A.137. Release 3.5.21

Speed up failover when all of backends are down. (Tatsuo Ishii)

If all of the backend are in down status, immediately give up finding primary node regardless search_primary_node_timeout and promptly finish the failover process.

Discussion: [pgpool-hackers: 3321]

pgpool-recovery extension and pgpool_setup is now ready for the next major release PostgreSQL 12. (Tatsuo Ishii)

Fix the wrong error message "ERROR: connection cache is full", when all backend nodes are down. (bug 487) (Bo Peng)

When all backend nodes are down, Pgpool-II throws an incorrect error message "ERROR: connection cache is full". Change the error message to "all backend nodes are down, pgpool requires at least one valid node".

Avoid exit/fork storm of pool_worker_child process. (Tatsuo Ishii)

pool_worker_child issues query to get WAL position using do_query(), which could throws FATAL error. In this case pool_worker_child process exits and Pgpool-II parent immediately forks new process. This cycle indefinitely repeats and gives high load to the system. To avoid the exit/fork storm, sleep sr_check_period.

Fix write_function_list's broken default value. (Tatsuo Ishii)

Fix "not enough space in buffer" error. (bug 499) (Tatsuo Ishii)

The error occurred while processing error message returned from backend and the cause is that the query string in question is too big. Problem is, the buffer is in fixed size (8192 bytes). Eliminate the fixed size buffer and use palloced buffer instead. This also saves some memory copy work.

Fix DROP DATABASE failure. (Tatsuo Ishii)

Fix wrong variable in read_status_file() function. (bug 493) (Takuma Hoshiai)

Abort session if failover/failback is ongoing to prevent potential segfault. (bug 481, bug 482) (Tatsuo Ishii)

Fix compiler warnings. (Tatsuo Ishii)

Fix memory leak in "batch" mode in extended query. (bug 468) (Tatsuo Ishii)