NetDiagnostics

• Analyze the performance status of business transactions – slow, very slow, or failed (due to errors).
• This can be viewed right at the tier, server, or instance view level.
• NDE helps to pin-point to the specific component/s causing performance issues.
• NDE also provides insight into business transaction health trend allowing engineering teams to proactively correct issues.

• Quickly check the health of your VMware host by monitoring CPU, memory utilization, number of virtual machines configured and running, and much more.

• Monitor and analyze availability and performance of the integration points, its thread pool usage and tuning.
• Get early warning before the threads used by an integration point get busy.

• Analyze server for the Ethernet throughput, TCP/IP connections, network errors, max connections to a single IP.
• This leads to finding the right solution to issues, e.g. need to add more IPs.

• Analyze CPU and the cause of its high usage. Determine whether the CPU time is in user space or kernel space.
• NDE provides insight into the root cause of high CPU, whether it is because of Java instance or because of some other parameters like garbage collection, batch jobs, Disk I/O, and high memory consumption etc.

• Evaluate load balancing across multiple data centers and between servers.
• Erratic load balancing can cause a server to go down because of unprecedented high load.

• Determine memory leak or any other heap issues and help re-evaluate the code and design of the application, which leads to memory leak and doesn’t allow garbage collection to clear up the unused objects from the heap.

• Identify the outlier metrics and pin-point the root cause of the slow response. Hotspot highlights where ever there is too much waiting in the response or a blocked-response.
• NDE Hotspots feature points to specific thread or method causing the slow response time.

• Analyze DB queries, number of active connections, leak connections, CPU, and the locks.
• This will lead to a conclusion whether DB is a cause of slow response.
• This upon fix can be compared for optimization.

• Determine whether the application design should be re-evaluated or that the disk needs to be partitioned appropriately for usage by other systems and users.

• Analyze GC metrics to determine if it is the cause of performance degradation.
• This helps to fine tune to the right GC setting for optimum performance.

• Capture detailed low level run-time information on how JVM and java applications are behaving.

• The root cause can be determined by pattern matching spike with other metrics showing the tendency to vary together.
• For example, the high response time could be because of a thread hotspot.

Analyze and verify the state of threads, whether they are running or stuck (wait or deadlock). For example, if a thread is stuck on a particular method then it affects the server response time. Thread dumps can be:

• Alert based,
• On-demand, and
• Schedule based

• Analyze the threads as per configuration along with their status – hogging or in blocked state.
• Thread pool gives information about idle thread count, total thread count that the application can handle, pending user requests.

• Analyze whether the network cache is offloading properly or not.
• NDE helps determining whether the network cache is able to manage appropriate traffic and divert the rest to the origin server.

• NDE provides easy to use comprehensive set of template based reports capturing numerous metrics, allowing users to export them in different formats.
• Visual illustration of monitoring and reporting data in a highly custom-ready user-friendly dashboard.

• A reporting feature to compare and analyze improvements and optimizations between releases before and after the fix is applied.

• Provides real-time insights into performance of critical business KPIs, including sales orders, revenue, etc.