This topic describes variables and panels on the Fluid control plane and JindoRuntime cache dashboards. Dashboard variables let you adjust observability dimensions such as monitoring interval and dataset namespace; panels help you assess component health and cache performance so you can detect issues and optimization opportunities early.
-
Fluid control plane dashboard — monitors the health and performance of Fluid's control plane components: dataset controller, runtime controller, webhook, and CSI plug-in.
-
Fluid JindoRuntime cache dashboard — monitors the cache efficiency and resource usage of a specific JindoRuntime cache system.
Detect component failures, diagnose cache performance issues, and spot optimizations before they affect workloads.
Prerequisites
Before you begin, ensure that you have:
-
Managed Service for Prometheus is enabled for the Fluid component. See Step 2: View the Fluid dashboard.
Fluid control plane dashboard
Dashboard variables
Variables control the scope and granularity of data across all panels. Changing a variable updates all related panels.
| Variable | Valid values | Description |
|---|---|---|
interval |
1m, 5m, 10m, 30m, 1h, 6h | The monitoring cycle duration. Shorter intervals show finer-grained trends; longer intervals smooth spikes. |
quantile |
0.5, 0.75, 0.90, 0.95, 0.99 | The percentile for latency and processing-time panels. For example, 0.90 = P90. |
runtime |
JindoRuntime, AlluxioRuntime, JuiceFSRuntime | The runtime type to monitor. Filters all runtime-related panels to the selected type. |
Runtime types:
-
JindoRuntime — the JindoFS execution engine, developed by the Alibaba Cloud Elastic MapReduce (EMR) team. Built in C++, provides dataset management, caching, and OSS support.
-
AlluxioRuntime — the open-source Alluxio execution engine. Supports dataset management, caching, and accelerated access to PVCs, Ceph, and Cloud Parallel File System (CPFS). Suited for hybrid cloud scenarios.
-
JuiceFSRuntime — a distributed cache acceleration engine based on JuiceFS. Supports scenario-specific caching and acceleration. See Introduction to JuiceFS.
Panels
Four panel groups: start with Component running status for a health check, then drill into Fluid Controller Detailed Indicator or Fluid webhook detailed indicators for root cause. Resource usage provides CPU, memory, and network data for all controller pods.
Component running status
Shows whether each Fluid component is running and how often pods restart. Frequent restarts signal instability.
| Panel | Description |
|---|---|
| Dataset Controller Ready Replicas | Dataset controller pods in Running state. If below expected replica count, dataset operations may stall. |
| History of Dataset controller restarts | Restart count of dataset controller pods. |
| Runtime Number of ready copies of controller | Runtime controller pods in Running state. |
| History Runtime Controller Restart Times | Restart count of runtime controller pods. |
| Fluid Webhook ready copies | Fluid webhook pods in Running state. |
| Number of historical fluid Webhook restarts | Restart count of Fluid webhook pods. |
| Fluid CSI Plug-in Ready Copies | Fluid CSI plug-in pods in Running state. |
| Historical Fluid CSI plug-in restarts | Restart count of Fluid CSI plug-in pods. |
| Fluid Component Restart | Top five Fluid components by restart count in the last 2-minute cycle. Identifies which component needs attention. |
Fluid Controller Detailed Indicator
Internal performance metrics of the runtime and DataLoad controllers. Check when dataset reconciliation slows or API server load increases.
| Panel | Description |
|---|---|
| Runtime Controller processing time | Runtime controller processing time per cycle, as percentile values. Sustained high values may indicate overload. |
| Number of Runtime controller processing failures | Failure types and counts during runtime resource handling: deployment failures and health check failures. Non-zero values require investigation. |
| Runtime Number of controller threads | Active and maximum thread count of the runtime controller. If active threads approach the maximum, the controller may bottleneck. |
| DataLoad Controller Threads | Active and maximum thread count of the DataLoad controller. |
| Controller Queue Length | Workqueue length of each Fluid controller. A growing queue indicates the controller is falling behind on reconciliation. |
| Total number of Kubernetes API requests | Total requests from all Fluid controller pods to the Kubernetes API server per cycle. Sudden spikes may cause API server throttling. |
| Runtime Controller Kubernetes API requests | Runtime controller requests to the Kubernetes API server, by HTTP status code. High 4xx or 5xx rates point to misconfiguration or permission issues. |
| Total time consumed by unfinished processing of controller | Cumulative time each Fluid controller spent on in-progress tasks. Persistently high values suggest stuck tasks. |
Fluid webhook detailed indicators
Monitors the Fluid webhook, which intercepts pod creation requests to inject FUSE sidecars. Webhook latency directly affects pod startup time.
