You can view the On-Demand recording here (pw: FpvBfrn6) and slide deck here, but here is a sampling of questions covered in the event.
Q1: Can I get some guidance on Agent installation and configuration?
A: For AppDynamics, there are many ways to install and configure an Agent. Below are the options:
- Manual configuration and installation
- Follow the getting started wizard on the Overview page
- Using automation like ansible or some other orchestration system to deploy applications and inject the agent and configuration
- If containerized and running in Kubernetes, you could use the auto instrumentation feature of the AppD cluster agent to deploy agents
- AppDynamics has a feature called SmartAgent that can install, configure and manage agents once smartagent is installed.
Documentation:
Q2: What is the best way to identify failure/error trends on a periodic basis (ex: after a software release of our internal product)?
A:
- Define the time frame you want to compare, where one is the older application code, and the second is the new release
- Create custom Dashboards and Health Rules displaying key error metrics (e.g., Error Rates, Response time, calls per minute) for your product’s overall application performance, tiers and business transactions.
- Dash Studio has a built in feature that allows you to timeline comparison using key metrics and business transactions using Widget time ranges
Documentation:
Q3: What are some best practices for filtering out “white noise”?
A: "White noise" in AppDynamics typically refers to excessive alerting especially one that are not actionable.
- Optimize Error Configuration and Slow Transaction Thresholds
- For Health Rules combine multiple conditions to get a clear technical problem definition for each monitor event (e.g., "Error Rate > 5% AND Calls Per Minute > 100")
- Fine-Tune Health Rules making sure we set conditions with realistic thresholds and proper baselines
- Use the <n> times in a time period to avoid single outliers creating monitor events.
- Optimize Business Transaction (BT) Detection and Configuration, exclude irrelevant BTs those that are not critical to your business or application performance, lock down the relevant BT and review on a regular basis.
- Don’t overdo it focus on alerts that actionable
- Take Advantage of Anomaly detection
Documentation:
Q4: Does AppDynamics support OpenTelemetry?
A:
- Yes, AppDynamics SaaS currently supports OpenTelemetry Traces and Spans
- AppDynamics On-Premise Virtual Appliance is planned to support OTEL in the near future
- AppDynamics also provides its own version of the OTEL collector that can be configured to send data to the controller OpenTelemetry end point
- In Americas, its supported in Oregon and Sao Paulo Regions
- In EMEA, its supported in Frankfurt and coming soon to London
- In APAC, its supported in Mumbai, Singapore and Sydney
Documentation:
Q5: Can you share some guidance around JVM Tuning?
A: Appdynamics recommends that in resource-intensive environments where Splunk AppDynamics features, such as asynchronous transaction tracking, developer mode that result in a high number of metrics or snapshots reported per minute can affect agent resource consumption. In these cases, its recommended to increase the Heap and PermGen settings for the agent.
- Maximum heap size (-Xmx): 100 MB in addition to the amount required by the application
- Maximum PermGen (permanent generation) heap size (-XX:MaxPermSize): 20 MB in addition to the amount required by the application. The setting for MaxPermSize is applicable only for Java 6 or Java 7. JVM option is no longer relevant in Java 8 and later versions
Documentation:
| Updated on
