Solved

Centreon 22.04.01 - Service check results not visibles in web interface

  • 25 August 2022
  • 18 replies
  • 456 views

Badge +2

Hello.

Since the last updating of our centreon platform (IT-100, 22.04.1), some service checks are no longer visible in Ressources Status view. In addition, the number of services checked reported is not correct: 605 instead of 854.

In “Configuration > Services > Services by host” all services are present and enabled.

In centengine.log, I saw Warning and Critical check result not reported in web interface.

Manual execution of check missing is ok.

Disk space is ok.

What could be the problem ?

Best regards,

Eric

icon

Best answer by cmeschin 30 August 2022, 09:40

View original

18 replies

Userlevel 2
Badge +6

Hi Eric,

Have you tried to export configuration and restart cbd and centengine since the update?

Kind regards.

Badge +2

Helllo cmeschin.

Thank’s to reply to my topic.

Yes. Entire server restarted a lot of time since the problem appeared. And configuration has been exported many time aswell.

What it is very strange is that the number of services checked reported is not the same at each export… Today, I have 743 services (“only” 111 are missing)

 

Userlevel 2
Badge +6

Can you send me, please:

  • a screenshot of service status deployed in top right of UI to clearly understand where are 743 services
  • screenshot of “Configuration > Poller” page (anonymised)

Have you some services with particular timeperiods (not 24x7) who should be in SOFT state and not reschedule yet? If yes, try to force check.

Badge +2

Eureka!

 I’ve found the problem!

2 versions of PHP are installed on this server and each has php-fpm processes launched (they had a different listening port). I just killed the old one (PHP 7.2) and everything is all right now.

Good to know...

 

 

Badge +2

Ah bein non ! The number of services decreased again to... 578   :-(

 

 

Instead of 854 services enabled:

 

 

Here is the screenshot of “Configuration > Poller” :

 

 

 

And no, all our services are in 24x7.

Userlevel 2
Badge +6

Hi,

Can you send me the result of these commands, please?

cat /etc/os-release

# if RedHat type:

rpm -qa | grep -i centreon | egrep -v "(plugin|pack)"

# if Debian:

dpkg -l | grep -i centreon | egrep "(plugin|pack)" -v | awk '{print $2 " - " $3}'

Badge +2

Hi.

It’s CentOS 7. Server has been installed with the Centreon ISO.

Here is the result of the package version command :

centreon-broker-cbmod-22.04.0-13.el7.x86_64
centreon-widget-servicegroup-monitoring-22.04.0-5.el7.centos.noarch
centreon-widget-live-top10-cpu-usage-22.04.0-5.el7.centos.noarch
centreon-22.04.3-4.el7.centos.noarch
centreon-broker-core-22.04.0-13.el7.x86_64
centreon-connector-22.04.0-13.el7.x86_64
centreon-widget-tactical-overview-22.04.0-5.el7.centos.noarch
centreon-widget-httploader-22.04.0-5.el7.centos.noarch
centreon-gorgoned-selinux-0.0.4-1.x86_64
centreon-trap-22.04.3-4.el7.centos.noarch
centreon-broker-storage-22.04.0-13.el7.x86_64
centreon-engine-extcommands-22.04.0-13.el7.x86_64
centreon-connector-ssh-22.04.0-13.el7.x86_64
centreon-widget-single-metric-22.04.0-5.el7.centos.noarch
centreon-widget-hostgroup-monitoring-22.04.0-5.el7.centos.noarch
centreon-widget-host-monitoring-22.04.0-5.el7.centos.noarch
centreon-license-manager-22.04.0-1.el7.centos.noarch
centreon-engine-selinux-0.0.10-1.x86_64
centreon-web-selinux-0.0.8-1.x86_64
centreon-perl-libs-22.04.3-4.el7.centos.noarch
centreon-central-22.04.3-4.el7.centos.noarch
centreon-engine-daemon-22.04.0-13.el7.x86_64
centreon-connector-perl-22.04.0-13.el7.x86_64
centreon-gorgone-centreon-config-22.04.0-9.el7.centos.noarch
centreon-widget-global-health-22.04.0-5.el7.centos.noarch
centreon-widget-service-monitoring-22.04.0-5.el7.centos.noarch
centreon-broker-selinux-0.0.6-1.x86_64
centreon-common-22.04.3-4.el7.centos.noarch
centreon-web-22.04.3-4.el7.centos.noarch
centreon-release-22.04-3.el7.centos.noarch
centreon-widget-engine-status-22.04.0-5.el7.centos.noarch
centreon-license-manager-common-22.04.0-1.el7.centos.noarch
centreon-broker-cbd-22.04.0-13.el7.x86_64
centreon-gorgone-22.04.0-9.el7.centos.noarch
centreon-widget-live-top10-memory-usage-22.04.0-5.el7.centos.noarch
centreon-widget-graph-monitoring-22.04.0-5.el7.centos.noarch
centreon-poller-22.04.3-4.el7.centos.noarch
centreon-broker-22.04.0-13.el7.x86_64
centreon-engine-22.04.0-13.el7.x86_64
centreon-widget-grid-map-22.04.0-5.el7.centos.noarch
centreon-pp-manager-22.04.0-2.el7.centos.noarch
centreon-common-selinux-0.0.11-1.x86_64
centreon-auto-discovery-server-22.04.1-2.el7.centos.noarch
centreon-nrpe3-daemon-4.0.3-0.el7.centos.x86_64
centreon-database-22.04.3-4.el7.centos.noarch
centreon-clib-22.04.0-13.el7.x86_64

