Sysdig regularly makes note of the fact that monitoring a container environment requires a monitoring tool that runs natively on containers. Otherwise, a monitoring tool has to rely on agents that more often than not are heavier than the container being monitored. Today Sysdig extended that reach of its namesake monitoring tool to the Mesosphere data center operating system platform and Marathon scheduling software.
Apurva Davé, vice president of marketing for Sysdig, notes that it’s not only important to be able to monitor the container, but also the orchestration framework managing it. For that reason Davé says Mesosphere is the first in a series of orchestration frameworks that Sysdig will be adding support for in both the open source version of Sysdig and Sysdig Cloud in the months ahead.
Sysdig will make use of application programming interfaces (APIs) to employ metadata provided by Mesosphere’s offerings to monitor and troubleshoot both physical resources as well as logical groups that make up a distributed set of applications and microservices. IT operations teams will be monitor individual hosts in Mesos environments using the core ContainerVision technology that Sysdig created for root cause analysis, performance analysis, system exploration, and security analysis.
Sysdig Cloud, the commercial version of the monitoring tool, comes with pre-built monitoring dashboards and templates to quickly analyze, for example, the performance of Mesos-specific frameworks, applications, groups and tasks across tens of thousands of containers running on any number of clouds, says Davé.
Of course, in all probability most enterprise IT organizations are going to wind up with multiple orchestration frameworks for managing containers. But that doesn’t necessarily mean they will want separate monitoring tools for each environment. It’s more likely they will want a common set of monitoring tools to better compare container application performance across multiple orchestration frameworks. In addition, Sysdig is betting that while developers and IT operations staff will want access to common pool of metrics, the dashboards they will want to view that data through will be substantially different. The end result should be greater DevOps collaboration being achieved in a way that doesn’t force one camp or another into a particular format for consuming those metrics. That’s critical at a time when IT organizations may soon find themselves managing more applications than ever as developers take advantage of containers to dramatically increase the total number of production applications running at any given time in the enterprise.
While there is a natural temptation to want to extend existing monitoring investment to track what is occurring in a container environment, Davé says it’s impossible to do that with breaking the fundamental lightweight container paradigm that make container so attractive to developers and IT operations teams alike. ContainerVision, explains Davé, provides insights into not only the containers, but also the infrastructure they run on and the applications inside them. For those reasons Davé says Sysdig has enjoyed considerable success at a time when both developers and IT operations teams are just now coming to terms with the implications of the impact containers will have across the enterprise.