Uptime simplifies system and server monitoring
- 13 January, 2015 07:42
Server and systems management tools have long been too expensive and too complex to actually use without spending thousands of dollars on consultants. Uptime Software has been trying to change that with a product that can be installed and useful within an hour. We tested the latest version (7.3.1) on a small network of Windows physical and virtual machines.
Like similar products, Uptime consists of a monitoring server that collects information from several scanning methods, including SNMP traps, WMI, and its own series of agents installed on each Windows and Linux PC.
The measure of a systems management tool is how wide a net it can cast and what it can manage. Uptime has been adding elements to both categories, and it can handle Oracle, mySQL, SQL Server, Microsoft Exchange, WebSphere, IIS, WebLogic, NFS and Windows file shares, various VMware events, along with many other applications and situations.
With its agents, Uptime also has the ability to automate some common self-healing tasks. You can set up profiles, for example, to detect when a Web server is down and can then automatically restart it without any need for operator intervention. This is a very powerful feature. You can also build scripts that can restart certain Windows services: these can run on the central monitoring station rather than as part of the agent code.
Like other systems management tools, there is a wide collection of user roles that can be set up in a very granular way. These roles can be applied to specific elements or action profiles, with the ability to view, add new ones, or change existing ones. This makes it easier to deploy it across your enterprise, where you might want to have the server folks only have access rights for monitoring those elements, and the network infrastructure folks to only have access to that portion of the product.
Once you install the monitoring server on any Windows or Linux 64-bit machine, (32-bit machines are no longer supported for running this server but can still be tracked) you bring up a recent vintage of a Web browser to configure it, run reports, and all other functions of the product.
This is similar to how CA's Nimsoft Snap monitoring tool and others work, which makes it easier for IT administrators to remotely track what is going on without having to be tied to a particular PC.
We had some issues getting the agents to communicate with the monitoring server running under Oracle's Virtual Box and had to allow traffic through the client PC on the monitoring port 9998 (or whatever you choose for the agent to communicate).
Elements such as monitored systems or particular application servers can be quickly added with an auto-discovery tool that can scan a range of IP addresses or look for specific systems. Once these elements are added into its database, the real utility of the product begins. Uptime will track outages, when a particular condition is met (such as a disk drive that is nearing its capacity or a CPU that is being heavily used), and provide copious reports for your viewing pleasure.
And by copious we mean dozens of reports on just about anything you can dream up: enterprise CPU usage, summaries of your service-level agreements, vSphere workloads, and more. Reports can be scheduled periodically, limited to particular date ranges, and sent out via email in a variety of formats, including HTML and PDF, and archived on the Web console for easy recall.
The latest version includes these enhancements:
- Better VMware infrastructure monitoring. Uptime has always had some integration with vSphere, but it has gotten better. Now you can create alerts on additional VM-specific metrics such as memory usage, and CPU ready/wait time. This also includes an understanding of your network topological dependencies. So if a VM host fails, Uptime knows that the VMs running on that host also have stopped working and they don't have send to alerts about those VMs.
- Expanded API-based element management. This allows integrators to build new synchronization, bulk management, or discovery tools to tightly integrate Uptime with tools such as CMDBs or asset management services. You can now add agent-based, WMI-based, and network-device elements, each with their own APIs. Also, all of the Uptime core features are now available through RESTful APIs, so you can query for current element status or alert acknowledgements. There is published documentation on these APIs here.
- Customizable dashboards and plug-ins. Uptime has always been dashboard-heavy, meaning that their analytics are approachable, clickable (to drill down for further analysis and understanding of particular events), and visually interesting. And they continue to add new dashboards with dozens of fancy plug-ins. New to this version is an extensions manager to help you do this from inside Uptime itself. We had a hard time finding this tool (there is a small gear icon in the corner of the screen that kicks off a series of dialogs to add new plug-ins). You can add completely new dashboards or new service monitors (such as one that checks to see if SSL certificates are about to expire) to the appropriate places. If you want to get an idea of what else is available, see this list.
- Improved auto-discovery. The automated scanning across your network happens faster, and you can control exactly what you wish to scan too.
Online documentation is well indexed and hyperlinked, making it easy to find explanations of commands and menus and configurations. You can download a fully functional 30-day trial and there is also a live demo environment where you can navigate around a working installation to see how a more complex network is being monitored.
Pricing for Uptime is very simple. Each element costs $99, with an element defined as whatever you are monitoring. So you can have the full version of this tool to try out on a few elements for a very reasonable cost, and expand it as your needs require. Volume discounts are available too. Pricing and support could change though as Uptime was recently acquired by Idera.
Strom is the founding editor-in-chief of Network Computing magazine and has written thousands of magazine articles and two books on various IT and networking topics. His blog can be found at strominator.com and you can follow him on Twitter @dstrom. He lives in St. Louis.