I am working for a customer that has a very complex .Net application. The application works as a micro service, with several .Net work process being executed and responsible for diferent funcitions of the application.
this leads to a single application having around 1500 agents spread around 15 servers.
My problem is on transaction flow. Due to this huge ammount of agents, the agent groups cannot be expanded. This is particularly difficult when analyzing the "domainless" group, it cannot be expanded and we use this group check for several services that are impacting the application.
Is there a way to change this behaviour on the transactoin flow? Are there any work arounds for this situation?
Any tips on this is much appreciated.
Have you tried breaking down the agent groups by function already? That may help you get under the limit per agent group and may make the TF more usable as well. I thought the limit was 100 instances before drill down was disabled.
Yes I have. Before we had a single group called PR_APP with 1500 agents. Now I have around 110 groups, a group for each function. We still face this problem. I have checked and none of the groups have more than 100 agents. The biggest group I could find had 73 agents.
The transaction flow got very big and complex, but now I am able to see each service individually. Is there a way of making this transaction flow less complex and still being able to isolate the services?
Would there be a best practice on how to organize such an application?
Igor, Another approach is to limit the data you're viewing. For example, drill into a specific web request(s), or select a specific user(s), then drill into Transaction flow. This might help provide a more digestible set of data.
Due to the large number of services and agents, your bumping into some limits of how AppMon visualizes this type of data. You might want to consider using the Dynatrace product, instead of the AppMon product. The Dynatrace product will provide visualization that will be easier in large scale environments like yours. Consider reaching out to your sales rep for a discussion, or just go to https://dynatrace.com/ and click on Free Trial, give it a spin for yourself.
Yes I understand that of limiting the data, but it makes the anlisys a little bit more complicated. Customer is currently undergoing a review on their application, because often it was found that some applications were reaching external services (domainless) when they shouldn't. Now we are not able to see the details on the domainless. Even though I clicked on "Focus on passing transactions" on that node, I still can't reduce data enough to get to expansion to get backt to work.
Something that you could try since this is just for problem analysis is to open the transaction flow dashlet and filter it by the agent group the makes calls to those domainless services.
Then you can uncheck the box that is at the bottom that says include correlating agents. This should only show the transaction flow for the tiers that directly connect to transactions that pass through your selected agents and should limit your view quite a bit.
Here is the definition of the correlating agents: "Check 'Include correlating Agents' to include also Agents of applications which communicate directly or indirectly with applications of selected Agents."