New Question
 
 
PRTG Network Monitor

Intuitive to Use.
Easy to manage.

200.000 administrators have chosen PRTG to monitor their network. Find out how you can reduce cost, increase QoS and ease planning, as well.

Free PRTG
Download >>

 

What is this?

This knowledgebase contains questions and answers about PRTG Network Monitor and network monitoring in general. You are invited to get involved by asking and answering questions!

Learn more

 

Top Tags


View all Tags


PRTG Performance Can we break down servers to balance load

Votes:

0

Your Vote:

Up

Down

Hi,

So we have 1 Core server, 1 Local probe and 9 remote probes. Altogether this holds around 7000 sensors.

"Rule of thumb: Typical PRTG installations almost never run into performance issues when they stay under 5,000 sensors, under 30 remote probes, and under 30 user accounts."

I understand we are over the "desired" amount of probes however are well under in probes and user accounts.

The last year we have noticed a massive drop in performance, for instance speed issues for all users, complete application outages. We are currently experiencing an issue where 5000 of our 7000 sensors are coming back as unknown and 6 of our are 9 probes are disconnecting constantly, we are looking for recommendations on how we can break PRTG down further to help with this burden on the core server.

We have taken the usual steps of turning off auto discovery on devices, tried to increase polling intervals. Removing unnecessary sensors etc.

I am wondering if we can create multiple core servers to distribute the load?

Or can we partition the core server into segments and assign certain probes to that partition?

Or any other suggestions that we can implement would be greatly appreciated.

Thanks.

core-server performance sensor

Created on May 31, 2018 8:44:59 AM by  DonHarris (0) 1



1 Reply

Votes:

0

Your Vote:

Up

Down

Hi there,

There is a significant difference in PRTG to "normal" applications, which usually only create 'load' when interacted with. Please bear in mind that PRTG is a constantly working application, 24/7, all year long. This calculates for example into 14,4 Million Measurements on a daily basis for a system with 10.000 Sensors and a scanning interval of 60 seconds. All these data values needs to be acquired, evaluated (against limits, etc.) and stored into the database.

The following steps are recommended to alleviate the load on your system memory and CPU:

  1. Disable the Cluster-Functionality or Install a second PRTG instance to spread the sensors on both PRTG Cores.
  2. Increase scanning intervals for the the sensor types that were mentioned above.
  3. Currently you have 8 CPUs on the server where the PRTG core is installed on. Please make sure they are all placed on one socket to assure best performance.
    Background: If you have less than 12 CPU cores in total, PRTG will only use the first CPU socket, independent of the amount of CPU cores. If you have more than 12, PRTG will use the first socket for the Core service and the seconds socket for the Probe service.

Additionally, I advise you to review the following KB-links which also provide insights and fine-tuning recommendations:

Distributing the installation across two Core Server would be an option but that would require an additional license.

Best regards.

Created on May 31, 2018 10:44:20 AM by  Dariusz Gorka [Paessler Support]



Please log in or register to enter your reply.


Disclaimer: The information in the Paessler Knowledge Base comes without warranty of any kind. Use at your own risk. Before applying any instructions please exercise proper system administrator housekeeping. You must make sure that a proper backup of all your data is available.