“Serverless” cloud services, such as AWS lambdas, are one of the fastest growing segment of the cloud services market. These services are lighter-weight and provide more flexibility in scheduling and cost, which contributes to their popularity, however the security issues associated with serverless computing are not well understood. In this work, we explore the feasibility of constructing a practical covert channel from lambdas. We establish that a fast and scalable co- residence detection for lambdas is key to enabling such a covert channel, and proceed to develop a generic, reliable, and scalable co-residence detector based on the memory bus hardware. Our technique enables dynamic neighbor discovery for co-resident lambdas and is incredibly fast, executing in a matter of seconds. We evaluate our approach for correctness and scalability, and perform a measurement study on lambda density in AWS cloud to demonstrate the practicality of establishing cloud covert channels using our co-residence detector for lambdas. Through this work, we show that efforts to secure co-residency detection on cloud platforms are not yet complete.