Understanding Event Grid Metrics

Question

Understanding Event Grid Metrics

Mark Pope 50

I've been publishing 4k messages/min for several days. Metrics show no event grid chart

failures for delivery. The published and delivered graphs showed uniform lines. Recently I'm getting sawtooth patterns. My test clients consuming data show accurate message counts. Is this a problem, is there ways to investigate the erratic delivery? Here is my chart

Sander van de Velde | MVP 36,146 Reputation points MVP

2025-04-17T15:33:09.51+00:00

Hello @Mark Pope,

This is a strange looking chart indeed.

How do you control the data is coming in evenly distributed over time?

Can you see improvements over a longer period?
Mark Pope 50 Reputation points

2025-04-17T15:48:01.6366667+00:00

I'm publishing data to Event Grid from a commercial MQTT broker. Event Grid has three clients and three topic spaces the filter mqtt messages for each client(a/#, b/#, c/#). Message quantities differ for each client but not by much. I'm using three instances of MQTT Explorer as my test clients. I think they may be the problem as I discovered dropped sessions chart mirrors the sawtooth pattern. I'll restart and report later today.
Sander van de Velde | MVP 36,146 Reputation points MVP

2025-04-17T19:38:14.32+00:00

Hello Mark,

So you forward messages from an MQTT broker to the Azure EventGrid Namespace MQTT broker?

Just to make ingest as simple as possible, can you try to connect with native MQTT clients as seen in this post.
Mark Pope 50 Reputation points

2025-04-17T22:00:10.9+00:00

Hi Sander, I used that post to create and configure my POC. Thanks for that. I restarted my 3 clients and the sawtooth pattern disappeared. I don't think MQTT Explorer was designed to sustain high workloads over extended time. Your help is much appreciated. Look for my next question :-)

Accepted answer

0 additional answers

Your answer

Sander van de Velde | MVP 36,146 Reputation points MVP

2025-04-17T15:33:09.51+00:00

Hello @Mark Pope,

This is a strange looking chart indeed.

How do you control the data is coming in evenly distributed over time?

Can you see improvements over a longer period?
Mark Pope 50 Reputation points

2025-04-17T15:48:01.6366667+00:00

I'm publishing data to Event Grid from a commercial MQTT broker. Event Grid has three clients and three topic spaces the filter mqtt messages for each client(a/#, b/#, c/#). Message quantities differ for each client but not by much. I'm using three instances of MQTT Explorer as my test clients. I think they may be the problem as I discovered dropped sessions chart mirrors the sawtooth pattern. I'll restart and report later today.
Sander van de Velde | MVP 36,146 Reputation points MVP

2025-04-17T19:38:14.32+00:00

Hello Mark,

So you forward messages from an MQTT broker to the Azure EventGrid Namespace MQTT broker?

Just to make ingest as simple as possible, can you try to connect with native MQTT clients as seen in this post.
Mark Pope 50 Reputation points

2025-04-17T22:00:10.9+00:00

Hi Sander, I used that post to create and configure my POC. Thanks for that. I restarted my 3 clients and the sawtooth pattern disappeared. I don't think MQTT Explorer was designed to sustain high workloads over extended time. Your help is much appreciated. Look for my next question :-)

Answer 1

Sander van de Velde | MVP 36,146 MVP

Hello @Mark Pope ,

welcome to this moderated Azure community forum.

Let me recap the outcome in an answer.

Based on the conversation it seems the test tool for generating the MQTT messages seems the source of this erratic metrics:

I don't think MQTT Explorer was designed to sustain high workloads over extended time.

By using native MQTT clients as seen in this post, the sawtooth pattern disappeared.

If the response helped, do "Accept Answer". If it doesn't work, please let us know the progress. All community members with similar issues will benefit by doing so. Your contribution is highly appreciated.

Ranashekar Guda 1,275 Reputation points Microsoft External Staff

2025-04-21T13:55:25.8466667+00:00

Hello @Mark Pope,

Just checking in to see if the above answer provided by @Sander van de Velde | MVP helped.

if you have any further query do let us know.

Share via

Understanding Event Grid Metrics

0 additional answers

Your answer