Mesosphere Blog

Mesosphere (https://mesosphere.com) | Remote | Software Engineers | Full-time Mesosphere is heavily involved in the development of Apache Mesos with the vision to build an infrastructure to utilize your entire datacenter or cloud instances like one big computer. In this role there will be a large variety of features to work on, including resource allocation, job scheduling, containerization, provisioning, metrics, logging, task preemption, checkpointing, fault-tolerance, auto-scaling, modularization, testing, and much more. All of these involve modern C++ libraries that facilitate asynchronous I/O handling for distributed computing and make concurrency and manual memory management fun again. As an added bonus, majority of your work will be upstreamed into Apache Mesos project giving your work global visibility and provide an avenue to become a committer of a popular OSS project! We are hiring in SF, Hamburg (Germany) and Remote. https://mesosphere.com/careers/?gh_jid=1265717 https://mesosphere.com/careers/?gh_jid=1265690 https://mesosphere.com/careers/?gh_jid=1229864 I'm the hiring manager and would love to hear from you.

Mesosphere, Inc. is hiring a variety of software engineers to help build the Datacenter Operating System, based upon Apache Mesos. If you're looking to work on distributed systems, large clusters at scale, containers & microservices, and big data frameworks, we'd love to hear from you. Languages we use include Scala, Java, Erlang, Go, C++, Python, & Javascript. We start the process with an intro call, then a coding challenge + review call, followed by a full day of onsite interviews. We are hiring for full-time roles in our SF, CA and Hamburg, Germany offices, and will sponsor visas wherever possible. If it makes sense for you and works for our team, we are open to remote, home office working arrangements, in both the US and EMEA. All openings: https://mesosphere.com/careers/

There are other open source container frameworks, we are using Aurora (http://aurora.apache.org/) but there is also Mesosphere (https://mesosphere.com/) and some other smaller ones...

Hi, Chronos Product Manager here ('air' on github). Status: Having fun creating issues in the wake of the Jepsen post : ) I can say that Mesosphere is 100% committed to both Marathon and Chronos. You're right that Marathon gets more attention for the moment, but we have grand plans for both. A strength of Mesos is that it's an open platform. If Chronos isn't right for you, there are solid alternatives. I'd appreciate if you raise issues against Chronos though so we can improve it : ) Excuse the plug, but if you'd like to help us improve these widely-used frameworks we'd love to hear from you: https://mesosphere.com/careers/

Mesosphere is a company, not a piece of software. I think you're referring to Mesosphere's dcos[1]. DCOS is some really nice packaging for marathon, chronos, etc, with a nice cli tool for downloading and installing new frameworks onto mesos. Personally, I find using Aurora a lot nicer. That being said, I do think the kubernetes-mesos gives you far and above the best of both worlds. You get the developer story of kubernetes, with the ops story of mesos. From an ops perspective, k8s is a bit clunky. I was actually shocked when I found out after bringing a new kubelet (worker node) online, you have to also update a service on the master. This was in a power training class on Kubernetes at this year's Redhat Summit in Boston. Really underwhelmed with the complexity of k8s compared to mesos, but they aren't an apples to apples comparison. [1] https://mesosphere.com/product/

It's more directly comparable to https://mesosphere.com/ - so, an abstraction further up the chain than Chef of Puppet. The general idea being that it creates a virtual operating system across multiple physical or virtual "hosts" and allocates application stacks (like containers, maybe) compute/memory/storage as required. It ideally handles fault tolerance and load balancing gracefully too.

A quick summary of what Mesosphere is working (DCOS - the datacenter opearting system) on is at https://mesosphere.com/ . Apache Mesos, the core of Mesosphere's infrastructure, has been in use by companies like Twitter for years to manage clusters with 10s of thousands of servers. http://mesos.apache.org/

It's worth noting that Kubernetes can also run on top of Mesos (https://github.com/GoogleCloudPlatform/kubernetes/tree/master/contrib/… and DCOS (https://mesosphere.com/learn/). So if you invest in Kubernetes and later want to take advantage of Mesos you can lift up k8s and put Mesos underneath. Mesos is definitely more mature, and you can use Marathon for container orchestration, which is a little more mature than Kubernetes, but it's also a bit simpler. Marathon doesn't have the service, secrets, or pod abstractions, for example. DCOS adds some additional benefits on top of Mesos, like the dcos-ui and dcos-cli. One of the more compelling features is one-step cluster package installation (ex: dcos package install kubernetes). Currently, the community edition is available on AWS (https://mesosphere.com/amazon/). I do agree that the kube-up scripts in k8s are a bit of a mess. There's really hard to read and reverse engineer, and each provider has their own divergent deployment methods. That's one of the things DCOS is trying to standardize, to have a consistent deployment pattern for Mesos frameworks that deploys their core components inside Marathon, giving them a more battle-tested platform to live within, and granting automatic resurrection in case of failure.

