Most software companies’ budget got stifled by the charges on dedicated server. So they are in need of solution that would relieve their finance resources from draining, but also include the same stable and supported infrastructure capability. Many CTOs and developers have already found such leverage for their projects by using PaaS.
Let's have a look at the case of Shiply, a British company that have migrated from dedicated servers to Jelastic PaaS empowered by Layershift hosting service provider. Their founder and CEO, Robert Matthams shared details how this shift helped to fix their hosting cost leak and fully satisfy their cloud requirements, as well as why Shiply decided to migrate and the benefits they got afterwards.
Shiply is an online transport marketplace. The customers list goods they need to move around the world, and Shiply provides a network of about a hundred thousand transport companies who bid for the work, as a way of keeping their trucks full along routes they're already doing. It's great profit for the both sides, for logistic companies because they are already making their journey and get additional orders for their part-full lorries, and for the customers because they are able to catch an offer with a low price. It's basically like a reverse auction format.
What platform did you use before Jelastic?
We've been with Layershift since we launched in 2008. In 2010 or so we saw that cloud hosting had been getting important, and Layershift didn't offer the platform at that time (but we really liked Layershift and we’ve been working with them since day one). They offer great customer service with a real personal touch.
In our first year when we were just getting started, Layershift would even jump in and helped us out with code. Which you don't get much these days from a big cloud provider, like AWS, they are just not able to help out on that level. As Layershift had that personal touch we really didn't want to leave them, but at the same time we realized we were paying a lot for the dedicated servers rather than for a cloud solution. So when Layershift announced they were going to offer Jelastic, it was great news for us and we were quick to make the switch. And we've been happy Jelastic PaaS customers since 2014 🙂
What were the main challenges that made you migrate from dedicated host to Jelastic and how did you overcome them?
We actually ended up running the site across two domains. Before the full migration, we ran Shiply on both - our dedicated boxes, as well as on the Jelastic environment. The traffic was running through each, giving us the confidence that it would handle the load OK on the new platform which it did. After that we migrated fully over to Jelastic.
I guess the only real challenge we were potentially worried about was MySQL. And Layershift was concerned about how well Jelastic would handle a large database in production, and at the end of the day it handled traffic just as well as the dedicated box did. We were happy with the performance, we had some initial concerns, but they were dealt with just fine.
So what was the key reason of migrating to Jelastic?
Could you please describe what topologies do you use?
We have several environments in production. The main technology we use are: PHP, NGINX, Apache, MySQL and Memcached.
Do you use multiple Apache nodes with the scaling feature to handle traffic spikes?
Yeah, we did. There were some occasions a few years back, where we had some TV and radio appearances and Jelastic gave us the possibility to prepare for the traffic spikes. When we had an initial PR and launched in 2008, we were running on dedicated boxes. We went onto something in the UK called the Chris Evans BBC Radio 2 Show which is listened to by several million people, and the spike just brought the site down because we weren't prepared for it. The problem was that stable cloud solutions didn't really exist back then. We learnt from mistakes and when a bit later we worked with PR (TV, radio) we used Jelastic to scale up, being sure that this site could handle the extra load well and not go down.
Do you have a large database with many gigabytes of data?
Yeah, it’s many millions of rows now, it's quite a big database.
And what about backup?
We have an off-site disaster recovery solution, just in case of the worst we can shift over and get the site back up and running with minimal impact for our users.
Are you using some features like collaboration tool?
Yeah, we have one senior developer who tends to do it, but equally other developers can jump into the environments from time to time in case of necessity, and it is very convenient with the collaboration feature.
Layershift has several hosting regions. Do you plan to use its locations in the the USA and Singapore in addition to the UK data center?
We`ve already started thinking about it. Just from redundancy point of view, if a service go down in Manchester, the ones in the US would be able to handle the traffic. So yeah, it's on our roadmap.
What factors or features do you like in Jelastic?
Cost is important, but uptime is even more important. It's worth mentioning that we've had no major downtime since we moved to Jelastic platform on Layershift. We used to have sporadic downtime on the dedicated boxes, whereas, Jelastic has been very very stable. What is also important is that the maintenance windows have always been during off-peak hours.
The other thing we like about Jelastic - it is intuitive and very easy to use, so the need to consult the documentation occurs very rarely. If any of our developers go in and need to change something, there is no necessity of deep learning - it's pretty quick to see where things are and how to apply changes.
A good quality of product is always highly evaluated among consumers, and this use case is only proving it. This is the perfect case when a combination of Jelastic PaaS and Layershift cloud hosting provide scalable, intuitive and cost-efficient solution with high-level customer support.
Struggling to choose a perfect cloud for development and hosting of your projects? Test drive Jelastic for free and feel a significant difference in cost, performance, and support.