Case Studies

LamaLoLi Partners with AllCloud: A State-of-the-Art Cloud Migration Case


LamaLoLi is a European e-commerce leader in licensed children’s apparel and merchandise, selling more than 60 mega hero brands such as Disney, Mattel and Hasbro.

The company was founded in 2007. It operates from Germany and serves France, the UK, Spain, Poland and 20 other European countries. The company’s vertical retail model goes from design, to high-quality manufacturing right through to sales. It has a broad online presence, including its own e-commerce site, as well as integration with a multitude of leading online marketplaces.


Smooth Cloud Migration

A healthy e-commerce business directly depends upon its website’s uptime, as every small amount of downtime translates into direct loss of sales. What’s more, it also makes a negative impact on marketing activity, such as harming the site’s online marketing campaigns including organic search, Google Ads, Facebook, TV and affiliate marketing programs. Not only that, but another challenge LamaLoLi faced was transferring to a complete new website domain while maintaining the site’s exact page structure, but with a new backend configuration. This included the different system tiers in order to maintain both the end user experience and integration with the different marketplaces.

Traffic Peaks

One of the major investments of every online retailer is in marketing. LamaLoLi often sees traffic peaks when running a campaign such as a TV commercial and may need to cope with up to 10 times more traffic than usual. While working in a colocation arrangement LamaLoLi didn’t have the confidence that its rigid environment could support both traffic spikes and their long-term business expansion plans.

Page Loading Time

In addition to the need to maintain the site and service uptime, LamaLoLi is sensitive to page loading times. Every page has its own browsing and processing requirements. For example, the product category pages make multiple queries and are at higher risk of slower loading times. By failing to build and manage the infrastructure to make their complex application stack perform, LamaLoLi not only risk their end user experience but also their marketplace presence. For example, repeated failure to complete Amazon Marketplace requests may not meet required service levels. This could result in account suspension and loss of revenue to the tune of millions of dollars a year.

Continuous Optimization

Last but not least, moving on with their decision to move to AWS, the LamaLoLi management team needed to acquire the skills and knowledge not only to migrate their complex stack to the cloud but also to manage it. What’s more, LamaLoLi were presented with new challenges in terms of cost and capacity control, which had completely changed by moving from steady colocated IT to an elastic cloud environment. They were faced with a bewildering number of different Amazon cloud portfolio options, such as instance types and costs. It was also important they avoid exceeding their IT budgets and achieve the cloud efficiency they were looking for.



The Design Phase

By leveraging its Cloud Architecture Workshop, AllCloud experts were able to get up to speed with the knowledge required to migrate all of LamaLoLi’s IT environments. Working closely with LamaLoLi, the AllCloud team developed a deep understanding of the stack and, in particular, the synergy between the application and the underlying physical resources. Analyzing the environmental requirements and potential points of failure, AllCloud’s cloud architects came up with a complete redesign of the infrastructure, which was tailored to the Amazon Cloud.

Using Beanstalk and RDS

All LamaLoLi services were deployed to AWS Elastic Beanstalk, which created a separation between the application environment and the underlying infrastructure. Leveraging AWS Beanstalk, AllCloud’s experts facilitated instance provisioning, scaling and load balancing. This enabled streamlining of each system service from migration to Amazon right through to management.

As part of the migration processes AllCloud also rebuilt the whole dev/test environment and brought in new tools to enhance LamaLoLi’s automated DevOps environment, which also supported the smooth migration. Over an intense two months, AllCLoud managed to completely migrate LamaLoLi’s systems to a well-structured AWS environment.


High Availability and Performance

In order to enhance environment uptime, AllCloud experts used AWS RDS, leveraging its inherent self-healing mechanisms to protect the migrated databases. The AllCloud team implemented application and database replication processes across two AWS Availability Zones (AZs) in order to support a secondary DR (disaster recovery) site. In addition, web tier instances were provisioned with Beanstalk’s native autoscaling capability so, in case of a demand peak, the system could automatically scale and maintain its performance and end user experience.

24×7 Monitoring and Optimization

As part of the migration and outsourcing of their cloud operation, LamaLoLi used AllCloud’s 24×7 NOC-as-a-Service to run their IT support tiers. An important part of the migration was to integrate the tools to continuously monitor and log LamaLoLi’s cloud. Event log data is stored within an isolated data store and processed with AWS Elasticsearch for fast data analysis. If an event occurs, a notification is generated to AllCloud’s NOC team to inspect and remediate the issue.

AllCloud’s NOC team has complete access to the environment and is responsible for detecting the root cause of issues and acting according to predefined protocols. If it’s an application issue it will communicate it directly to the LamaLoLi R&D team. What’s more, to maintain complete visibility over the dynamic environment, AllCloud uses the DataDog proprietary monitoring-as-a-service solution.
Lastly, one of AllCloud’s most important responsibilities is to monitor and control usage and utilization and continuously optimize the LamaLoLi cloud footprint. This includes resizing resources and purchasing EC2 Reserved Instances (RIs) in order to meet performance and demand as well as predefined budgets.