Disclosure: CoreOS employee; Kubernetes contributor. After reading the title I was a bit tempted to call BS since I think Kubernetes should have the rights to "The first Datacenter OS" tagline[1]. However, after reading up on the project details at https://mesosphere.com/learn I can see how Mesosphere came to the conclusion of being a DCOS, if not necessarily the first. Mesosphere goes a bit further than Kubernetes and offers a solution to the storage problem and attempts to address other "userland" concerns by shipping Apache Spark, Cassandra, Kafka, and Hadoop. So maybe it would be more accurate to call this a datacenter distro on top of the Kubernetes kernel? Regardless, I think the concept of a datacenter OS will be the key to commoditizing IaaS providers and leveling the playing field in terms of features and usability for those who have not given up on the dream of running a "private" cloud. Why will the DCOS work where others have failed? Current solutions aimed at taming the datacenter operate at the machine/VM level, which exposes the OS for each machine, and completely punts on the application. Guess who gets to stitch it all back together? A DCOS is designed to manage applications directly, commonly via application containers, which means we can treat the OS running on the underlying machines like firmware and limit our interactions to basic updates and minimal configuration -- think CoreOS. What about PaaS? That's a topic worthy of a lengthy discussion, but…

Why is this awesome? Production ready Mesos hosted in the cloud. Now I dont have to run a $500M company to cut server costs and save dev ops time. Apache Mesos is a system predominatly used by large companies, to more efficiently utilize servers in their data centers. The traditional way of organizing your servers requires a certain number of servers per cluster. For example: 50 webservers for your web app in one cluster, 5 servers for your database in another cluster, and 10 servers for redis, etc. On average, you'd utilize 50% of each cluster, but have the extra capacity for spikes. Mesos makes your servers way more efficient by treating them as raw power, allowing any server to run any kind of app/task. It chops a server up into many linux containers and can shuffle around your tasks, so that a web app, redis server, and database server could all be running in different areas of the same "physical" server. I read that one company saved ~40% of their server costs and served the same load just by switching to Mesos. Marathon is made by Mesosphere and acts as the brains or controller for your servers. You can allocate what percentage of resources go to certain apps/tasks, and it also handles deployment, scaling, failover and restarting. Mesos supports Docker, so converting an existing app to run this on DO with Mesosphere is simple. Some applications are built for Mesos already, like Spark (MapReduce replacement), and will be inherently faster without the Docker…

Yea I don't understand the original comment... however it is true Java isn't (or at least wasn't until recently) ideal for containerization, since the JVM would not obey things like cgroup limits for memory and CPU. This was fixed in Java 9 and backported to the Java 8 (update 131) however there are still potential issues regarding CPU isolation that are not trivially solved without some explicit configuration. For a good overview, see: https://mesosphere.com/blog/java-container/

I mean, clearly Kubernetes has won. That shouldn't be a matter of debate when Mesosphere is doing product releases with top-line functionality designed to make it easier for you to operate Kubernetes.[1] And Swarm has been a clear loser for a while now. I say this despite us running a reasonably sized Aurora/Mesos deployment! Three years ago this was probably the right choice. Today I think you end up debating things like whether or not you should go with self-hosting Kubernetes vs managed vs what Mesosphere is pitching now. [1] https://mesosphere.com/blog/dcos-1_11-overview/

Fun fact: Mesosphere promoting Kubernetes on DC/OS. Example: https://mesosphere.com/blog/dcos-1_11-kubernetes/

There are definitely Mesos clusters running in production. My company is one. Mesosphere publishes case studies around it mostly used as part of DC/OS. https://mesosphere.com/resource/category/case-study/

Here is another good tutorial - about Tensorflow on DC/OS: https://mesosphere.com/blog/2017/05/11/deep-learning-tensorflow-gpus-d…

Mesosphere has had an open-source free version for nearly a year now. You can read about it here: https://mesosphere.com/blog/2016/04/19/open-source-dcos/

Yup! https://mesosphere.com/blog/2016/04/19/open-source-dcos/ (I am a Mesosphere employee.)

