Pool creation timeout

[PaulWasTaken]

Hello, could you point me out about pool initializing.
I wonder how pool is created, but I am not that good in Cython.
I noticed that pool creation time depends on TRANSPORT_CONNECT_TIMEOUT, so I expected an log messages when this param is lower than expected. There is https://github.com/oracle/python-oracledb/blob/main/src/oracledb/impl/thick/pool.pyx#L147 here, but it is never thrown
The approach is "We want to acquire some connections (min param) beforehand (in create_pool), but if we fail, we will establish connections on requests"
Am I right?

[anthony-tuininga]

First, there are some differences between thick mode and thin mode in how pools are initialized. In thin mode, the connections are established in a background thread and failure to create these connections will be reported when the first call to pool.acquire() is made. Thick mode, however, currently establishes min connections and if any of those fails, the call to oracledb.create_pool() itself will fail. This may change at some point to use a background thread similarly to thin mode.

The value of the parameter TRANSPORT_CONNECT_TIMEOUT indicates how long to wait for the initial socket connection to be established with the listener. I'm not sure what log messages you were expecting. Can you clarify?

[PaulWasTaken]

I'm not sure what log messages you were expecting.

For some reason I thought it would be reported as soon as possible.

I was confused because the script run twice long than expected. I thought it is because of TRANSPORT_CONNECT_TIMEOUT - the connection was tried to establish two times

UPD: I figured it out - if pool.acquire() is called before timeout first we will wait for TRANSPORT_CONNECT_TIMEOUT