Release Date: 2018-02-13
Note: This release fixed the bug with socket writing added in Pgpool-II 3.7.0, 3.6.6 and 3.5.10. Due to this bug, when the network load is high, an illegal message may be sent to the frontend or backend. All users using 3.7.x, 3.6.6 or later, 3.5.10 or later versions of Pgpool-II should update as soon as possible.
Allow to build with libressl. (Tatsuo Ishii)
See [pgpool-hackers: 2714] for more details. Patch by Sandino Araico Sanchez.
Set TCP_NODELAY and non blocking to frontend socket. (Tatsuo Ishii)
TCP_NODELAY is employed by PostgreSQL, so do we it.
Change systemd service file to use STOP_OPTS=" -m fast". (Bo Peng)
Change pgpool_setup to add restore_command in recovery.conf. (Bo Peng)
Fix writing transaction flag is accidentally set at commit or rollback. (Tatsuo Ishii)
Throw a warning message when failover consensus settings on watchdog nodes differs. (Muhammad Usama)
Doc: Fix document typo. (Bo Peng)
Fix bug with socket writing. (Tatsuo Ishii)
pool_write_flush() is responsible for writing to sockets when pgpool's
write buffer is full (this function was introduced in 3.6.6 etc). When
network write buffer in kernel is full, it does retrying but it forgot
to update the internal buffer pointer. As a result, broken data is
written to the socket. This results in variety of problems including
too large message length.
Fix pgpool child process segfault when ALWAYS_MASTER is on. (Tatsuo Ishii)
If following conditions are all met pgpool child segfaults:
Streaming replication mode.
fail_over_on_backend_error is off.
ALWAYS_MASTER flags is set to the master (writer) node.
pgpool_status file indicates that the node mentioned in #3 is in down status.
Fix segfault when %a is in log_line_prefix and debug message is on. (bug 376) (Tatsuo Ishii)
Fix per node health check parameters types. (Tatsuo Ishii)
Fix queries hanging in
parse_before_bind with extended protocol and replication + load-balancing. (bug 377) (Tatsuo Ishii)