Server Health

Server health is one of the first areas to check when diagnosing NCache issues. A cache may appear unavailable, slow, or unstable because one or more server nodes are experiencing service failures, high resource usage, network interruptions, or configuration problems. By reviewing server health, administrators can quickly determine whether the issue is related to the cache itself, the underlying server node, or the communication between nodes and clients.

This page explains how to diagnose NCache issues by checking server availability, NCache service status, CPU and memory usage, network connectivity, cluster node status, and related logs or monitoring indicators.

Common Symptoms

You may need to review server health if you observe any of the following issues:

• A cache server node is not responding.
• A cache cannot be started or stopped from the NCache Management Center.
• Client applications cannot connect to the cache.
• Requests are slow or timing out.
• One or more nodes leave the cluster unexpectedly.
• State transfer is delayed or does not complete.
• Cache operations fail intermittently.
• Server resource usage is unusually high.
• NCache services stop unexpectedly.
• Monitoring counters or dashboards show abnormal activity.

Step 1: Verify Server Availability

Start by confirming that the server node itself is reachable. If the server is offline, unreachable, or blocked by network rules, NCache services and caches hosted on that node may not respond correctly.

Check the following:

• The server machine is powered on and accessible.
• The server can be reached by IP address or host name.
• DNS resolution is working if host names are used.
• Firewall or security rules are not blocking communication.
• The server is reachable from other cache nodes and client machines.

If the server is not reachable, resolve the network or machine-level issue before troubleshooting NCache configuration.

Step 2: Check NCache Service Status

After verifying server availability, confirm that the required NCache services are running on the server node. If the NCache Service is stopped, cache management operations and hosted cache processes on that server may not work as expected.

Check the status of the following services based on your environment:

• NCache Service: Manages cache runtime, cache lifecycle, and management operations on the server node.
• NCache Execution Service: Runs server-side tasks such as loaders, refreshers, and backing source providers.
• NCache Bridge Service: Manages bridge operations for Multi-site Replication scenarios.

If any required service is stopped, start or restart it and then verify whether the issue is resolved.

Step 3: Review Resource Usage

High CPU, memory, disk, or network usage can affect cache performance and stability. Review server resource usage to determine whether the node is under pressure.

Check the following:

• CPU usage on the server node.
• Available physical memory.
• Cache memory usage.
• Network throughput and packet loss.
• Disk space for logs or temporary files.
• Any other process consuming excessive resources.

If server resources are constrained, cache operations may become slow, client requests may time out, or the node may become unstable.

Step 4: Verify Cache and Cluster Node Status

Next, check whether the cache is running correctly on each server node. In clustered environments, one unhealthy node can affect cluster communication, state transfer, data distribution, and client connectivity.

Check the following:

• The cache is started on all expected server nodes.
• All cluster nodes are visible in the NCache Management Center.
• No node is repeatedly joining or leaving the cluster.
• State transfer is not stuck or repeatedly restarting.
• Data load balancing is not causing unexpected delays.
• Client connections are distributed across available nodes.

If a node appears unhealthy, isolate whether the problem is caused by the server, NCache service, network configuration, or cache configuration.

Step 5: Check Network and Port Connectivity

NCache relies on network communication between cache servers and between clients and servers. If required ports are blocked or the server is bound to the wrong network interface, connectivity issues can occur even if the cache is running.

Check the following:

• Required NCache ports are open.
• Client-server communication is not blocked.
• Cluster communication between server nodes is working.
• IP binding is configured correctly in multi-NIC environments.
• Server nodes are using reachable IP addresses.
• Firewalls, VPNs, or network security groups are not blocking traffic.

If the environment uses multiple NICs, verify that the correct IP address is configured for cluster communication and client-server communication.

Step 6: Review Logs and Event Information

Logs provide detailed information about service failures, node connectivity, cache lifecycle events, state transfer, and runtime errors. Use logs to confirm what happened before and during the issue.

Review the following:

• NCache server logs.
• Cache logs.
• Client logs, if the issue affects application connectivity.
• Windows Event Viewer or Linux event logs.
• Cache health alerts, if configured.
• Event tracing data, if deeper diagnostics are required.

Use timestamps to correlate events across server logs, client logs, and monitoring dashboards.

Step 7: Use Monitoring Counters

Monitoring counters help confirm whether the server node is actively processing cache operations or showing signs of abnormal behavior.

Check counters such as:

• Requests/sec
• Average operation latency
• CPU usage
• Memory usage
• Client connections
• Cache count
• Additions/sec
• Fetches/sec
• Evictions/sec
• Expirations/sec
• State transfer-related activity

If counters stop updating, increase unexpectedly, or show abnormal values, use them with logs to narrow down the cause.

See Also

Manage Services
IP Binding with Multiple NICs
Logging
Cache Health Alerts
Trace Events
Monitor Caches
NCache Counters