Website Woes and Launch Failures

HSE's free GP care Registration site crashes on launch
HSE’s free GP care Registration site crashes on launch

This is every web project manager’s worst nightmare. It making front page news of one of the country’s biggest online news portals is even worse. On Monday, the HSE launched a new site for Free GP Care Registration in Ireland, and due to overwhelming demand, the website could not handle the initial ‘rush through the door’.

Spill in Aisle 4

Leaving the ‘who’ aside here for a moment to look at the issue, this is not an uncommon event. A new website or online resource is launched whose aim is to engage people in a self-service/online manner to shift them away from footfall into offices and administration staff resources will always be faced with this issue. Anyone whose ever tried to buy a ticket for the hottest show coming soon in Ticketmaster knows exactly what this is like; the website becoming unresponsive, grinding to a halt or ‘falling over’.
In my time in the Cloud Services space, I’ve dealt with everything from online store launches to Dragon’s Den contestants appearing on the show, utility companies relaunching their customer management portals to financial institutions launching new portal sites. And often, the results are the same despite best advice. The website takes a pasting like a Conor McGregor opponent.

Why does this happen and do people allow it to happen?

Let’s be clear here. No-one deliberately wakes up to the launch of a new online resource where a significant financial investment has been made  and consciously says “Hey, who cares if it falls over as the boss looks on.” It is utterly naive to also assume there’s a level of carelessness about these kind of failures. It is similarly over-simplified to parlay it into online speak like ‘EPICFAIL’ or ‘n00b-move’.What in actuality has happened is down to usually one of two thing;
  1. Poor planning
  2. A lack of tools

The Million Dollar Question: ‘How to you stop it happening?’

Planning is key. It’s not enough to meticulously plan the content and functionality of your online resource. Nor is it enough to make sure you’ve enough compute, disk IO and network traffic availability. Most of the failure stems from the omission to include a plan to test these resources under load. This form of testing will demonstrate several things:
  1. The performance of your infrastructure design to handle the incoming flood/waves of requests for pages
  2. Whether your network can handle what could inadvertently become a by-product unwitting DoS attack on your network and compute resources due to the nature of the content being served
  3. How interactive portions of the site that process input (i.e. forms) perform while a mass of simple webpage requests are going on
  4. Where optimization under load is needed for the content being served.
  5. How your scaling technology or ability to scale on the fly works (if it works at all)
The second key is understanding that in today’s landscape of commodity compute services, there are tools available for this and services using these tools available.
Won’t somebody please think of the Children?
At Cogneto, we’ve been involved with large scale website launches where using our CTS product (a boutique part of our StillAlive Enterprise Services), we’ve handled the performance and load testing of interactive and passive parts of sites testing at various user loads; simultaneous waves of registrations, ramped page request loads across ‘hot’, ‘tepid’ and ‘cold’ content on the site (‘hot’ being the most likely hit content, with ‘cold’ being the most infrequently hit content), all of which was able to kick back great performance metrics and alerts on what was going on with resource requests from the application layer, and even allow the coordination of incident response alerts using our platform to the affected parties managing the various levels in the stack, i.e. IaaS, Network, Content Platform – with them all being able to co-ordinate their responses through the socialised experience Twist offers).

If you’ve an upcoming web project launch or an existing one you’ve concerns about, please get in contact with our Still Alive CTS Experience team today, and we can help you avoid the front-page woes.

Hello Cogneto-verse. We are here.

We are Cogneto. 

As a valued customer, I wanted to share some exciting news with you and to take this time to introduce ourselves. We are We have acquired Point DNSStillAliveStatusHub and from Copper Cloud Inc.. We are a team who understands developers and the online space, with the last decade spent working in the Cloud services space with everyone from start-ups to large multinationals and Government agencies across the globe.

This acquisition will have a positive impact for you, our customer as we bring our experience to shape, guide and make the services you use better. As part of this acquisition, your billing will now be paid to Cogneto and appear same on your payment methods with us. You will see this billing change during your next billing cycle.

All existing methods of support will remain valid, and we will be working to improve support services in the coming months to make these great services you’ve come to depend on even better.
If you use Twitter, please follow us @CognetoApps where we look forward to engaging with you, our customers more.

Thank you for your on-going support and we look forward to continuing to provide you with the best services, user experience and with an exciting line of new developments in the future for the services you love to use.

Ian Bergin Founder & General Manager