Files
DataStack/sdk-python
Team Dark Devil 8132087085 feat(auth): enhance security and audit logging
- Integrate webhook alerts for admin login lockouts.
- Implement detailed audit logging for login, logout, and setup actions.
- Improve client revocation event tracking.
- Standardize error handling in profile updates.
- Enforce 'operator' role for event API access.
2026-11-21 00:06:55 +06:00
..

DataStack Python SDK (datastack-sdk)

License: MIT Python 3.8+ Maintained by Vib Tools

Official Python client library for DataStack — Self-Hosted Real-Time Telemetry & Event Logging Platform, developed by Vib Tools.

datastack-sdk is a compatibility alias for the canonical datastack package located at ../sdk/python/datastack. Both from datastack import DataStack and from datastack_sdk import DataStackClient share a single source of truth with config.datastack.py auto-discovery, CLIENT_ID wiring, Socket.IO streaming, automatic HTTP fallback, offline disk spooling, and direct .csv/.json/.jsonl file streaming.


Installation

Install the canonical package directly from sdk/python/datastack:

pip install ./sdk/python/datastack

Or install the compatibility alias from sdk-python:

pip install ./sdk-python

Quick Start

from datastack_sdk import DataStackClient

# 1. Initialize the client with a signed JWT token from /vcon/clients or /vcon/sdk
client = DataStackClient(
    server_url="http://localhost:4000",
    api_url="http://localhost:3000",
    token="YOUR_CLIENT_JWT_TOKEN",
    client_id="YOUR_CLIENT_ID",  # optional (auto-derived from JWT sub/clientId)
)

# 2. Connect to the DataStack Socket.IO server (or auto HTTP fallback)
if client.connect():
    # 3. Stream a single structured event and wait for server acknowledgment
    res = client.sync(
        data={
            "event_type": "service_heartbeat",
            "service": "billing-worker",
            "latency_ms": 18.4,
            "status": "healthy",
        },
        timeout=5.0,
    )
    print("Sync result:", res)

    # 4. Stream a batch of events in a single round-trip
    batch_res = client.batch_sync(
        records=[
            {"metric": "cpu_load", "value": 42.1, "unit": "%"},
            {"metric": "mem_usage", "value": 68.5, "unit": "%"},
            {"metric": "disk_free_gb", "value": 124.0, "unit": "GB"},
        ],
        timeout=10.0,
    )
    print("Batch result:", batch_res)

    # 5. Disconnect cleanly
    client.disconnect()

API Reference

DataStackClient(server_url="http://localhost:4000", token="", reconnection_attempts=5, reconnection_delay=1)

Creates a new Socket.IO client instance configured for DataStack handshake authentication.

client.connect(wait_timeout=5) -> bool

Establishes a WebSocket/Socket.IO connection to the DataStack Port 4000 ingestion server using auth={"token": self.token}. Returns True on success or False on connection failure.

client.sync(data, event_id=None, timeout=5.0) -> dict

Emits a single event (sync_data) and blocks via threading.Event.wait(timeout) until the server responds.

  • Returns: {"success": True, "eventId": "evt_..."} or {"success": False, "eventId": "evt_...", "error": "..."}.

client.batch_sync(records, timeout=10.0) -> dict

Emits a list of event dictionaries (batch_sync) and blocks via threading.Event.wait(timeout) until the server acknowledges the batch.

  • Returns: {"success": True, "processed": N, "rejected": M}.

client.disconnect()

Cleanly closes the active Socket.IO connection.


Canonical Documentation Suite

For exhaustive documentation on configuration auto-discovery, HTTP fallback, offline disk spooling, direct file streaming, and error handling, consult the canonical documentation:


Author & Support

Developed and maintained by Vib Tools:

License

Released under the MIT License.