Skip to main content

Troubleshooting MBI

It is strongly advised that you install the Centreon MBI connector which allows you to monitor the status of your MBI server.

Before going further, make sure that MBI is up to date (update or upgrade it if needed).

During the daily compute, the ETL may encounter several problems:

  • database errors (mysql crashed table, disk full)
  • database connection errors (mysql timeout, DNS issues)
  • computing errors (a server is down).

If you are monitoring your MBI server with your Centreon, some of these errors will be returned by the connector.

Running a diagnostic​

Use the following command to check that MBI is properly configured:

/usr/share/centreon-bi/tools/diagnostic.sh

Expected result:

#################### Check connection to databases ####################


 --> Connection to monitoring databases:


########## Java ##########

    [OK]      Java 17 installed
    [OK]      Connection to centreon database on db successful
    [OK]      Connection to centreon_storage database on db successful

 --> Connection to reporting server:

    [OK]      Connection to centreon on db-bi successful
    [OK]      Connection to centreon_storage on db-bi successful

####################          CBIS deamon          ####################

    [OK]      CBIS is running

####################       ETL configuration       ####################

   [INFO]     Use large memory tweaks  option is disabled
    [OK]      ETL log file exists
    [OK]      ETL cron activated 

####################    Retention configuration    ####################

    [OK]      Retention file exists
    [OK]      Purge cron activate 
    [OK]      Purge option is enabled

Use this command to check the status of the CBIS service:

systemctl status cbis

Expected result:

● cbis.service - Centreon MBI Scheduler
   Loaded: loaded (/usr/lib/systemd/system/cbis.service; enabled; vendor preset: disabled)
   Active: active (running) since Wed 2025-08-06 09:55:09 IST; 41min ago
 Main PID: 584 (java)
    Tasks: 39 (limit: 24325)
   Memory: 381.4M
   CGroup: /system.slice/cbis.service
           └─584 /usr/bin/java --add-opens java.base/java.lang=ALL-UNNAMED --add-opens java.base/javax.crypto=ALL-UNNAM>

If necessary, use the following command to start CBIS:

systemctl start cbis && systemctl enable cbis

Locating missing data or partitions using the --partitions and --db-content commands​

The /usr/share/centreon-bi/etl/centreonbiMonitoring.pl script has the options --db-content and --partitions which allow you to potentially identify issues with data.

  • --db-content will indicate the date of the last data of each table that has a problem
  • --partitions will indicate the number of missing partitions between the first partition of the table and now, and since when they have been missing. Note that this doesn't work if there are many non-consecutive periods missing.

--db-content​

Check the date of the last data found in the MBI database:

/usr/share/centreon-bi/etl/centreonbiMonitoring.pl --db-content

If everything is OK, you should have the following message: ETL execution OK, database is up-to-date. However, if problems have been found, you will get something like:

[mod_bam_reporting_ba_availabilities: 2025-07-27 00:00:00] [hoststateevents: 2025-07-28 00:00:00] [servicestateevents: 2025-07-28 00:00:00] [mod_bi_hoststateevents: 2025-07-28 00:00:00] [Table mod_bi_servicestateevents: EMPTY] [mod_bi_time: 2025-07-28 00:00:00] [mod_bi_hostavailability: 2025-07-27 00:00:00] [Table mod_bi_serviceavailability: EMPTY] [Table mod_bi_hgmonthavailability: EMPTY] [Table mod_bi_hgservicemonthavailability: EMPTY] [data_bin: 2025-07-27 23:59:55] [mod_bi_metricdailyvalue: 2025-07-27 00:00:00] [Table mod_bi_metricmonthcapacity: EMPTY]

In this case, you must rebuild some of your data. For instance, in the example above, you will need to rebuild data_bin for the 28th of July.

