[Python] InfluxDB 2.0 Client Stops Writing After Few Hours (Notifications Still Received,...

I’m using InfluxDB 2.0 in Docker along with a custom notification receiver (influx-adapter-api container) that writes to InfluxDB using the Python influxdb-client in synchronous mode.

• The setup works fine for a few hours, but then:
• Notifications are still received correctly
• Writes to InfluxDB silently stop (no errors, no logs, no exceptions)
• Restarting the influx-adapter-api container makes writing work again
• InfluxDB itself is never restarted and seems unaffected

So it seems likely the issue is with the InfluxDB client or the runtime inside the influx-adapter-api container.

What I’ve Tried

1. -Switching between SYNCHRONOUS and ASYNCHRONOUS modes
2. -Reconnecting every hour
3. -Handling session timeouts manually
4. -Verifying the container continues receiving real-time data
5. -Removing excess logging

Question:

• 
  What could cause the InfluxDB Python client or the influx-adapter-api container to silently stop writing to InfluxDB after running for some time?
  
  Could it be a client session leak, open connection timeout, thread hang, or socket exhaustion?
  
  Any known issues or recommended workarounds for long-lived InfluxDB write clients in Docker?
  

influxdb:

image: influxdb:2.0
restart: unless-stopped
expose:
- "8086"
ports:
- "8086:8086" # For testing only
environment:
- DOCKER_INFLUXDB_INIT_MODE=setup
- DOCKER_INFLUXDB_INIT_USERNAME=influxdb
- DOCKER_INFLUXDB_INIT_PASSWORD=influxdb
- DOCKER_INFLUXDB_INIT_ORG=Polito
- DOCKER_INFLUXDB_INIT_BUCKET=test
- DOCKER_INFLUXDB_INIT_ADMIN_TOKEN=password
volumes:
- influxdb_data:/var/lib/influxdb2


writer container:

influx-adapter-api:
image: influx-adapter-api:latest
restart: unless-stopped
expose:
- "8008"
volumes:
- setting:/influx-adapter-api


the code:

from influxdb_client import InfluxDBClient
from influxdb_client.client.write_api import SYNCHRONOUS
from datetime import datetime
import time, threading, logging

class InfluxWriter:
def __init__(self, url, org, bucket, username, password):
self.client = InfluxDBClient(url=url, username=username, password=password, org=org)
self.write_api = self.client.write_api(write_options=SYNCHRONOUS)
threading.Thread(target=self._periodic_reconnect, daemon=True).start()

def _periodic_reconnect(self):
while True:
time.sleep(3600)
self._reconnect()

def _reconnect(self):
self.write_api.close()
self.client.close()
self.client = InfluxDBClient(url=self.url, username=self.username, password=self.password, org=self.org)
self.write_api = self.client.write_api(write_options=SYNCHRONOUS)

def write_notification_to_influxdb(self, measurement, fields, timestamp):
data = {"measurement": measurement, "fields": fields, "time": timestamp}
try:
self.write_api.write(bucket=self.bucket, record=data)
self.write_api.flush()
except Exception as e:
if "session not found" in str(e):
self._reconnect()
self.write_api.write(bucket=self.bucket, record=data)
self.write_api.flush()
else:
logging.error(f"Write error: {e}")

Continue reading...