How to deploy PRTG Network Monitor in AWS, Azure and GCP

 Originally published on June 29, 2020 by Greg Ross
Last updated on January 23, 2024 • 16 minute read

In October 2017, Paessler released the first hosted offering of PRTG built on the AWS platform. Since launching that first version of “PRTG Hosted by Paessler” our cloud team has gained a tremendous amount of experience with managing and automating AWS services.

By now you know the solution as our full-fledged product Paessler PRTG Hosted Monitor. This makes monitoring the cloud, from the cloud, a breeze.

Hosted monitoring for small  & medium IT infrastructure  Monitor all the systems, devices, applications,  traffic, and more in your IT infrastructure with  our cloud-based monitoring solution. Learn more

 

To learn how we deal with running PRTG in the cloud, check out our own Greg Campion in the video below, as he explains the back end functions for maintaining a hosted solution for our customers running PRTG in the cloud. Over that time, we now have customers across every industry using our hosted offering as it provides nearly all the benefits of an on-premise installation of PRTG without the concerns of updates, backups, redundancy or mobile app connectivity within a flexible SaaS subscription model.

 

Cloud-based Network Monitoring

✔️ Every sysadmin needs a reliable network monitoring tool. Traditionally installed on an in-house server, a network monitoring tool monitors every aspect of local and distributed networks, including bandwidth, hardware, and the CPU usage of virtual machines. A cloud-based network monitoring tool, on the other hand, runs in the cloud. As a result, it does not need to be hosted by a dedicated in-house server.

☁️ A cloud-based monitoring tool such as PRTG Hosted Monitor boasts all the same features as an in-house tool. What’s more, cloud-based tools are usually more flexible, less expensive, and easier to use.


Outside of our cloud team, the Paessler presales team runs all their PRTG demo installations from AWS EC2 m5 instances. These instances are geographically dispersed in different AWS regions including the United States, Ireland, Germany and Australia. This configuration allows us to leverage the durability and availability of cloud infrastructure for a global sales organization.

But what if you want to run PRTG in the public cloud on your own?

Is it supported? What type of instance should you use? What type of storage? How is it secured? How do you use remote probes? How is email delivery and SMS setup? Can you use mobile apps and PRTG Desktop?

Though not exhaustive we will review some of the most common questions for deploying PRTG in Amazon Web Services (AWS), Microsoft Azure and Google Cloud Platform (GCP).

Amazon Web Services (AWS)

Instance Types

T2 instances (more information)

aws-instance-type-prtg-01

T3 Instances (more information)

Name vCPUs Memory in GB Network
t3.large 2 8 5 Gbps max
t3.xlarge 4 16 5 Gbps max
t3.2xlarge 8 32 5 Gbps max

M5 Instances (more information)

Name vCPUs Memory in GB Network
m5.large 2 8 10 Gbps max
m5.xlarge 4 16 10 Gbps max
m5.2xlarge 8 32 10 Gbps max

Storage

Both the t3 and m5 are EBS-Only storage. In some cases, customers will use a single GP2 volume and have PRTG installed along with the data directory reside on the same volume as the OS. In other cases customers will use Throughput Optimized HDD st1 for the PRTG data disk. (more information)

One of the biggest concerns with moving infrastructure to the cloud are the ongoing costs. With that in mind, our sensor team built the AWS Cost Sensor for PRTG Network Monitor to give you better visibility to monthly and yearly costs along with forecasted costs by querying data from the AWS Cost Explorer API.

Microsoft Azure

Instance Types

azure-instance-type-prtg-01

Storage

To find out, what disks are available in Microsoft Azure, just check the Microsoft Virtual Machine documentation.

Cost is very important regardless of the public cloud platform you utilize and with that in mind our developers are working on an Azure Cost Sensor like our AWS cost sensor. Please see our PRTG public roadmap for information on this and other upcoming features. 

Google Cloud Platform (GCP)

Instance Types

In GCP the instances are known as machine types. Check the GCP documentation for more information.

Name vCPUs Memory in GB  
n1-standard-2 2 7.5  
n2-standard-2 2 8  

Storage

Zonal Persistent Disks: standard HDD and SSD are both suitable for PRTG though larger installations will benefit from SSD to help with graph cache. (more information)


Standard ports needed for PRTG

To connect to the Web server on your PRTG core server to use the web UI remotely, connect your PRTG Desktop app or one of our mobile apps, you will need to open TCP port 443 (SSL) or a custom web port. We covered this topic in our knowledge base, simply check out this link.

Remote probes require TCP port 23560 (default) to be open as well. You will need to evaluate the security groups, ACLs and other security settings for each public cloud platform to use each of these PRTG features. PRTG email delivery can leverage the built in SMTP server or your own mail server. In either case, these mail delivery options will require outbound connectivity.

Here are the steps required for two of the most common email delivery setups our customers use for email relay: Microsoft 365 and G-Suite.

Share your cloud experience with us!

Do you have experience running PRTG from the public cloud? What have you learned? Please let us know in the comments below!