DC/OS from Mesosphere is a an alternative to AWS that you should consider looking into. (Ref: https://mesosphere.com/product/) That will reduce your reliance on Amazon to just hardware, that's what are you paying for. In addition, you can use Google Cloud, Microsoft Azure, your datacenter alongside as a single homogenous cluster.

understood. The ones I've been doing are on average 4 hrs ( eg: https://mesosphere.com/careers/challenges/distributed-applications/). Did one last week that was 7 hrs. I am just tired and exhausted at this point.

This is cool. I also wrote a Marathon/Mesos version, without the vendor lock-in: https://mesosphere.com/blog/2016/04/06/lets-encrypt-dcos/

It's cool seeing technology developed at CERN in the spotlight. There are a lot of interesting tools developed there that can solve real problems outside CERN and academia. One such technology, featured in the article, is CernVM File System that is used to distribute terabytes of scientific software to hundreds of datacenters all over the world. A shameless plug: Apache Mesos recently integrated it to solve container image distribution problem (https://mesosphere.com/blog/2016/03/08/cernvmfs-mesos-containers/).

Have anybody works with Mesos? How does it compare to Yarn, particularly in a production perspective? Is it stable, easy to integrate? We are currently thinking to switch from a Kafka-Spark(on Yarn)-Mongo stack to a SMACK stack (Spark, Mesos, Akka, Cassandra, Kafka)[1]. It seems that there is a good integration between theses projects. Also you can run Docker on Mesos using Marathon[2] so not only our data-driven stack could be on Mesos but the full stack. [1] https://mesosphere.com/blog/2015/07/24/learn-everything-you-need-to-kn… [2] https://github.com/mesosphere/marathon

They talked a bit more about the Siri setup at a talk they gave for the Mesos guys (Siri runs on Mesos with a custom schedule Apple wrote - one of the biggest Mesos installs out there) https://mesosphere.com/blog/2015/04/23/apple-details-j-a-r-v-i-s-the-m… They say the clusters have 1k worker nodes

This looks like inverse offer in Mesos https://mesosphere.com/blog/2015/10/07/mesos-inverse-offers/

> Mesos has been simulated to scale to 50,000 nodes, although it is not clear how far scale has been pushed in production environments Apple are "many thousands"[0] to provide Siri, and Twitter use it for pretty much everything, though I can't remember any numbers (15000 springs to mind? But mesoscon was all a bit of a blur now) [0]: https://mesosphere.com/blog/2015/04/23/apple-details-j-a-r-v-i-s-the-m…

> I'm going to assume he means Stevenson Avenue, Polmont, Falkirk, Stirlingshire, United Kingdom -- since that's the first Google result I got for "Stevenson" as an address. Perfect point! Of course, somebody can't imagine the people exist who aren't in the same city as he is. https://mesosphere.com/contact/ "88 Stevenson St, San Francisco, CA 94105" And it's often forgotten that Google is not the same for everybody, customizing their result based on their own heuristics of who is asking and from where he's asking. But maybe the author of that post actually knows the person he communicates to is actually in the same city? I don't know.

"Adding k8s on mesos you get the next-generation cloud native scheduler and the ability to run existing workloads. by running k8s by itself you get a lightweight cluster environment for running next gen cloud native apps." @cmcluck More references: [1] https://mesosphere.com/blog/2015/04/22/making-kubernetes-a-first-class… [2] http://blog.kubernetes.io/2015/04/kubernetes-and-mesosphere-dcos.html [3] http://thenewstack.io/mesosphere-now-includes-kubernetes-for-managing-…

You can use it for web apps/services (https://mesosphere.com/docs/getting-started/service-discovery/). But I'm with you on databases. I guess I just like knowing that those hosts are being managed and cared for (pets vs. cattle). With that said, I often wonder how many people are using Mesos/Marathon before they have any need for it? Using it on for 4 hosts vs. 40 or 400?

This mesos 0.22 talk goes into a bit more what they are imagining - http://mesosphere.com/2015/03/27/mesos-0-22-0-released/ Essentially the current mesos disk quotas are just so that tasks that run, run with a minimum amount of disk space. I think what they are trying to accomplish in future releases is some way you could build an EBS Style disk space management on top of mesos.