FineOps Version
The application monitoring dashboard monitors the operation of FanRuan applications.
Unlike the dashboard in Overview, the application monitoring dashboard allows querying detailed operation data of individual nodes.
Ensure all prerequisites are met to use the dashboards described in this document normally. For details, see Prerequisites of Using Monitoring Dashboards.
1. Log in to FineOps as the admin, select the O&M project, and choose Project Monitoring > Application Monitoring.
2. The dashboard is divided into multiple sections that display FineBI 6.1-specific indicator overview, node status overview, node details, and application information, respectively.
Dashboard Example
It provides an information overview of nodes in the selected O&M project. Indicators with anomalies will be marked in red or yellow. If you click a node name, the sections below will display detailed indicators of the node.
Dashboard Indicator Description
It is exclusive to FineBI 6.1 projects.
Application Node: FineBI - Application Node
Engine - Calculation Node: Engine - Calculation Node
Metadata Node: Engine - Metadata Node
It displays the actual physical memory occupied by the application.
Unit: GiB
The real load of the Java virtual machine (JVM) cannot be well reflected through the memory usage rate due to the garbage collection mechanism (GC).
Therefore, the load score is introduced by FanRuan to reflect the real JVM load. The GC data can be obtained through GC callbacks. Various states of the JVM load can also be accurately measured.
Memory usage rate = Currently used memory/Maximum available memory
Currently used memory: the memory size currently used (unit: bytes)
Maximum available memory: the maximum memory (unit: bytes) that can be allocated to JVM
It displays the available space of the checked disk directory.
This section displays detailed information about the specified node in the selected O&M project. Indicators with anomalies will be highlighted in red/yellow. If you click a node name, the sections below will display detailed indicators of the node.
It monitors the JVM load.
Load is a quantitative measure of application pressure calculated by FanRuan based on GC information. Compared to memory, it reflects the actual application pressure more accurately.
A high load indicates significant application pressure. Attention is required when the load exceeds 80 to check for potential abnormalities.
It tracks actual physical memory used by the application over time.
It shows the memory usage of NIO, including the currently used and the maximum available memory.
It shows the memory usage of MetaSpace, including the currently used and the maximum available memory.
It displays the maximum and currently used number of connections in the connection pool of the project's configuration database.
Note that the connection pool refers to that available to the FanRuan application, not the native connection pool of the database.
It displays the maximum and currently used number of connections in the connection pool of the project's Redis.
Note that the connection pool refers to that available to the FanRuan application, not the native connection pool of Redis.
Currently, only Redis standalone and sentinel modes are supported. Monitoring of the connection pool status of other types of Redis is not supported.
It shows the disk space usage rate for directories related to the application, including root, project, backup, tempDir, schedule, and logdb directories.
A usage rate exceeding 80% will be highlighted in red. Insufficient available disk space may affect the normal operation of the application or even cause crashes. Disk cleanup or expansion should be performed promptly.
It shows the growth trend of disk space usage for the various application directories.
If the remaining space is less than 10 GB, it may affect the normal operation of the application or even cause crashes. Disk cleanup or expansion should be performed promptly.
It displays the trend lines of the number of current active, new, and closed sessions.
They can be used to understand the concurrency peaks of the application.
It is exclusive to FineBI projects of V6.1.3 and later versions.
It monitors specific indicators of the FineBI configuration cache, including the count of hits, misses, evictions, and total visits.
It monitors the hit rate of the FineBI configuration cache.
Indicators include:
Threads Currently Used for HTTP
Max. Threads for Container HTTP (Real-Time)
Threads Currently Used for HTTPS
Max. Threads for Container HTTPS (Real-Time)
It monitors the total number of threads used by the current node of the application (including container threads and other threads). Indicators include:
Total Threads
blocked Threads
runnable Threads
time_waiting Threads
waiting Threads
Threads Currently Used for Chart
Max. Threads for Chart (Real-Time)
Current Long Query Threads for Spider Chart
Waiting Long Query Threads for Spider Chart
Max. Long Query Threads for Spider Chart
Current Short Query Threads for Spider Chart
Waiting Short Query Threads for Spider Chart
Max. Short Query Threads for Spider Chart
Spider chart thread monitoring is only available for BI applications, and the chart is empty for FR applications.
1. If Threads Currently Used for Container = Max. Threads for Container, adjust the maximum number of threads for the container.
For non-container deployment: Modify the value of maxThreads in the server.xml configuration file in /Tomcat/conf to an appropriate value (recommended range: 200 - 2000), which will take effect after a restart.
For container-based deployment, modify the value of the same parameter under Business Tool > Business Configuration Modification in FineOps, where the recommended range is the same as above.
2. If Threads Currently Used for Spider Chart = Max. Threads for Spider Chart, adjust the maximum number of threads for the chart.
For non-container deployment: Modify the value of SystemOp timizationConfig.chartDataThreadPoolNum in the FINE_CONF_ENTITY table in FineDB to an appropriate value (recommended range: 8 - 50), which will take effect after a restart.
It displays basic information about the selected application in the O&M project.
Specific indicators are provided for FineBI 6.1 projects to assist you in project monitoring.
The number of alive Engine - Metadata Node nodes
The number of terminated tasks (TaskState: FINISHED, FAILED, and CANCELLED)