This script is reading from yuewiktionary and writing to extension1.wiktionary_cognate.
There could be something wrong with the abstraction waiting for rep lag while using 2 different clusters?
Or some replag?
addshore@mwmaint1002:~$ mwscript extensions/Cognate/maintenance/populateCognatePages.php --wiki yuewiktionary
Started processing.
1 rows processed.
Pass finished.
[Tue Jan 22 15:40:54 2019] [hphp] [93233:7f0aad4fc3c0:0:000001] [] SlowTimer [60000ms] at runtime/ext_mysql: slow query: SELECT MASTER_GTID_WAIT('0-171966669-4075108480,171966669-171966669-3589226378,171974792-171974792-120056328,180359174-180359174-94123433,180363367-180363367-134174373', 60)
0 rows processed.
Pass finished.
[Tue Jan 22 15:41:54 2019] [hphp] [93233:7f0aad4fc3c0:0:000002] [] SlowTimer [60000ms] at runtime/ext_mysql: slow query: SELECT MASTER_GTID_WAIT('0-171966669-4075108480,171966669-171966669-3589226378,171974792-171974792-120058635,180359174-180359174-94123433,180363367-180363367-134174373', 60)
0 rows processed.
Pass finished.
♥[Tue Jan 22 15:42:02 2019] [hphp] [93234:7f0aad4fc3c0:0:000001] [] Lost parent, LightProcess exiting
[Tue Jan 22 15:42:02 2019] [hphp] [93235:7f0aad4fc3c0:0:000001] [] Lost parent, LightProcess exitingThis script has worked in the past with no problem.
The code running the query that times out is:
$loadBalancerFactory->waitForReplication();
I last ran this in Feb 2018 with no such issues https://tools.wmflabs.org/sal/log/AWGZVMQuL7tQ11ghvPqn