On Wednesday, we announced an SDK, developer program and certification program for the Mesosphere Datacenter Operating System. We consider it a big step in our company's maturation, as we seek to turn the DCOS into a true platform for running the next generation of distributed services and applications. You can learn more about it on the product page, or by reading our accompanying press release
and blog post
However, the promise might be best explained by developers and software companies who are already working with us to integrate their products with the DCOS. Here's what they have to say about why the DCOS matters and why they're getting on board with it early.
"As an early stage, venture-funded startup, our focus is on continuously innovating ArangoDB's functionality and growing our market share," said Claudius Weinberger, CEO and Founder of ArangoDB. "The Mesosphere SDK, the primitives in the Mesos kernel, and the DCOS extensions that allow for packaging and distribution are allowing us to expand our footprint rapidly. With DCOS, organizations are able to use and deploy our database in the cloud and on-prem, without us having to write and maintain all of that infrastructure code. Our partnership with Mesosphere gives us access to the market of distributed developers who are already highly leveraging Mesos and the DCOS."
"The need for distributed applications and distributed data services is causing major challenges for companies using traditional architectures," said John Kaiser, Director of Partners and Alliances at Basho. "By combining Basho's Riak with the Mesosphere Datacenter Operating System, we're able to deliver an easy-to-deploy platform for real time data processing, including back-ends for large scale data processing, web, mobile, and Internet-of-Things applications. The new abstractions provided by Mesosphere reduce the complexity of large-scale, distributed environments and let enterprise customers gain immediate value from the hottest applications and distributed technologies."
"At Concord, we're building a stateful stream processor that can handle mountains of data in real time," said Shinji Kim, Co-Founder and CEO of Concord, an early stage startup building a new data streaming processor on Mesos. "With Mesosphere's SDK we are able to build on a platform that handles all of operational mechanics and scheduling, letting us focus on delivering a more robust way of processing data. Moreover, Mesosphere's developer program gives us access to formal developer support and a growing community of users. For a young company like ours, this will be an exciting distribution channel for our product when we launch it."
"Apache Kafka lets you build real-time stream data pipelines that handle gigabytes of event streaming data per second in a highly elastic way, and we're seeing Mesos and the Mesosphere DCOS at the forefront of innovation around how datacenter resources enable data flow in big data analytics stacks." said Neha Narkhede, Co-founder and Head of Engineering at Confluent. "What's so compelling about the DCOS is the promise of running Kafka alongside distributed datastores like MemSQL and Cassandra, and Spark for analytics -- in a datacenter operating system platform that includes single command installation and native functionality for deploying and scaling to clusters. These are the hardest devops challenges for running modern big data analytics stacks, and DCOS is creating a simplified adoption path that makes running these technologies in tandem a user-friendly experience similar to downloading and running applications on a laptop or mobile device."
"The Mesosphere DCOS is really the only distributed systems platform that allows an enterprise user to easily administer distributed applications on a cluster that auto-scales under load," said Christian Lutz, COO at Crate.IO. "The operating system analogy is appropriate in the sense that you just don't think about how applications consume resources on your personal computer, and that level of abstraction is now being brought to distributed applications in the datacenter. We think a lot of ISVs are going to be attracted to DCOS for the promise that enterprises that normally couldn't consider giving a time slice to play around with a new distributed application can now simply install it and try it. This is a game changer in the application distribution status quo for enterprise."
"In the same way that containers have become synonymous with how distributed applications are packaged and shipped -- Mesos and the Mesosphere DCOS is becoming synonymous with orchestration of resources across clusters," said Alexis Le-Quoc, CTO at Datadog. "We're using the Mesosphere APIs to collect performance metrics and events in enterprise-scale container environments, running the DCOS. The Mesosphere SDK makes it very easy to build our integration and deploy our agents on all nodes so that we can provide the kind of comprehensive and real-time monitoring that our customers need in these highly distributed environments."
"We're excited to work with partners like Mesosphere because Cassandra and Mesos are a great combination," said Matt Rollender, VP, Cloud Strategy at DataStax. "With the Mesosphere DCOS users can deploy a modern microservices applications and add DataStax Enterprise to the same cluster with one command. That data can then be easily flushed to Cassandra or DataStax Enterprise without having to worry about problems you would normally encounter with a traditional relational database. The Mesosphere DCOS SDK opens up opportunities to deepen the integration points between Mesos and Cassandra or DataStax Enterprise for distributed developers working with high data volumes."
"We're seeing a lot of Mesos out there in the field today," said Gabriel Monroy, CTO at Engine Yard. "We see Mesosphere and the DCOS making the power of Apache Mesos more broadly available to the ecosystem of vendors tackling those problem areas and more, and this new SDK opening up some great opportunities around the promise of so-called developer 'self service' in solving some of distributed computing's hardest problems. The services that DCOS and Mesos offer for running large scale batch analytics and big data workloads are unparalleled. And nothing handles the combination of containerized workloads with non-containerized workloads better."
"Hedvig's distributed storage application, like any distributed application, needs to solve the problem of enabling customers to easily deploy the technology in microservices architectures with containerized workloads," said Avinash Lakshman, founder and CEO of Hedvig, and inventor of Apache Cassandra. "From a storage perspective, the trickiest detail is data persistence. This is a major strength of the DCOS. We believe that the DCOS approach to persisting data across containers and between clusters is optimal for the new generation of applications that requires a much more flexible underlying datastore."
"As the leader in production deployments for Hadoop, we see increasing demand for convergence of infrastructure between operational and analytics applications, eliminating idle resources and complex data pipelines," said Will Ochandarena, Director of Product Management, MapR Technologies. "We are excited to work together with Mesosphere to make this a reality, with DCOS providing resource scheduling and the MapR Data Platform providing reliable, persistent data services across different applications. With Apache Myriad, this solution will support the widest ecosystem, including native Mesos, YARN, and Docker applications."
"Mesosphere and MemSQL share a vision of easy deployments of distributed systems, and that's why we're excited to be working with Mesosphere's Developers' Program," said Eric Frenkiel, CEO and co-founder, MemSQL. "Developers can deploy MemSQL in under two minutes using the DCOS and then combine MemSQL with other technologies like Apache Kafka and Spark. DCOS is a new distribution channel for MemSQL, and we look forward to powering real-time applications on Mesosphere that need the in-memory speed and scale of MemSQL."
Project Calico (Metaswitch)
"As organizations move from a fairly static VM-based infrastructure to a containerized microservices application infrastructure, one of the challenges they will have to manage is the order of magnitude increase in the number of containers, each of which may be very short lived," said Alex Pollitt, Evangelist / Director of Engineering for Project Calico. "The Mesosphere team have been working hard to deliver the key technology pieces that allow companies deploying and building at scale to focus on the applications, not the new demands and pressures on how the underlying resources are orchestrated."
"Developers want tools that allow them to manage processes without stepping on themselves, and that's why we're excited to work with Mesosphere's developer program to distribute Spark on top of Mesosphere's DCOS," said Dean Wampler, Architect for Big Data Products and Services at Typesafe, who earlier this year announced commercial support for Apache Spark on the Mesosphere Datacenter Operating System. "Mesosphere's DCOS provides convenient tools for deploying applications, as well as managing and monitoring the cluster."