Senior Performance Engineer

Palo Alto, CA

What's the best way to achieve the fastest standing quarter mile? More horsepower? Better traction? Adding lightness? As a performance engineer you will be responsible for driving and establishing the critical performance metrics for MongoDB, from macro benchmarks through workload characterizations. You will work with cutting edge customers (internal & external) to characterize what is changing in the software, capture and analysis of results and determine how to achieve a set of performance goals (code changes, application changes etc.). You will determine and recommend changes to the core kernel to improve the performance. You need to be hands-on and drive building frameworks, workloads and understanding customer performance and capacity planning issues.
 
Responsibilities

- Set up of topologies: Create a library of large scale topologies that can be deployed to the cloud
- Load generation: Create tools to generate loads with a precise mix of operations
- Create a library of loads to run against target systems
- Response verification: Create tools to measure the response times, throughputs and correctness of responses during load testing
- System monitoring: Track the utilization of system resources across time at all nodes involved
- Stress tests: An increasingly complex set of stress tests aimed at verifying that systems behave in a healthy fashion
- Soak tests: Run heavy loads for a period of weeks in an attempt to uncover and fix longevity issues
- Functional throughput tests: Find the cost of isolated functions by running load tests and measuring throughput at saturation
- Compare behavior across builds to provide early warning of degraded performance
- System performance tests: Measure how performance is impacted as a cluster is expanded to hundreds of nodes
- White box testing: Instrument key code paths. Monitor performance of these code paths from build to build
- Provide recommendations on how code needs to change, refactored to improve performance 
 
Requirements

- Bachelor’s Degree in Computer Science, Math, or Engineering
- 6+ years hands-on experience in performance testing, data collection, analysis and workload characterization, bottleneck identification, and capacity planning
- Experience diagnosing the full stack; application, database, o/s, storage, and network layers
- Experience with large-scale, large volume, distributed 24x7 systems supporting millions of transactions / sec
- Experience in applying appropriate mathematic modeling techniques in deriving and validating projected performance improvement
- Excellent written and verbal communications skills; not afraid of standing in front of 100+ of your peers to discuss and defend your findings
- Excellent programming knowledge of (C or C++ preferred, Java / Scala ok)
- Excellent scripting skills (Shell, JavaScript, Python or Go)
- Must be an expert in UNIX/Linux O/S (Windows a bonus)
- Must have strong problem-solving skills
- Should have experience with cloud providers (Amazon EC2, Azure, Joyent , Rackspace, Softlayer)
- Should have database experience (SQL or Non-relational) 
- Should have experience with/knowledge of hypervisor technologies (KVM, VMWare, Xen or Hyper-V) 
 
 

About MongoDB

MongoDB (from humongous) is reinventing data management and powering big data as the leading NoSQL database. Designed for how we build and run applications today, it empowers organizations to be more agile and scalable. MongoDB enables new types of applications, better customer experience, faster time to market and lower costs. MongoDB has a thriving global community with more than 6 million downloads, 125,000 Online Education registrations, 25,000 MongoDB User Group (MUG) members and 20,000 annual MongoDB Days attendees. The company has more than 900 customers, including many of the world’s largest organizations.

To all recruitment agencies: MongoDB does not accept unsolicited agency resumes. Please do not forward resumes to our jobs alias or MongoDB employees. MongoDB is not responsible for any fees related to unsolicited resumes.