View Issue Details

IDProjectCategoryView StatusLast Update
0000071Pgpool-IIBugpublic2013-11-03 17:46
ReporterjnatafAssigned Tot-ishii 
PrioritynormalSeveritymajorReproducibilitysometimes
Status resolvedResolutionopen 
PlatformLinux centos 6.3.2OSOS Version
Product Version 
Target VersionFixed in Version 
Summary0000071: Sometime healcheck failed and backend
DescriptionSometimes pgpool healtcheck exluded one posgres backend with error.

2013-09-04 05:14:22 DEBUG: pid 7611: health_check: 0 th DB node status: 2
2013-09-04 05:14:22 LOG: pid 7611: connect_inet_domain_socket: select() interrupted
2013-09-04 05:14:22 ERROR: pid 7611: make_persistent_db_connection: connection to vm-estim-recpsql1.dmzprive.cite-sciences.fr(5432) failed
2013-09-04 05:14:22 ERROR: pid 7611: health check failed. 0 th host vm-estim-recpsql1. at port 5432 is down
2013-09-04 05:14:22 DEBUG: pid 7611: reap_handler called

we can see any errors in backend posgresql.
This error appeared with no rules (random) one each backends DB.


postgresql92-9.2.4-1PGDG.rhel6.x86_64
postgresql92-libs-9.2.4-1PGDG.rhel6.x86_64
pgpool version 3.2.4

Pgpool.conf health_check parameters
health_check_period = 20
                                   # Health check period
                                   # Disabled (0) by default
health_check_timeout = 20
                                   # Health check timeout
                                   # 0 means no timeout
health_check_user = 'healthcheck'
                                   # Health check user
health_check_password = 'healthcheck'
                                   # Password for health check user
health_check_max_retries = 10
                                   # Maximum number of times to retry a failed health check before giving up.
health_check_retry_delay = 3
                                   # Amount of time to wait (in seconds) between retries.


Have you ever seen this kind of errors ?
Did you recommend to upgrade in 3.2.5 version
Thanks

TagsNo tags attached.

Activities

t-ishii

2013-09-05 10:33

developer  

pool_connection_pool.c.patch (651 bytes)
*** pool_connection_pool.c.orig	2013-09-05 10:28:44.949127284 +0900
--- pool_connection_pool.c	2013-09-05 10:29:54.687126569 +0900
***************
*** 665,671 ****
  					pool_log("connect_inet_domain_socket: select() interrupted. retrying...");
  					continue;
  				}
! 				pool_log("connect_inet_domain_socket: select() interrupted");
  				close(fd);
  				return -1;
  			}
--- 665,671 ----
  					pool_log("connect_inet_domain_socket: select() interrupted. retrying...");
  					continue;
  				}
! 				pool_log("connect_inet_domain_socket: select() interrupted: %s retry: %d", strerror(errno), retry);
  				close(fd);
  				return -1;
  			}

t-ishii

2013-09-05 10:34

developer   ~0000320

I need more detailed info. Can you please apply attached patch and show the log next time when the error occurs?

jnataf

2013-10-29 00:09

reporter   ~0000349

Hi,
finally we upgrated our version to 3.2.6, defined host and ip in resolv.conf file
it should be a network resolve problem
thks for your response

t-ishii

2013-11-03 17:46

developer   ~0000350

Thanks for your resposnse. I have changed the Mantis status to "resolved".

Issue History

Date Modified Username Field Change
2013-09-04 23:35 jnataf New Issue
2013-09-05 10:33 t-ishii File Added: pool_connection_pool.c.patch
2013-09-05 10:34 t-ishii Note Added: 0000320
2013-09-05 10:34 t-ishii Assigned To => t-ishii
2013-09-05 10:34 t-ishii Status new => assigned
2013-09-05 10:35 t-ishii Status assigned => feedback
2013-10-29 00:09 jnataf Note Added: 0000349
2013-10-29 00:09 jnataf Status feedback => assigned
2013-11-03 17:46 t-ishii Note Added: 0000350
2013-11-03 17:46 t-ishii Status assigned => resolved