[pgpool-general: 8491] Signal 4 errros

[Ferhat Can]

Hello all,

First of all, thank you for making such a unique software! I'm using pgpool
with AWS RDS on AWS EKS with the following pgpool.conf configuration:

listen_addresses = '*'

port = 5432

socket_dir = '/var/run/pgpool'

pid_file_name = '/var/run/pgpool/pgpool.pid'

pool_passwd = '/etc/pgpool/pool_passwd'

pcp_listen_addresses = '*'

pcp_port = 9898

pcp_socket_dir = '/var/run/pgpool'

backend_hostname0 = '*primary*.rds.url'

backend_port0 = 5432

backend_weight0 = 1

backend_flag0 = 'ALWAYS_PRIMARY|DISALLOW_TO_FAILOVER'

backend_hostname1 = '*replica*.rds.url'

backend_port1 = 5432

backend_weight1 = 19

backend_flag1 = 'DISALLOW_TO_FAILOVER'

sr_check_period = 0

health_check_period = 0

enable_pool_hba = off

allow_clear_text_frontend_auth = on

backend_clustering_mode = 'streaming_replication'

num_init_children = 128

max_pool = 4

child_life_time = 300

child_max_connections = 0

connection_life_time = 0

client_idle_limit = 305

connection_cache = on

load_balance_mode = on

log_statement = off

log_connections = off

log_client_messages = off

log_per_node_statement = off

log_error_verbosity = terse

statement_level_load_balance = on

replication_mode = off

master_slave_mode = on

master_slave_sub_mode = stream

fail_over_on_backend_error = off

ssl = on

use_watchdog = off

memory_cache_enabled = off

enable_shared_relcache = off

primary_routing_query_pattern_list = '.*/\*.*NO_LOAD_BALANCE.*\*/.*'

failover_on_backend_error = off

authentication_timeout = 60

client_idle_limit = 0

The pgpool version I'm using is 4.3.3 obtained from Alpine linux
distribution, and it works great most of the time. However, sometimes I get
the following error:

"child process with pid: xxx exits with status 4 by signal 4"

Signal 4 seems to be "SIGILL" and is mainly related to hardware issues. Yet
this issue happens on all of the environments we are using and trying the
configuration settings or pgpool versions did not help. I wonder if there
is a solution to the issue since it causes the connections to be closed,
and as a result, server-side errors happen.

Thanks a lot in advance for your guidance.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.pgpool.net/pipermail/pgpool-general/attachments/20221114/89a7a3b0/attachment.htm>