As the decoupled (often services-driven) componentisation of computing continues apace, so does the proliferation of vendors aiming to deliver attuned wider tooling adjuncts to new lifeblood of code construction.
In this vein, Container Solutions brings forth its minimesos project, an open source testing and experiment tool for Apache Mesos.
The project can be accessed here: https://minimesos.org/try/
NOTE: Apache Mesos is a distributed systems kernel i.e. as Jason Hibbits described it here, [it is] an open source technology that brings together multiple servers into a shared pool of resources — effectively, it’s like an operating system for the datacentre.
Minimesos aims to solve a problem common amongst developers of container-based applications. Moving microservice applications from a laptop to a production environment is challenging because the target platform is different to the local one.
So, consequently, developers need a way to create a production-like environment on their desktops for building, experimenting and testing.
Minimesos allows developers to bring up a containerised Apache Mesos cluster on their laptop, complete with Apache Zookeeper and Weave Scope, for visualisation.
“When we started building a number of Mesos frameworks, we found it hard to run and test them locally,” said Jamie Dobson, CEO of Container Solutions. “So, we ended up writing a few scripts to solve the problem. Those scripts became minimesos, which lets you do everything on your laptop. We later integrated Scope so that developers could visualise their applications. This made minimesos even more useful for exploratory testing.”
Minimesos is simple says Dobson, users can start a Mesos cluster on the command line or via the Java API.
It is logically isolated, Mesos master, slave and Zookeeper processes run in separate Docker containers — and minimesos is integrated i.e. it exposes framework, state and task information to its Cluster State API.
A new version of minimesos, v0.9.0, is now available. It adds the minimesos ps command, which shows what is running on the cluster. It prints the framework, the task name and its state. An uninstall command is added in the latest version, as well as tokens for IP addresses in Marathon JSON files.