Automatic plot and update of an real-time infrastructure maps with New Relic

Real-time infrastructure maps, plot, and update automatically.

After having a deep look into New Relic service maps, I found that we can plot a diagram of ours systems monitored by New Relic using service maps. But with service maps, we cannot integrate all the entities with their relationships.
NR Observability maps come with the possibility to add custom connections between entities and custom entities, but observability maps don’t provide automatic plot like service maps.

So I would like to know if there is a possibility that I have not gone through?
If not, is there a way to use New Relic service maps APIs (with nerd graph APIs, maybe) to create a custom solution for that, please?

For larges infrastructures, it will be good to have a map that points where is an incident, then allow the developers to properly going through the resolution.
And this can be a very good feature for New Relic, which could generate a lot of money :slightly_smiling_face:

@loic.nguepi1 I’m not sure if this will be an answer to what you are looking for but wanted to make sure you were aware of this feature update:

I took a look at Automap.
But automap displays only the entities related to a source entity that have experienced a change in health status within three hours after the automap was launched.
Is it possible to have a map of all the infrastructure (independently of a particular service/entity)?
As observability maps can’t be plot and update automatically, and service maps is a bit limited. conf. here

@loic.nguepi1 Thank you for the clarification. I’m going to tag @jvajda to see if they are aware if this type of map is possible. So you are aware this topic category is not monitored by our support team.

1 Like

I actually don’t know if that is possible, I’ll see if any internal teams know more about this.

1 Like

Hi, it’s possible to use the New Relic graphql api to retrieve monitored entities (service-tier and infra like hosts, containers, Kubernetes pods, etc) and the auto-discovered relationships between them. From this you could build your own map visualization but it’s quite a project.

Product manager research question: what types of infra entities are most important for you to see in the auto-discovered maps like service map and Automap?

3 Likes

Thank you for your response @jkaron, and sorry for the late reply.
Let me share with you some more details. We have many types of resources (APIs, Services, Hosts, Containers, Mobile Apps, Web Apps, Functions, Lambdas, Kafka servers, Azure VMs, etc), a large infrastructure. Currently, we maintain an infrastructure diagram manually, and we are looking at how New Relic can help us to do this automatically.
Several of our resources (most critical) are all connected with Kafka, and we are not sure New Relic can easily plot this out.
If you can give us an answer on this, please? If so, what to do?
If not, we’ll go with the ‘project’ idea. But this is first an investigation.

Thanks.

1 Like

@JoiConverse, @jvajda, @jkaron, is there an update on my last post, please?

@loic.nguepi1 Unfortunately my knowledge of Service maps is limited, so I’ve asked some internal teams if that had any ideas for you, I was hoping for some additional responses to your question.

1 Like

Hi @loic.nguepi1 , I did some more digging on this. Two things I wanted to share…

As you noted, our service maps and distributed trace maps don’t represent infrastructure-level components like Kafka queues in the flow diagrams. You can manually pass trace context as part of your message payload so that our distributed tracing engine sees the connection between the producing and consuming service. This essentially makes the queue invisible to the trace so you wouldn’t see the queue represented in a map between services.

We’re researching ways for our agents to automatically detect Kafka dependencies so that maps and relationships in NR1 show you which broker and topic sits between services. I don’t have an estimate on when this capability will ship.

4 Likes