Last week, Zenoss set up a follow-up rematch between Zenoss Service Dynamics (AKA ZSD, Zenoss Commercial and Zenoss Enterprise) and its open-source foundation, Zenoss Core. While the initial Core vs. Commercial webinar focused on general differences, this one offered a point-by-point comparison between the two flavors.
Zenoss Marketing Strategist Deepak Kanwar moderated the discussion, which featured Zenoss Community Manager Andrew Kirch as the Zenoss Core representative and Zenoss Senior Technical Product Manager, Matt Maloney, as the Zenoss Commercial one.
First off, this was a friendly match. Both sides agreed either flavor of Zenoss is best of breed in the unified monitoring space. Given that most companies tout their products as best of breed, Andrew referenced Network World’s July 2013 open source monitoring showdown, which Zenoss Core won handily. Network World reporter Susan Perschke wrote:
Zenoss is our top pick due primarily to its intuitive and professional-grade admin interface...we were able to configure our environment and run reports easily, and when help was needed, we found the user guide to be an excellent resource, a rare find in the open source world.
In other words, both versions are viable unified monitoring solutions, but your needs will determine the best one for your situation. I’ve rounded up a few examples below to get you started...
CHOOSE CORE IF...
You’re a small to medium-sized organization.
If you’re running a business with less than 1,000 or so managed resources (which includes everything from servers to routers to pilot instances and more), Zenoss Core should provide you with a feature-packed monitoring tool. Andrew described Zenoss Core as a turnkey solution with “one source collecting all of your performance and monitoring data.”
In addition to its excellent user interface, Zenoss Core [offers] ample extensibility using Core’s Python-based plugins, known as ZenPacks.
Zenoss is written in a manner where you can extend almost any portion of the product in a comprehensive and integrated way with a ZenPack. If you’re looking for a [certain] functionality in the user interface or a way to integrate ticketing systems, you can write a ZenPack to do that. ZenPacks are flexible in what they can do.
Chances are you won’t need to write that ZenPack because you can find over 300 Community ZenPacks at the Zenoss Community Wiki. And if you don’t find what you’re looking for, go to the Zenoss IRC Channel and ask whether anyone is already building that ZenPack. You may just get to test out that code before it gets released.
Zenoss Core is not in itself a distributed architecture, however, and once you get north of 1,000 resources, “it does start to fall down,” Andrew said.
You are comfortable using Linux (or have a system admin who is).
Zenoss Core is a Linux-based application that uses resources like SQL, RabbitMQ for messaging, and components in Python and Java that need to communicate with one another, among other things.
If you don’t have a reasonably competent in-house Linux system administrator, you’re going to find yourself in the deep weeds. You have to have a knowledge set that gets you at least most of the way through the installer and to the point where you can identify a problem and assist in troubleshooting if you’re going to effectively leverage the community. It’s critical.
If Linux isn’t part of your IT organization’s skillset, you should probably defer to Zenoss Commercial.
You enjoy participating in a vibrant open source community.
Network World was impressed by the amount of documentation available with Zenoss Core, adding:
We appreciated the well-organized, current and comprehensive 200+ page PDF user manual, a rare find in the open source community.
Under Andrew’s stewardship, the Zenoss Core community has rolled out an extensive Wiki that contains loads of documentation, including an interactive user guide with video and a listing of every ZenPack currently available. And Andrew marveled at how quickly the community has grown over the last year or so:
The IRC channel went from being very slow to being extremely active during U.S. business hours. And we're going to be releasing brand new forum software here in the next 30 days that is going to infuse new life. There’s quite a lot of infrastructure for getting support.
But again, a strong Linux-based background is essential for making use of these resources because the community is not designed to provide the sort of troubleshooting that you would receive with Zenoss Commercial:
Support is somewhat on the community's terms, and there are definitely ethical and social mores there where people expect you to come up with an intelligent question.
In other words, you can’t log into the Zenoss IRC channel (#zenoss at irc.freenode.net) and say, “Hey, the Amazon Web Services ZenPack isn’t working!” You need to be able to give users details about your installation, the devices you’re monitoring, and all other relevant information that would help members of the community glean an understanding of your situation. As talented and knowledgeable as these community members are, they’re volunteers, not your personal help desk.
CHOOSE COMMERCIAL IF...
You need automated Service Assurance (SA) and Root Cause Analysis (RCA) capabilities.
Before discussing Service Assurance capabilities in Zenoss Commercial, Matt noted that to some degree basic monitoring is Service Assurance:
You’re monitoring your infrastructure; you know what state it's in.
But once you go beyond 1,000 or so managed resources, sticking with this paradigm becomes increasingly time-consuming and pricey.
If your environment is highly virtualized, and you’re creating and destroying VMs [and] moving a lot of them around, and moving workloads among different servers to optimize your infrastructure deployment and capacity planning, it becomes difficult to efficiently map these targets back to the service.
Zenoss Service Impact, a component of Zenoss Service Dynamics, takes a “top-down” approach to Service Assurance that places services at the top and maps them to your critical infrastructure components at the bottom. Service Impact is a combination automated Root Cause Analysis engine and service modeling engine:
Service Impact defines the services people care about, whether internal or [customer-facing]. We don’t care about a specific server in your environment [but] about the ability to access the service people are trying to consume. Not only can you map that, [you can] provide context around low-level things you’re monitoring in your environment.
Impact adapters are built into Zenoss Commercial ZenPacks. If you use, say, the VMware ZenPack to numerate specific VMs, and you move a VM from one hypervisor to another, Service Impact updates those relations for you. If you have standardized on a CMDB, you can choose to feed (or receive) this information to (from) a CMDB, so that you have context whenever something goes down, obtain an analysis of the most probable culprits, and prioritize its importance.
You need seamless integration with third-party IT service management tools like ServiceNow and BMC Remedy
According to Matt, Zenoss Commercial integrates with both ServiceNow and BMC Remedy. With the current integration into ServiceNow for example, you can generate trouble tickets directly into ServiceNow that include actionable probable root cause arranged by confidence levels.
As you add services to ServiceNow, those actual business services then map directly into [Zenoss] Service Impact, as well as their subsequent dependencies, [such as] servers, routers, switches, and whatnot. Effectively you can control how monitoring is done in Zenoss Commercial through ServiceNow.
From a monitoring perspective (not service assurance), Andrew noted that Zenoss Core could be made to do most anything Zenoss Commercial did, including integration. But he doesn’t see the point of spending research and development time (and dollars) duplicating a solution that’s already been developed:
Could you build a Remedy ZenPack? Can you build some of these higher-level features? Certainly! It's just expensive to do, and you're still having to provide your own support, do your own dimensioning during the deploy time, and you just don't have the expertise that turns a piece of software into a solution. That transition isn't there without the support or functionality we already provide with Zenoss Commercial.
Regardless, Choose Zenoss
This webinar makes additional comparisons between Zenoss Core and Zenoss Commercial, and I recommend watching it if you’re still trying to figure out the Zenoss for your monitoring needs. If you have additional questions, please post them in the comments!