By Don Thomas Jacob
Maintaining web application and website performance is critical for a business because any unplanned downtime can lead to lost productivity, dissatisfied customers and lost revenue. This is why delivering a seamless online user experience is critical during peak shopping periods. So with the holidays fast approaching, IT admins are on heightened alert to ensure the business’ online presence is uninterrupted and can handle expected spike in traffic.
Here are some tips to help retailers prime their systems for the very serious silly season:
Test the impact of changes to a website from the end-user's perspective before the big event. A lot of testing takes place from the developer’s side, but some issues won’t manifest themselves until you put yourself in the end-user’s shoes – or, indeed, get a layperson in your organisation to stress-test the system. Set alerts and establish normal performance baselines then monitor each crucial step of transactions to identify problems. Once done, monitor from multiple locations to make sure all are performing optimally.
Ensure adequate capacity for influx of user requests. Communication among all stakeholders is essential. Talk with other internal teams, such as marketing, to anticipate challenges from any planned activities that may drive a spike, such as customer offers or planned advertising and media coverage of a particular product. Then use monitoring systems' historical data to gauge the capacity of your network (for example, conduct 95th percentile load planning) and allocate sufficient buffer, or the ability to scale, in preparing for those likely peaks.
Monitor availability and performance of the supporting infrastructure. Monitor everything you can, including shared storage, visualisation infrastructure, database, firewall and core network components and the WAN links to the outside world. Utilise IT staff or hosting providers to ensure any problems identified by monitoring solutions are quickly resolved; and support them with rigorous automation which can either fix minor problems, or buy your team enough time to solve larger ones without dramatically affecting performance.
Make sure your suppliers are one dial away. If you are reliant on any outsourced services (like bandwidth, IT support) or your online shopping website is hosted via an online or cloud service, make sure that you have informed your vendor of your expected spike in traffic. It’s also crucial to have a direct contact in case any issues arise, as well as an awareness of the supplier’s ability and costs in scaling up and down. And do your research to ensure the third-party’s technology meets their promise: while SLAs may guarantee you some level of financial recompense in the event of an outage, they’ll do little to repair the damage to consumer loyalty and reputation which may occur.
Maximise your bandwidth. Ensure that the bandwidth that connects between your consumers and your online IT resources are uninterrupted by day-to-day business operations, such as third-party updates. Once again, this is an issue best solved by scaling and monitoring: if you can track your traffic across the network, and rapidly provision additional bandwidth if you see a spike in progress, you’ll be able to ensure a seamless experience for both your customers and staff on the day.
Don’t skimp on security. While collecting customer details, ensure that you understand your responsibilities when collecting and storing this information. With recent password breaches in 2013 of Adobe, LinkedIn and LivingSocial’s databases, consumer confidence in privacy and protection is low. Combining fundamental security measures such as a comprehensive security information and event management (SIEM) tool with a structured approach to storing data will allow IT admins to more easily identify and track access to sensitive information.
Have an offline backup. This is perhaps the most important point. If your system is down at any time, what other ways can you still capture order? Make sure that the transition between primary and backup systems (or “failovers”) is automated so that you minimise downtime for users.
Rigorous network monitoring coupled with the ability to rapidly scale your resources – and topped with a good dose of foresight – will put your business in the “nice” category for customers.
Don Thomas Jacob is the head geek of SolarWinds, an IT management software provider.