AnsweredAssumed Answered

APM Team Center is showing bad performance

Question asked by Tmcmm93 on Jun 15, 2018
Latest reply on Jun 22, 2018 by SergioMorales

We're experiencing slow performance on TeamCenter, sometimes it takes up to 20-30 secs to view groups of components in the map.

According to Ca documentation:

Some of the Top 5 most important new metrics for APM Team Center are:

1. Metric Name: Number of Active Workers Metric Path:

Enterprise Manager|ApplicationTriageMap:Number of Active Workers What does it measure: Indicates how many threads are currently used from common thread pool of APM Team Center. It can point to overloading issues. The maximum value is 8 and it should be below 2 or 3.

2. Metric Name: Database:Average Response Time (ms) Metric Path:

Enterprise Manager|ApplicationTriageMap|Database:Average Response Time (ms) What does it measure:
Indicates what is the average time of data processing in APM Database for APM Team Center.
It can point to a bad APM Database condition.

3. Metric Name: REST API:Average Response Time (ms) Metric Path:

Enterprise Manager|ApplicationTriageMap|REST API:Average Response Time (ms) What does it measure:
Indicates what's the average response time of all REST API resources for APM Team Center User Interface It can point that users are affected by a slow backend. Values under 1000ms (1s) are fine, for more than 5000ms, investigate REST API:Concurrent Invocations and REST API:Responses Per Interval to determine if the system is overloaded by requests.

4. Metric Name: Database:Edge Count Metric Path:

Enterprise Manager|ApplicationTriageMap|Database:Edge Count What Does It Measure:
Indicates how many dynamic changes we are tracking for topology of ATC in APM Database. It shows a size of environment. It grows up continuously because the Database stores expired entities as well.

5. Metric Name Ontology engine:Average Process Time (ms) Metric Path:

Enterprise Manager|ApplicationTriageMap|Ontology engine:Average Process Time (ms) What Does it Measure: Indicates what's the average time of processing incoming static and dynamic topology changes for ATC. It can point to a problem with processing of changes. There can be some peaks, but average values should be much lower.


So we're guessing it's a database issue, as anyone experienced this kind of behaviour? What can we do in order to increase the performance on the database side ? We're attaching the images on a 7 day scale and you can see the degradation of the database as we add more and more correlation in the appmap.