Badge +2

Hi.

The OS is CentOS 7.

Centreon package versions are :

 

 

Userlevel 2
Badge +6

Hi,

It seems ok by this side.
the last update was done since which version?

On 22.04, only php 8.0 must be enabled.

Can you run this command,please?

systemctl list-unit-files | grep -i php

and 

ps -ef | grep -i php

Regards

Badge +2

22.04.03 was installed over 22.04.01.

Here are the results:

  • systemctl list-unit-files | grep -i php

 

  •    ps -ef | grep -i php

 

In addition, here are the PHP installed packages:

  • yum list installed | grep -i php

 

 

Userlevel 2
Badge +6

Have you some errors in centreon-broker logs, enable debug for sql and bbdo if needed

Be careful, because it is very verbose.

You need to export configuration to take effect.

Badge +2

I turned on debug on SQL and BBDO logs but no specific problem came out. Checks are performed and committed without error.

Now, I put this logs back in ERROR view and I can see these errors:

[2022-08-29T17:08:24.040+02:00] [sql] [info] SQL: processing log of poller 'Central' generated at 1661785703 (type 5)
[2022-08-29T17:08:24.040+02:00] [sql] [info] SQL: processing log of poller 'Central' generated at 1661785703 (type 5)
[2022-08-29T17:08:24.040+02:00] [sql] [info] SQL: processing log of poller 'Central' generated at 1661785703 (type 5)
[2022-08-29T17:08:24.040+02:00] [sql] [info] SQL: processing log of poller 'Central' generated at 1661785703 (type 5)
[2022-08-29T17:08:24.725+02:00] [core] [info] main: configuration update requested
[2022-08-29T17:08:24.725+02:00] [core] [info] modules: attempt to load '/usr/share/centreon/lib/centreon-broker/50-tcp.so' which is already loaded
[2022-08-29T17:08:24.725+02:00] [core] [info] modules: attempt to load '/usr/share/centreon/lib/centreon-broker/20-unified_sql.so' which is already loaded
[2022-08-29T17:08:24.725+02:00] [core] [info] modules: attempt to load '/usr/share/centreon/lib/centreon-broker/15-stats.so' which is already loaded
[2022-08-29T17:08:24.726+02:00] [core] [info] multiplexing: engine started
[2022-08-29T17:08:30.261+02:00] [sql] [error] mysql_connection: could not store service status:  Out of range value for column 'notification_number' at row 1
[2022-08-29T17:09:00.162+02:00] [sql] [error] mysql_connection: could not store service status:  Out of range value for column 'notification_number' at row 1
[2022-08-29T17:09:05.176+02:00] [sql] [error] mysql_connection: could not store service status:  Out of range value for column 'notification_number' at row 1
[2022-08-29T17:09:15.299+02:00] [perfdata] [error] In service (176:1731), metric name empty before 'd=''...'
[2022-08-29T17:09:15.299+02:00] [perfdata] [error] invalid perfdata format in service (176:1731): equal sign not present or misplaced ''
[2022-08-29T17:09:40.366+02:00] [perfdata] [error] In service (176:1729), metric name empty before 'entag0...'
[2022-08-29T17:09:40.366+02:00] [perfdata] [error] invalid perfdata format in service (176:1729): equal sign not present or misplaced ''
[2022-08-29T17:09:50.390+02:00] [perfdata] [error] In service (93:1052), metric name empty before '24'=20...'
[2022-08-29T17:09:50.390+02:00] [perfdata] [error] invalid perfdata format in service (93:1052): equal sign not present or misplaced ''
[2022-08-29T17:09:50.393+02:00] [sql] [error] mysql_connection: could not store service status:  Out of range value for column 'notification_number' at row 1
[2022-08-29T17:10:00.215+02:00] [sql] [error] mysql_connection: could not store service status:  Out of range value for column 'notification_number' at row 1
[2022-08-29T17:10:10.238+02:00] [perfdata] [error] In service (176:1728), metric name empty before '��0...'
[2022-08-29T17:10:10.238+02:00] [perfdata] [error] invalid perfdata format in service (176:1728): equal sign not present or misplaced ''
[2022-08-29T17:10:30.297+02:00] [perfdata] [error] In service (115:1091), metric name empty before 'ns'10...'
[2022-08-29T17:10:30.297+02:00] [perfdata] [error] invalid perfdata format in service (115:1091): equal sign not present or misplaced ''
[2022-08-29T17:10:30.299+02:00] [sql] [error] mysql_connection: could not store service status:  Out of range value for column 'notification_number' at row 1
[2022-08-29T17:10:50.351+02:00] [perfdata] [error] In service (115:1089), metric name empty before 't='' 0...'
[2022-08-29T17:10:50.351+02:00] [perfdata] [error] invalid perfdata format in service (115:1089): equal sign not present or misplaced ''
[2022-08-29T17:11:25.251+02:00] [perfdata] [error] In service (115:1088), metric name empty before '��0...'
[2022-08-29T17:11:25.251+02:00] [perfdata] [error] invalid perfdata format in service (115:1088): equal sign not present or misplaced ''
[2022-08-29T17:12:50.388+02:00] [perfdata] [error] In service (93:1055), metric name empty before 'atus0...'
[2022-08-29T17:12:50.388+02:00] [perfdata] [error] invalid perfdata format in service (93:1055): equal sign not present or misplaced ''
[2022-08-29T17:12:55.403+02:00] [sql] [error] mysql_connection: could not store service status:  Out of range value for column 'notification_number' at row 1
[2022-08-29T17:13:00.127+02:00] [sql] [error] mysql_connection: could not store service status:  Out of range value for column 'notification_number' at row 1
[2022-08-29T17:13:10.151+02:00] [sql] [error] mysql_connection: could not store service status:  Out of range value for column 'notification_number' at row 1
[2022-08-29T17:13:30.205+02:00] [sql] [error] mysql_connection: could not store service status:  Out of range value for column 'notification_number' at row 1
[2022-08-29T17:14:00.194+02:00] [sql] [error] mysql_connection: could not store service status:  Out of range value for column 'notification_number' at row 1
[2022-08-29T17:14:05.209+02:00] [sql] [error] mysql_connection: could not store service status:  Out of range value for column 'notification_number' at row 1
[2022-08-29T17:14:15.235+02:00] [perfdata] [error] In service (176:1731), metric name empty before '0...'
[2022-08-29T17:14:15.235+02:00] [perfdata] [error] invalid perfdata format in service (176:1731): equal sign not present or misplaced ''
[2022-08-29T17:14:40.305+02:00] [perfdata] [error] In service (176:1729), metric name empty before '��0...'
[2022-08-29T17:14:40.305+02:00] [perfdata] [error] invalid perfdata format in service (176:1729): equal sign not present or misplaced ''

