taosX-Agent is a core component of the TDengine TSDB Enterprise edition, designed to handle edge data collection in industrial environments. It deploys close to data sources, collects data from various industrial devices and systems, and forwards it through a secure channel to the taosX service for storage in TDengine.
Core capabilities
Multi-protocol support. taosX-Agent supports OPC UA, OPC DA, MQTT, and other industrial protocols, letting it connect to a wide range of equipment without additional gateway software.
Edge computing. The agent can preprocess and filter data at the edge before transmission. This reduces bandwidth consumption and lightens the load on the central TDengine cluster by discarding irrelevant data points early.
Store-and-forward. When the network connection to taosX is interrupted, the agent automatically caches collected data in local storage. Once the connection is restored, it resends the cached data to ensure no gaps in the time series.
Secure transmission. Data is compressed and encrypted during transit, protecting sensitive industrial telemetry as it moves from edge locations to the central system.
Typical deployment scenarios
PI System data collection. Deploy the agent on the PI server itself. The PI-Connector reads data from the PI System and the agent synchronizes it to TDengine. This setup minimizes network hops and keeps the PI-to-TDengine path short.
OPC UA/DA device collection. Place the agent inside the device network to collect data directly from OPC servers. No intermediate relay or protocol converter is needed.
Network-isolated environments. In sectors such as defense or energy where security policies require strict network segmentation, deploy the agent in the DMZ. It bridges the secure operational network and the data platform without exposing internal systems directly.
Configuration Details
The configuration file defaults to /etc/taos/agent.toml on Linux.
Core Parameters
| Parameter | Description | Default |
|---|---|---|
| endpoint | taosX service gRPC endpoint (required) | – |
| token | Authentication token (required) | – |
| instanceId | Instance identifier | 1 |
| compression | Enable data compression | false |
| in_memory_cache_capacity | Memory cache capacity in MB | 64 |
| client_port_range | Client port range | 49152-65535 |
The [log] section controls logging behavior with standard settings for level, directory, and rotation.
Deployment Recommendations
Deploy as close to the data source as practical. Assess memory and CPU requirements based on expected data volume and the number of concurrent protocol connections. Verify that the agent can establish a gRPC connection to the taosX service through any firewalls between edge and central locations.
Start and manage the agent with systemctl:
systemctl start taosx-agent
systemctl status taosx-agent
systemctl enable taosx-agentData Collection Flow
The complete pipeline runs: data source devices (OPC, PI, etc.) to taosX-Agent (edge collection) to taosX (data ingestion) to TDengine (time-series database). Each stage handles a distinct responsibility: protocol adaptation at the edge, data transformation and routing in taosX, and optimized storage and query in TDengine.
Troubleshooting
Use journalctl to inspect agent logs. Check that the endpoint configuration is correct and that no firewall blocks the gRPC port. Verify that the authentication token is still valid and has not expired.
Performance Tuning
Adjust in_memory_cache_capacity based on peak data throughput. If the agent runs in a bandwidth-constrained environment, enable compression to reduce network traffic. Plan the port range to avoid collisions with other services on the same host. Monitor resource usage regularly and revisit the cache size as data volumes grow.
Summary
taosX-Agent addresses the challenges of complex network topologies and multi-protocol environments in industrial data collection. By placing collection, preprocessing, and buffering at the edge, it provides a reliable data pipeline from shop-floor equipment to the central TDengine TSDB.