| Panel | Description |
|---|---|
| Fluid Webhook Pod CPU Usage | CPU utilization of each Fluid webhook pod per cycle. |
| Fluid Webhook Pod Memory Usage | Memory usage of each Fluid webhook pod per cycle. |
| Total number of requests processed in Fluid Webhook | Total requests handled by the Fluid webhook per cycle. |
| The number of requests processed in each Fluid Webhook Pod | Requests per Fluid webhook pod per cycle. Spots load imbalance across replicas. |
| Fluid Webhook Request Processing Delay | Request processing latency of the Fluid webhook, as percentile. High P99 slows pod startup cluster-wide. |
| Request processing delay of each Fluid Webhook Pod | Per-pod request processing latency, as percentile. Identifies slow pods causing tail latency. |
Resource usage
CPU, memory, and network metrics for all Fluid controller pods. Use to detect resource pressure behind issues shown in other panel groups.
| Panel | Description |
|---|---|
| CPU usage | CPU utilization of each Fluid controller pod per cycle. |
| Memory usage | Memory usage of each Fluid controller pod per cycle. |
| Network Send Rate per Pod | Network transmit rate of each Fluid controller pod per cycle. |
| Network Receive Rate per Pod | Network receive rate of each Fluid controller pod per cycle. |
Fluid JindoRuntime cache dashboard
Dashboard variables
Select a dataset by namespace and name to scope all panels to that dataset's cache system.
| Variable | Description |
|---|---|
namespace |
Namespace of the target dataset. |
fluid_dataset |
Name of the target Fluid dataset. |
Panels
Three panel groups: start with Dataset overview to confirm cache pod health, then check Cache system metrics for cache efficiency and bandwidth. For FUSE-level issues such as high application latency, use FUSE metrics to isolate the cause.
Dataset overview
| Panel | Description |
|---|---|
| Ready Pod Num | Ready pod count per component (master, worker, FUSE) of the selected cache system. |
| Pod Overview | Per-component pod info: restart count (last hour), CPU and memory requests and limits. |
Cache system metrics
Core cache health indicators: capacity usage, data-serving effectiveness, and bandwidth delivered to applications.
| Panel | Description |
|---|---|
| Cache Capacity Usage (%) | Proportion of cache capacity currently in use. |
| Cache Capacity Usage | Maximum and current cache capacity in absolute values. |
| Cache Hit Ratio Per Minute | Per-minute cache hit rate of the selected cache system. |
| Read Bytes Per Minute | Per-minute reads split into cache hits (Cache Hit) and misses from backend storage (From Backend). High From Backend share means most reads bypass cache. |
| Cache System Aggregated Bandwidth |
Total outbound traffic across all worker pod network interfaces — the bandwidth the cache system delivers to applications.
Note
If worker pods run on the host network, this value may be inflated. For accurate readings, run worker pods on the container network. |
| Cache Worker Pod Network I/O |
Per-worker-pod network I/O.
Note
If worker pods run on the host network, this value may be inflated. For accurate readings, run worker pods on the container network. |
| Cache System Pod Memory Usage | Memory usage of master and worker pods. If workers use process memory as cache medium, cache capacity is included. |
| Cache System Pod CPU Usage by Cores | CPU usage of master and worker pods. |
| Aggregated File Operation Requests | Aggregated frequency of file metadata operations. Only GetAttr and ReadDir are counted. |
FUSE metrics (via CSI)
Monitors FUSE pods injected via the CSI driver. Check when applications report high file-access latency or slow metadata operations.
| Panel | Description |
|---|---|
| FUSE Network I/O |
Per-FUSE-pod network I/O.
Note
If a FUSE pod runs on the host network, this value may be inflated. For accurate readings, run FUSE pods on the container network. |
| FUSE Memory Usage/Limit (%) | Memory usage as percentage of limit per FUSE pod. Empty if no limit is set. |
| FUSE CPU Throttled Percent | CPU throttling percentage per FUSE pod. Empty if no limit is set. |
| Meta Ops Per Second | Per-second metadata operations (GetAttr, ReadDir, Open) per FUSE pod. |
| Meta Ops P99 Latency | P99 latency of metadata operations (GetAttr, ReadDir, Open) per FUSE pod. |
| Read/Write Ops Per Second | Per-second read/write operations per FUSE pod. |
| Read/Write Ops P99 Latency | P99 latency of read/write operations per FUSE pod. |
FUSE metrics (via sidecar)
Monitors FUSE sidecar containers injected into application pods. Equivalent to the CSI metrics, but scoped to sidecars.
| Panel | Description |
|---|---|
| FUSE Memory Usage/Limit (%) | Memory usage as percentage of limit per FUSE sidecar. Empty if no limit is set. |
| FUSE CPU Throttled Percent | CPU throttling percentage per FUSE sidecar. Empty if no limit is set. |
| Meta Ops Per Second | Per-second metadata operations (GetAttr, ReadDir, Open) per FUSE sidecar. |
| Meta Ops P99 Latency | P99 latency of metadata operations (GetAttr, ReadDir, Open) per FUSE sidecar. |
| Read/Write Ops Per Second | Per-second read/write operations per FUSE sidecar. |
| Read/Write Ops P99 Latency | P99 latency of read/write operations per FUSE sidecar. |