CDNs have reshaped the Internet architecture at large. They operate (globally) distributed networks of servers to reduce latencies as well as to increase availability for content and to handle large traffic bursts. Traditionally, content providers were mostly limited to a single CDN operator. However, in recent years, more and more content providers employ multiple CDNs to serve the same content and provide the same services. Thus, switching between CDNs, which can be beneficial to reduce costs or to select CDNs by optimal performance in different geographic regions or to overcome CDN-specific outages, becomes an important task. Services that tackle this task emerged, also known as CDN broker, Multi-CDN selectors, or Meta-CDNs. Despite their existence, little is known about Meta-CDN operation in the wild. In this paper, we thus shed light on this topic by dissecting a major Meta-CDN. Our analysis provides insights into its infrastructure, its operation in practice, and its usage by Internet sites. We leverage PlanetLab and Ripe Atlas as distributed infrastructures to study how a Meta-CDN impacts the web latency. custom logic to direct traffic to a CDN that currently offers better performance and/or lower cost (e.g., at certain geographic regions or times). Since the routing approach employed by the Meta-CDN customers is unknown to the involved CDNs, directed traffic and thus generated revenue gets harder to predict. In particular, since decisions can be based on active performance measurements by the Meta-CDN, a (single) delivery of bad performance by the probed CDN can result in rerouting traffic to a competing CDN and thus losing revenue. Thus, while Meta-CDNs can offer cost and performance benefits to content providers, they also challenge CDN business models. Concerning Internetusers, performance-based routing decisions can yield better Internet performance and benefit end-users while cost-based decisions can have other effects (as for any server selection approach run by CDNs). While the concept is known and related work covering service specific implementations, i.e., Conviva's streaming platform [4,10,16], exists, the empirical understanding of a generic Meta-CDN and its operation in practice is still limited. We posit that this understanding is necessary.In this paper, we thus shed light on the Meta-CDN operation by dissecting the Cedexis Meta-CDN as a prominent example that is used by major Internet companies such as Microsoft (Windows Update and parts of the XBox Live Network), Air France, and LinkedIn [1]. Given its current adoption, understanding its functionality and its usage by customers provides a first step towards understanding currently unknown implications of Meta-CDNs on Internet operation. We thus investigate the infrastructure and services powering this Meta-CDN and provide insights about its operation in practice. We analyze for what kind of services, e.g., media, API backends, or bulk data transfers, customers utilize Cedexis and how different CDNs are employed. We further investigate how...