Valentin Kovalenko

@sunnycalsoft The presented `com.mongodb.MongoTimeoutException`s originating from `DefaultConnectionPool` don't by themselves signify a driver problem. They happen when the pool is asked for a connection, but does not have enough [available](https://github.com/mongodb/specifications/blob/master/source/connection-monitoring-and-pooling/connection-monitoring-and-pooling.rst#marking-a-connection-as-available-internal-implementation)...

I have a suspicion that `BookingDatesRepository` works on top of an ACID DBMS (most likely PostgreSQL in the case of this project), and this DBMS has a single node that...

> Are you talking about to handle this lock inside Postgres instead of Redis? Yes, you may be able to do this even with the guarantees given to you by...

If you don't want transactions to be failing as a result of concurrent requests (this is a downside of MVCC), you can use explicit advisory PostgreSQL locks. Centralized LM implemented...

> the lib that I'm using already supports this lock Just make sure that the locks you acquire (whether at session level or transaction level) are acquired in the same...

@toktarev > Why we can't send all requests with the same roomId on the same node ? Alex, there are no multiple nodes here, there is a single writable PostgreSQL...

@juliofalbo > The way that I found to do this was using the Isolation level (SERIALIZABLE). And I didn't like the solution, since I need to lock the whole table....

@ceki Besides having the explicitly called apply method (which merges MDC's of two threads such that the transferred MDC has precedence), `TransferableMdc ` also has an implicitly called (by try-with-resources...

I am assuming that "_this message should not bear data for "client" but probably will_" in the third task is the specific problem you are referring to. Executions of all...

> By later time, I meant ... a recycled thread In this case the situation is analogous to a single-threaded executor with all submits happened from the same thread (with...