peak performance of your cloud-based applications.

DO #1: Don't rush

Many organizations begin their cloud journey with no clear requirements or measurable objectives — and often as a reaction to demands placed by the business. Huge mistake! Slow down; take the time to put together a Cloud Overview document containing goals and objective success criteria, constraints and high-level risk. Then review it and get consensus with all your stakeholders (sponsoring, business and technical). A couple of quick pointers for you.

• Carefully assess which applications to move to the cloud. Some apps may not work with cloud services due to their requirements, configurations or use patterns. Determine the value of different cloud services and map these capabilities to the needs of the apps in your portfolio. For example, the EU (European Union) Data Protection Act, mandates to keep personal information within the EU. Therefore, if you have European customers, select a cloud vendor that will keep European customer data on servers located in Europe.
• Consider a phased-approach -- don’t move all applications at once: Less is more. Start small, moving less mission-critical apps first to make sure that you get it right, prior to migrating other business-critical applications and data.

DO #2: Triage your applications before you move them. Bottlenecks, memory leaks, underperforming SQL queries and other problems will intensify, and it will be much harder to identify the root cause of problems later on (a cloud-based architecture adds more potential failure points to the Web application delivery chain). Instead, fine-tune applications in your datacenter now (get rid of memory leaks, potential deadlocks, etc.) before moving to the cloud.

DO #3: Think about your Application Performance Monitoring strategy from the get go. Whether you are planning to use an Infrastructure as a Service (IaaS) or Platform as a Service (PaaS) model, there is an added layer of complexity and management challenges. Problems could happen now in your custom-built app code, in your datacenter/cloud provider infrastructure, within a 3rd-party Webpage component (CRM integration, Social Media widgets, new feds, etc) and so on. Therefore build a comprehensive application performance monitoring strategy that covers all your bases, including:

• Datacenter and cloud infrastructure performance monitoring. For example Web server, mail servers, application servers, DNS servers, EC2 & RDS instances, VMware vSphere host servers and guest virtual machines (VMs) and so on
• Connectivity with back-end databases. Oversee database response time and throughput as well as database operations by caller (item, inventory, category, etc).
• Application performance (response time, throughput, error rates) as well as transaction traces (to identify application code problems such as underperforming methods, or hanging threads)
• End-to-end transactions. Oversee key Web functions and critical multi-step transactions (e.g. new app code, marketing registration forms, shopping carts, Salesforce record retrieval, HR functions, etc) on a 24x7x365 basis. Monitor, as well, performance of 3rd-party components and services (marketing analytics, social media feeds, webpage forms, ad servers..)- Their failure can also halt your business and be very visible to your customers.
• Customers SLAs. Traditional metrics like average response time can be significantly distorted by just a few long responses. Monitor as well Apdex scores (user satisfaction levels with your applications – “Satisfied,” “Tolerated” or “Frustrated”) – to immediately detect problems and turn the situation around before it is too late

In addition, look for a multi-platform solution that supports all popular languages and frameworks (Ruby, PHP, Python, Java, .NET…). Don't focus only on the development framework you use today, but think about future development plans as well.

DO #4: Pay attention to both internal and external customers. When your public website underperforms, your visitors leave to a competitor. When internal applications are slow, complaints and escalations will increase on the inside. In either case, upper management and your boss will not be happy, so strive to assure both internal and external customer satisfaction.

DO #5: Make performance optimization an integral part of your culture. Speed is crucial to your business success. Did you know that 75% of online shoppers who experience a site that freezes or crashes, is too slow to render, or involves a convoluted checkout process would no longer buy from that site (Akamai sponsored research)? If your applications are slow, your customers will flock your competitors’ websites. Don't use your cloud migration initiatives as an excuse to forget your competitive benchmark efforts as you settle-in with your new cloud provider. Continuously keep an eye on your performance (and your competitors' performance) before and after you move to the cloud.

Whether you are researching the cloud, planning a cloud migration, in the midst of a migration or fully running in the cloud, our latest Site24x7 cloud monitoring release has you covered. Check it out!