--partitions​

You can also check that partitions are up to date (to have a more detailed description of what is missing) using the following command:

/usr/share/centreon-bi/etl/centreonbiMonitoring.pl --partitions

The command should return All partitions are up-to-date. If a problem has been found, you will get a message of this type:

[mod_bi_hostavailability, last partition:2025-07-28 00:00:00 missing 42 part.][mod_bi_serviceavailability, last partition:2025-07-28 00:00:00 missing 42 part.][mod_bi_metrichourlyvalue, last partition:2025-06-05 00:00:00 missing 67 part.][mod_bi_metricdailyvalue, last partition:2025-07-28 00:00:00 missing 42 part.][data_bin, last partition:2025-07-28 00:00:00 missing 14 part.]

In the example above, 42 missing partitions have been found for the mod_bi_hostavailability table (and the last partition was created on 28th July).

In this case, you must rebuild some of your data.

Understanding the commands' output​

If the --partitions and --db-content commands indicate that there is a problem with one of your tables, you may have to do a partial rebuild of your data. Make sure you use the correct options in all cases when rebuilding (including the --no-purge option which stops you from deleting important data).

TablesMeaningWhat to do
hoststateevents, servicestateevents,
mod_bam_reporting*, data_bin		Problem with raw data imported from Centreon.Investigate and fix the issue with raw data. (Maybe you need to compute events with eventReportBuilder on the central server). After resolving the issue, run the import script to import missing data, using the correct options (/usr/share/centreon-bi/etl/importData.pl).
mod_bi_servicemetrics,mod_bi_hosts, mod_bi_services, mod_bi_hostgroupsIssue with dimensions data.After resolving the issue, run the dimension script to restore consistency in dimensions, using the correct options (/usr/share/centreon-bi/etl/importData.pl). (/usr/share/centreon-bi/etl/dimensionsBuilder.pl)
mod_bi_*availabilityIssue with aggregated availability data, not raw data.After resolving the issue, run the availability aggregation script using the correct options (/usr/share/centreon-bi/etl/importData.pl) (/usr/share/centreon-bi/etl/eventStatisticsBuilder.pl).
mod_bi_metric*Issue with aggregated metrics data (e.g., performance), not raw data.After resolving the issue, run the metrics aggregation script using the correct options (/usr/share/centreon-bi/etl/importData.pl) (/usr/share/centreon-bi/etl/perfdataStatisticsBuilder.pl).

If any issues occur, double check your MBI is configured according to our post-installation configuration procedure.

CBIS does not start​

Use SSH to connect to your MBI reporting server and switch user to root to check the following logs:

  • /var/log/centreon-bi/cbis.out
  • /var/log/centreon-bi/cbis.date-of-the-day.log

If you find a relevant issue in one of these two files, fix the problem (e.g. error when connecting to the database) and manually restart the CBIS service:

systemctl restart cbis

Where can I find the logs?​

Use SSH to connect to your MBI server and switch to root.

Navigate to the MBI logs located in var/log/centreon-bi.

CBIS creates a new log file for each day, located in /var/log/centreon-bi, called bis.date-of-the-day.log. The date has the following format: YYYY-MM-DD.

The report I generated is empty​

When a report is empty, you should analyse it this way:

image

I can't download the generated reports​

  • Confirm that the Default publication rule using the SFTP protocol is applied and configured correctly. (The Default rule transfers generated reports from the MBI server to the central server.)

  • Check that the central server and MBI server are time-synced: the timezone and date must be exactly the same on both servers for reports to be downloadable.

timedatectl

If the timezones are different, edit the corresponding PHP file. For example, add date.timezone = Europe/Paris on both servers. You then need to restart php-fpm:

systemctl restart php-fpm
  • Check that the CBIS process starts with the same time zones as both servers. Check in the CBIS logs if the time displayed in front of the logs is consistent with the time of the server. First, restart CBIS:
systemctl restart cbis

Then check the time and the time zone by executing the command date then compare the date with the one in the logs.

tail -f /var/log/centreon-bi/cbis.YYYY-MM-DD.log

The date and time must be consistent with the one which appears in the logs.

I cannot see the report design/the hosts I need​

If you cannot see some report designs or some resources, it is possible you have not been authorized to do so by your administrator. User rights on MBI can be configured by an administrator using the Administration > ACL > MBI Options > ACL Rules page. Here, administrators can choose which report designs, jobs and job groups each user is allowed to access.