You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Proposal:
This client should allow configuring the underlying urllib3 connection pool size just like the older client: influxdata/influxdb-python#349
Current behavior:
Currently this client throws the following error if there are more than 20 connections made: WARNING:urllib3.connectionpool:Connection pool is full, discarding connection: eu-central-1-1.aws.cloud2.influxdata.com
the default connection_pool_maxsize is multiprocessing.cpu_count() * 5, which in my case on a raspberry pi 4 evaluates to 20. Not enough when one is making 50+ threads to send data to influx. Use case: a ping tool which pings 50 hosts parallely and sends results to influx (so its a very valid use case).
Proposal:
This client should allow configuring the underlying urllib3 connection pool size just like the older client:
influxdata/influxdb-python#349
Current behavior:
Currently this client throws the following error if there are more than 20 connections made:
WARNING:urllib3.connectionpool:Connection pool is full, discarding connection: eu-central-1-1.aws.cloud2.influxdata.com
Desired behavior:

Connection pool size should be configurable. E.g. in the old influx-python client there is a
pool_size
argument in the client's constructor:https://influxdb-python.readthedocs.io/en/latest/api-documentation.html?highlight=pool#influxdb.InfluxDBClient
Use case:
This is necessary to avoid the above error and allow multi threaded clients to hit influx db 2.0 cloud properly
The text was updated successfully, but these errors were encountered: