Trove sought to replace its dynamic, server-dependant application architecture on AWS with a containerized environment. Successfully doing so could introduce auto-scaling and cost optimization, enhance engineering recruitment and business development, and modernize in the face of the particularly fast growth Trove was seeing as the pandemic unfolded. However, as a quickly scaling business with a small engineering team that did not have much existing container experience, Trove needed a cloud services provider with proven expertise containerizing applications on AWS.
AWS introduced Trove to Mission as a recommended partner with particular acumen around container modernization. Mission assisted Trove by first validating the company’s expectations around the benefits of its containerization plans, and then by leading Trove’s technology strategy in selecting and utilizing Amazon Elastic Container Service (ECS). Mission proceeded to build out Trove’s container architecture, including dev, QA, staging, and production environments, along with a CI/CD pipeline. Mission also initiated the process of rearchitecting and containerizing 17 Trove applications, while providing close support and training to prepare Trove’s team to deploy future containerized applications into production on its own.
Mission successfully set Trove on a path to realize the full advantages of its new container architecture. Following the completed initiative, Trove then chose to utilize Mission Cloud One for ongoing AWS monitoring, support, cost optimization, DevOps, and enterprise tooling, replacing Trove’s existing NOC provider. Trove also engaged with Mission Managed Detection and Response (MDR) as a new SOC solution.
“For any organization like ours looking to modernize with containers on AWS, I would say run – don’t walk – to engage with Mission. Mission came in as a trusted partner, let us lean on them to do the heavy lifting, and helped us build the container architecture that had long been on our wish list. I highly recommend Mission both for the deep expertise and delightful customer experience it provides.”
Colleen Blakelock
Director of Engineering, Infrastructure, and Security
Trove provides logistics and white-label websites for its clothing resale partners, utilizing a multi-tenant database. Under its previous dynamic application architecture using Amazon EC2, Trove lacked the auto-scaling capabilities necessary to manage its resources efficiently. Retail partners would often run temporary promotions that created spikes in traffic that could last for several days. These periodic events caused uneven demand for AWS resources. However, without options for easy scalability, Trove’s solution had been to estimate the maximum potential demand for a promotion and maintain that server capacity throughout the event to ensure 24/7 availability. Trove recognized this approach as inefficient from a cost optimization perspective. It also carried risks to reliability, as the estimated maximum server demand could be surpassed.
Trove envisioned moving to a modern container architecture to address its scalability and reliability needs. With a container architecture providing auto-scaling, the company would be able to efficiently scale server capacity to match demand – optimizing costs while improving resiliency. At the same time, the new architecture would be more inviting to potential new hires. It would also enable Trove’s business development team to tout the company’s modernized capabilities and facilitate any potential new partner requirements.
However, with a small engineering team managing Trove’s technical requirements as a rapidly growing startup – and with minimal in-house container experience – Trove didn’t have the capacity to take on the containerization project alone. “We wanted to get there,” said Blakelock. “But taking on that project while keeping up with all our other duties was out of reach for us as a team. It was a wish list item we weren’t going to get to.” Trove needed a cloud managed services provider with expertise in AWS and container architecture to help drive the project to completion.
Trove prefers AWS as the most powerful and cost-effective cloud provider, and is the cloud its team has the most familiarity and experience with. “We use Terraform, and it’s very easy to do infrastructure-as-code with AWS services,” said Blakelock. “We also use S3 buckets in all of their amazingly powerful ways. AWS just has everything: whatever you need to do, there's at least two ways to do it in AWS.”
Before engaging in the containerization project, Trove received a fully-funded offer from Google Cloud Platform for support in adopting container architecture, if Trove would adopt GCP. Trove declined. “It would have been a huge migration for us,” said Blakelock. “We're solid in our AWS architecture. We said we weren’t going to make that leap, but thanks.”
Soon after, AWS itself made a similar offer to help fund Trove’s move to containers through the AWS Jump Start Program – a popular program specifically designed for startup businesses. Trove readily accepted.
