Tradesy

Tradesy_Case_Study_Cover_Perspective-Thumbnail.png

Screen_Shot_2016-01-20_at_11.55.45_AM.png

Problem

Tradesy enables women to maximize both their personal fashion and their personal finances. The mobile and online clothing re-sale service launched in October 2012. In three short years, it has grown from a nimble, seven-person startup to a 130-person company that is a hip destination for millions of users eager to buy and sell clothes. As with all user-driven platforms, Tradesy gives control of data input to its users. Though this is a huge benefit to the users, it also makes database management more challenging. Between Tradesy’s rapid company growth and the huge volume of data in its system, it has extremely high standards for how well its developers can navigate its systems.

By design, Tradesy’s IT team is comprised of full-stack engineers, meaning each of its developers also takes care of his or her own database administration and sys-ops. That kind of structure has proven to be a huge boon to the company, reducing bottlenecks and improving utilization efficiency. But such generalization also means that there’s no one at Tradesy with specialized DBA expertise.

Despite the obvious advantages of such a set-up, when data activity gets dicey and complex, it can be dangerous.

"Even with only a few members of our team initially using VividCortex, it was so easy to spot problems. We were able to bring maximum CPU utilization spikes down from 80% to 10%.... VividCortex gave us the necessary information to determine what was problematic. We have been safely fluctuating under 10% utilization for many months now. ”                                                                                                                        - Jesse Forrest, Tradesy Engineering Manager

VividCortex's Impact

Before Tradesy began using VividCortex, database incidents were frequent, causing reactive fire-drills. Devs spent valuable time trying to troubleshoot performance issues with cumbersome log file analysis.

Previously, Tradesy relied on query logs and Nagios. “In the past,” Jesse Forrest, Tradesy’s Engineering Manager, said, “it was extremely difficult to understand inefficiencies in our databases. VividCortex has a swath of tools that empower the user to truly understand everything that is happening under the hood of their databases. One of the most powerful features we have found is the ability to pinpoint ineffciencies in queries.”

Tradesy has now deployed VividCortex across its team. All of Tradesy’s engineers can proactively monitor their database and maintain high awareness of its activity. Time-consuming fire-drills are a thing of the past.

Benefits

  • Enable development teams to self service
  • Drastically cut CPU usage
  • Avoid time-cosuming fire-drills
  • Identify bottlenecks and solutions
  • Empower full-stack developers with total confidence in handling their database

Results

Since introducing VividCortex, Tradesy has achieved higher deployment frequency. Deployments are faster too, and Tradesy developers find any setbacks “super easy to spot.” From there, they’re pros at making quick fixes, adding indexes, or addressing whatever else needs to be handled. “VividCortex is the only tool we’ve seen that provides query-level performance data. We could never do this before. We’re able to rollout code, see if anything spikes on VividCortex, rollback if problems occur.”


Read the full case study here:

Tradesy_Case_Study_Cover_Perspective.png

Want to Know More?