Paeesler is a German company founded in 1997 by Dirk Paeesler, who wanted to create an easy to deploy and operate network monitoring solution, having no idea that what was at the time a “one man show” would have become in 20 years a company with 150,000 deployments worldwide and a staff of 170 people based in different countries. Paessler is a quite peculiar company as it is completely independent and owned by its founders and staff. It is also interesting to note that PRTG, Paessler’s solution, is adopted by 70% of Fortune 100 companies worldwide and acknowledged in Gartner’s magic quadrant.
The strenght of PRTG is that it started as a pure network monitoring solution and over the years it evolved into a “unified” one capable of consolidating the monitoring needs of a composite IT organization. Monitored objects can be networking, computing and storage hardware items, virtualization infrastructure components, applications and even whole clouds. This remarkable result is achieved without any special magic, simply by leveraging the power of good old SNMP, only sometimes with a little help from WMI or SSH when needed.
The other peculiarity is that PRTG is extremely simple to deploy, configure, mantain and kept up to date; Paeesler believes in the “Continuous Rollout” paradigm to ensure that PRTG is constantly and seamlessly kept up to date to provide new features, support new technologies, have bugs fixed, the quality of its code improved, all of this without headaches on the user’s side. This is achieved by Paessler through three different “release channels”: the “Stable Channel” is updated once or twice a month only with super tested new features and it is meant for production environments. Then, for more adventurous users (or users who have a test environment they can play with) there is the “Preview Channel” which is updated more frequently with code that will eventually end up in Stable (you get it first but you accept some risk) and finally there’s the “Canary Channel”, which is mainly for Nightly Builds (a big no-no for production environments!).
The PRTG UI is designed to be accessed from a wide range of devices, from PC/Macs through a fast and responsive AJAX web interface, to mobile devices by means of dedicated Apps; even wearables like Apple Watch or Pebble Time are supported! In addition to the many options just mentioned, the Windows only Enterprise Console application can be used to control multiple instances of PRTG from one single entry point. As you can imagine, Paessler has you covered when it comes to access the tool!
In terms of evolution of monitoring scope, Paessler believes they have now reached stage 4.0. They started with version 1.0 when the scope of monitored objects was mainly network devices; version 2.0 introduced support for virtualization (hypervisors, hosts, storage, VMs etc), then with 3.0 PRTG became “Cloud Aware”. Finally in version 4.0 PRTG’s reach was extended to the “Internet of Things”. Quite a remarkable progress I’d say.
So, how does PRTG get data from the monitored objects? It does so through SNMP as almost any device can be queried through this protocol and reply with useful information that PRTG can then report. PRTG comes with more than 240 embedded SNMP-based sensors (some of them based on vendor-specific MIBs provided by Dell, HP etc), but its capabilities can be easily expanded by ingesting custom MIBs through the MIB Importer tool; these MIBs can then be converted into OID Libraries for future use and inclusion into templates: import MIBs once, use the new sensors anywhere and anytime!
So, what is a “sensor” in PRTG jargon? A sensor is a collection of channels all pointing to an object: imagine a specific physical server, then a sensor would point to it, collect and present HW related info like CPU Temperature, Voltage and Ampere statuses, Fans RPM etc, each one through a dedicated channel.
Each of these values comes from OIDs, so you can imagine the amount of channels available for any kind of SNMP supported object and the wealth of useful info PRTG can gather and present. For those of you who are curious of how PRTG collects info from a vSphere environment, that is done out of the box by means of a specific VMware VM SOAP Sensor that can provide basic VM monitoring: Paessler suggests to enable the SNMP daemon on vCenter to enable full monitoring capabilities on the vSphere layer. Whatever is not covered by embedded sensors, imported MIBs, WMI or SSH can still be monitored as long as the end users are capable to write their own scripts that produce a valid XML output. Paeesler support is always available to provide assistance to customers with customizing PRTG.
Collected data is then organized in a hierarchical tree fashion:
Root Group > Probe (the PRTG node or remote collector) > Group (e.g. Virtualization, Network, Storage etc.) > Device > Sensor > Channel
This should be self explanatory and this organizational structure allows for simple navigation of the infrastructure.
Sensor definition brings directly to Licensing: PRTG is free for up to 100 sensors (so no reason not to download and try it) then it progressively scales up to unlimited sensors in reasonable steps (adequately priced). This should address the neeeds of organizations of any size. All the features are available regardless of the license purchased, what changes is only the amount of sensors supported. Simple as this.
The question that comes next is about scaling out the PRTG installation to support all the needed sensors: Paessler suggests not to exceed 5000 sensors per node if the deployment is on a VM, 10000 if on a physical machine. Once those limits are reached, one just deploys another node (as long as licenses purchased allow so). As mentioned above, the whole installation, regardless of the number of nodes, can be managed through the Enterprise Console. Remote probes can also be used to collect data from remote offices or segregated networks and send them to the Core PRTG Server. You can have unlimited remote probes as long as (again) you are covered by licensing in terms of number of allowed sensors.
The installation process is as simple as double-clicking the Windows installer package and in two minutes PRTG is reasy for Auto Discovery: once credentials are provided objects are found and the topology is populated according to the tree structure mentioned before. PRTG collection and monitoring immediately begins. PRTG also provides for HA and protection of collected data: clustering comes with any license size, just deploy a second node, enable the cluster feature and the secondary node will kick in in case of failure of the primary one.
Christian Zeh ran a live demo of PRTG during the TFDx seession. I believe that seeing is believing so I would suggest you to head to the Tech Field Day website and see for yourself either from here or from the TFD Website:
PRTG is deifintely an interesting product, it might not cover extremely specialized needs that require vendor or technology specific tools, but it does an incredible job at putting together a broad wealth of information coming from any corner of your infrastructure, visualizing them in a simple and effective way and allowing the IT admins to monitor their assets with little if no overhead at all. Being free for 100 sensors (or fully unlocked for 30 days in Eval Mode) makes it a no brainer to test it before eventually deciding to purchase it.
Disclaimer: I have been kindly invited by Gestalt IT to attend Tech Field Day Extra sessions while I was in Barcelona for VMworld Europe. Gestalt IT did not pay for travel, accomodation and meals and I was no under obligation to write blogposts, tweet or endorse any of the vendors that presented at TFDx.