AWS introduced Trove to Mission as a cloud services provider with proven experience helping organizations implement containerized architecture.
“We had solid rapport with Mission starting from the first call,” said Blakelock. “Mission immediately spoke our language and knew where we were coming from. We saw a path forward right away, in terms of how Mission approached the project, explained what was possible, and how they would work with us. To do this project in-house, I would have had to hire at least two more people and have them on it full time for probably six months. It was a no-brainer to have Mission come in ready to go.”
“Mission has probably the best project manager I’ve ever worked with. From day one, they were organized with every piece of information you'd need. It was a delight, and we just liked them from the get go. They answered any question we had, and were really patient with me and my team members who were new to containers.”
Colleen Blakelock
Director of Engineering, Infrastructure, and Security
The engagement started in January 2021 with Mission guiding Trove in selecting ECS to operate its container architecture as a lower-overhead, easier-to-learn alternative better fit for Trove’s use case than EKS. This recommendation best accommodated Trove’s request that its internal team receive training and support from Mission to be able to later containerize and manage future applications in production.
Mission then built out Trove’s container architecture on AWS, consisting of dev, QA, staging, and production environments. Mission rearchitected and containerized 17 Trove applications through dev, QA, and staging, while providing constant close training to prepare Trove to launch future applications in production. While funding from AWS came short of providing for a planned CI/CD pipeline, Mission completed other aspects of the project with such efficiency that Mission was able to still create the pipeline within the scope of its project hours.
After building Trove’s trust through the successful containerization project, the company further engaged Mission to take advantage of several of its AWS managed services. Trove has now replaced its previous NOC solution with Mission Cloud One, which it found to be a more capable, more proactive and prescriptive, and far more affordable service for meeting its monitoring, support, cost optimization, DevOps, and enterprise tooling needs. Trove also recently added Mission Managed Detection and Response (MDR) as its SOC solution.
“I've learned that if our Mission contact reaches out and wants to show me something, it always makes my day,” said Blakelock. “When I get an email from him, I know he's going to solve a problem for me. For example, I now really think of the Mission Cloud One group as partners we can rely on.”
With its Mission-built container architecture in place, Trove is realizing cost benefits and gains around scalability and reliability. Trove projects savings of 30% on its AWS expenses once the cost optimization from its container architecture takes hold. These benefits extend to the company’s customers in the form of increased uptime, and the ability to automatically scale to meet any quantity of demand. Containerization has improved Trove’s abilities to recruit new engineering talent as well. Trove’s business development team can also now cite its resilient container architecture and autoscaling capabilities to win the business of larger partners driving tremendous traffic volumes.
“Over time, yes, containerization is going to save us money,” said Blakelock. “But it’s also about making our architecture right-sized, responsive, and resilient. Auto-scaling allows us to stop worrying about server capacity in the middle of the night. It’s also about modernizing our architecture. The world keeps changing, and we want to keep up and be competitive. It was a lot harder to bring on engineers who wanted to work with our old school infrastructure before containers. Now we’re on the cool stuff that engineers want to be working on. Having this work behind us instead of on a wish list is amazing. Now as Trove grows, instead of accruing a bunch of technical debt, we’re ready for a fast-moving future.”
“I'll tell you what I didn't expect: I thought that Mission would need a lot more help from us. I thought ‘oh our stuff’s weird,’ and that they’d get stuck and not understand why something is the way it is. So that first week when Mission’s team said they had the first repo up in a container, I was like ‘no you don’t, our stuff’s not that easy.’ But Mission’s team was self-sufficient, and they were fast.”
Colleen Blakelock
Director of Engineering, Infrastructure, and Security
Trove plans to complete the containerization of future applications into production and leverage even more of the benefits of its new container architecture, with Mission’s ongoing support. Trove will also continue to utilize Mission Cloud One and Mission MDR managed services.