Tango Database Mode¶

workflow:

Start Tango DB
      ↓
Register device in DB
      ↓
Run device server
      ↓
Connect using DeviceProxy

list of commands¶

“”"

1. TANGO_HOST=localhost:11000 uv run python -m tango.databaseds.database 2

• returns:

  • Ready to accept request

2. in another terminal export TANGO_HOST=localhost:11000 uv run scripts/2_register_devices.py

• returns:

  • Connected: MacBook-Pro-649.local:11000

  • registered: test/haadf/1

  • registered: test/microscope/1

  • property: haadf_device_address = test/haadf/1

  Done!

3. in another terminal export TANGO_HOST=localhost:11000 uv run python -m asyncroscopy.hardware.SCAN scan_instance

• returns:

  • Ready to accept request

4. in another terminal export TANGO_HOST=localhost:11000 uv run python -m asyncroscopy.ThermoMicroscope microscope_instance

uv run python -m asyncroscopy.ThermoDigitalTwin microscope_instance¶

• returns:

  • True Client connecting to [localhost:9090]... {‘haadf’: ‘test/haadf/1’, ‘AdvancedAcquistion’: None} Ready to accept request

5. client side export TANGO_HOST=localhost:11000 DeviceProxy(“test/haadf/1”) “”"

Why use Tango Database Mode?¶

Using the Tango database provides several advantages over running devices in non-database mode.

1. Centralized device registry¶

All devices are registered in a single database. Clients only need the device name, and Tango resolves where the device server is running.

2. No manual port management¶

Clients do not need to know host addresses or ports for each device. The database handles device location, avoiding the need to open and manage multiple ports.

3. Deterministic system startup¶

The system can be brought up in a reproducible way:

Start Tango DB
Register devices
Start device servers
Connect via DeviceProxy

This makes it easy to automate instrument initialization.

4. Device discovery¶

Clients can query the database to discover available devices, classes, and servers, enabling dynamic workflows.

# List all devices
import tango

db = tango.Database()
devices = db.get_device_name("*", "*")

for d in devices:
    print(d)

# List devices of a specific class
db.get_device_name("HAADF", "*")

# List all device classes
db.get_class_list("*")

# List all device servers
db.get_server_list("*")

5. Configuration via database properties¶

Device configuration (e.g., dependencies between devices) can be stored as database properties, avoiding hardcoded relationships in code.

db.put_device_property(MICRO_DEVICE,{"haadf_device_address": [HAADF_DEVICE]})

6. Supports distributed instruments¶

Device servers can run on different machines while clients connect using only the device name.

7. Scalable system architecture¶

Database mode enables building modular systems where higher-level devices orchestrate multiple lower-level devices (e.g., microscope → detectors → acquisition).

✔ In practice, this allows a workflow where the entire system can be initialized via scripts and then accessed from tools like Jupyter notebooks using simple DeviceProxy calls.