 

 

Badge +2

No specific errors in Broker logs after activation of DEBUG view for SQL and BDDO. Checks are performed and committed normally.

Now, I put these logs back in ERROR view and I can see these errors:

 

Userlevel 2
Badge +6

Hi,

Here is the error.

We have recently a similar issue with another customer.

The field “notification_number” is a smallint and in some cases it can be too small.

You can update the fields on centreon_storage database with this query:

ALTER TABLE centreon_storage.hosts MODIFY notification_number integer;

and

ALTER TABLE centreon_storage.services MODIFY notification_number integer;

After this modification, services should be displayed correctly.

 

For errors on perfdata, it's not blocking but I encourage you to check what comes back from the plugins and to filter invalid metrics with the "--filter-perfdata" option in order to improve the performance of the broker side processing.

Kind regards

 

Badge +2

Hi.

You are the best !

 

 

 

After execution of your 2 commands in MariaDB, all services checks enabled are now visible in the UI.

Problem solved, thank you.

Userlevel 3
Badge +11

To be transparent, two fixes are in progress regarding the issue :

  • The size of these columns are about to be changed to avoid such errors in the future.
  • Broker’s handling for this type of error is being improved to avoid being impacted to that extent by such errors.
Badge +3

To be transparent, two fixes are in progress regarding the issue :

  • The size of these columns are about to be changed to avoid such errors in the future.
  • Broker’s handling for this type of error is being improved to avoid being impacted to that extent by such errors.

Hello,

 

About this problem, does the version 22.04.7 or another version fix this problem ?

Userlevel 3
Badge +11

Hi @Cyril_31 

  • The first fix has been released in centreon-web 22.04.6 and is referred to as “[Core] Improved database storage to avoid blocking Broker when maximum values are reached”
  • The second one has not been released yet but will likely be part of centreon-collect 22.04.2 and should be available in the next couple of weeks.

EDIT: the first fix is enough to avoid the issue.

Reply