Splunk AppDynamics

Sum of memory usage in processes is misleading.

Gustavo_Marconi
Path Finder

Hi everybody,

I was doing an internal demo presentation with the demo1, and someone noticed that in a server the memory usage was high (87.8%), when we checked the processes that were running, there was no process consuming memory, except 3% of the machine agent, so we don't understand why is showing that 87.8% peak.
image.pngMemory usage 87.8%image.pngBut no process is consuming memory except for the machine agent.

In other example happens the opposite, the memory usage in the server is 34.6%.

image.pngMemory usage is 34.6%

but the sum of the processes is way more than 100% image.pngSum of processes are using way more than 100%.

Is this an interpretation problem of us or just an issue of the demo1? In other demos the sum is correct according to each process. 

Thanks in advance.

Hope you're having a great day.

Labels (1)
0 Karma
1 Solution

Troy_Partain
Engager

Hi Gustavo,

Excellent question and I can appreciate the interest in the discrepancies.

With demo systems, we're generally not dealing with live data. The way it's generated can vary and, at times, can cause abnormalities within the context of the data. This is what is going on here.  Beyond this, in production, it's advisable to ensure your looking at the most specific time range possible to reduce the likelihood of data aggregation complexities.  e.g. looking at 24 hour data is less effective than looking at 5 minute aggregation levels.  The following two links may be beneficial too.

Server Visibility: https://docs.appdynamics.com/appd/24.x/24.4/en/infrastructure-visibility/server-visibility

Troubleshooting Applications: https://docs.appdynamics.com/appd/24.x/24.4/en/application-monitoring/troubleshooting-applications

View solution in original post

iamryan
Community Manager
Community Manager

Hi @Gustavo.Marconi,

I reached out to a few people and Anderson B. should be getting in touch with you about this. 

0 Karma

Troy_Partain
Engager

Hi Gustavo,

Excellent question and I can appreciate the interest in the discrepancies.

With demo systems, we're generally not dealing with live data. The way it's generated can vary and, at times, can cause abnormalities within the context of the data. This is what is going on here.  Beyond this, in production, it's advisable to ensure your looking at the most specific time range possible to reduce the likelihood of data aggregation complexities.  e.g. looking at 24 hour data is less effective than looking at 5 minute aggregation levels.  The following two links may be beneficial too.

Server Visibility: https://docs.appdynamics.com/appd/24.x/24.4/en/infrastructure-visibility/server-visibility

Troubleshooting Applications: https://docs.appdynamics.com/appd/24.x/24.4/en/application-monitoring/troubleshooting-applications

Gustavo_Marconi
Path Finder

Hi @Ryan.Paredez  and @Troy.Partain ,

Thank you for the reply, that clarifies the issue for me, I'll be more careful with my demo presentations in the future, especially with potential customers.

Hope you both have a great day!

0 Karma
Get Updates on the Splunk Community!

Splunk Answers Content Calendar, June Edition

Get ready for this week’s post dedicated to Splunk Dashboards! We're celebrating the power of community by ...

What You Read The Most: Splunk Lantern’s Most Popular Articles!

Splunk Lantern is a Splunk customer success center that provides advice from Splunk experts on valuable data ...

See your relevant APM services, dashboards, and alerts in one place with the updated ...

As a Splunk Observability user, you have a lot of data you have to manage, prioritize, and troubleshoot on a ...