Scootsy — Changing up the delivery game

Anuranjani Rathore
CloudCover
Published in
3 min readNov 8, 2019

About Scootsy

Scootsy is a Mumbai-based on-demand delivery firm. They deliver a broad selection of gourmet food, fashion, home decor, gifts, and SOS items all across the city. Scootsy differentiates itself from popular Indian online delivery firms like Zomato and Swiggy by offering only high-end products from selected restaurants and stores that meet their quality benchmark.

Within three years of its launch, Scootsy raised a total of $3.6M in funding over 1 round and is looking at expanding to other metros within India.

Challenges

  • Scootsy delivers gourmet food, packages, and bakery items all across Mumbai. During peak times, they receive a high number of requests that significantly increases their website traffic; whereas off-peak hours/ during weekdays, there is less website traffic. In order to serve their customers at all times, they need to meet the demand in both scenarios; which eventually leads to them overprovisioning their infrastructure.
  • To build a responsive cloud infrastructure that supports thousands of users. Scootsy noticed that the end-users were often running into errors while browsing or placing orders. Their multiple services (food delivery and logistics) had a tight dependency on each other since they were running on the same virtual machine. An impact on one service would automatically impact all other services. This would increase the downtime and impact the end-user experience.
  • With all services dependant on each other, the main challenge was isolating them so that when one service is updated, others are not impacted.

Scootsy wanted to make sure that their business and team is scaling together. They used CloudCover’s deep knowledge of products and best practices to build technical expertise and introduce a DevOps culture within their team.

Goals

  • To complete the migration with minimum impact on their business
  • To improve the scalability of their services on Google Cloud
  • To introduce a structured approach to their current development and deployment cycle

Infrastructure Components

Third-Party Tools

Solution

CloudCover performed a big bang migration of Scootsy’s application and database stack in order to prevent any service outage that might arise due to latency issues during live migration. To achieve high durability, improve data resiliency and reduce running costs, CloudCover recommended Scootsy to offload all their static content to Google Cloud Storage.

From a scalability point of view, CloudCover helped them isolate their services from each other, and implemented containerization (to prevent domino effect in case of failures) by using a fully-managed Kubernetes service available in Google Cloud. This saved Scootsy an ample amount of time and their engineers can now focus on improving the existing app through containerization.

By introducing a well structured CI/CD pipeline, their change management and release management processes were improved drastically. The combination of automation and third-party build tools like Jenkins has helped them eliminate human intervention during code releases.

Architecture Diagram

Results/Benefits

  • Decoupled 15 web applications while migrating to Google Cloud, ensuring that their customers do not face any service outage
  • Transformed the application stack to run on Google Kubernetes Engine
  • Implemented the right set of auto-scaling policies to ensure that they are not incurring any extra costs while serving their daily traffic
  • Significantly improved user experience by helping them manage any unexpected load or traffic that might come in

About CloudCover

CloudCover delivers the insane potential of the public cloud to start-ups & agile enterprises through a combination of weaponized geekiness, extreme automation, and battle-scarred experience.

--

--