Chapter 11. Maintaining Satellite Server

This chapter provides information on how to maintain a Red Hat Satellite Server, including information on how to work with audit records, how to clean unused tasks, how to recover Pulp from a full disc, and how to reclaim disc space from MongoDB.

11.1. Deleting Audit Records

Audit records are created automatically in Satellite. You can use the foreman-rake audits:expire command to remove audits at any time. You can also use a cron job to schedule audit record deletions at the set interval that you want.

By default, using the foreman-rake audits:expire command removes audit records that are older than 90 days. You can specify the number of days to keep the audit records by adding the days option and add the number of days.

For example, if you want to delete audit records that are older than seven days, enter the following command: 

# foreman-rake audits:expire days=7

11.2. Anonymizing Audit Records

You can use the foreman-rake audits:anonymize command to remove any user account or IP information while maintaining the audit records in the database. You can also use a cron job to schedule anonymizing the audit records at the set interval that you want.

By default, using the foreman-rake audits:anonymize command anonymizes audit records that are older than 90 days. You can specify the number of days to keep the audit records by adding the days option and add the number of days.

For example, if you want to anonymize audit records that are older than seven days, enter the following command: 

# foreman-rake audits:anonymize days=7

11.3. Configuring the Cleaning Unused Tasks Feature

Satellite performs regular cleaning to reduce disc space in the database and limit the rate of disk growth. As a result, Satellite backup completes faster and overall performance is higher.

By default, Satellite executes a cron job that cleans tasks every day at 19:45. Satellite removes the following tasks during the cleaning:

  • Tasks that have run successfully and are older than thirty days
  • All tasks that are older than a year

For Satellites Upgraded from Previous Versions

Until BZ#1788615 is resolved, this functionality works only on fresh installations of Satellite 6.8 and later. If you upgrade Satellite from previous versions, this functionality is disabled by default. To enable Satellite to perform regular cleaning, enter the following command: 

# satellite-installer --foreman-plugin-tasks-automatic-cleanup true

Optionally use this procedure to adjust the configuration to serve your needs.

Procedure

  1. Optional: To configure the time at which Satellite runs the cron job, set the --foreman-plugin-tasks-cron-line parameter to the time you want in cron format. For example, to schedule the cron job to run every day at 15:00, enter the following command: 

    # satellite-installer --foreman-plugin-tasks-cron-line "00 15 * * *"
  2. Optional: To configure the period after which Satellite deletes the tasks, edit the :rules: section in the /etc/foreman/plugins/foreman-tasks.yaml file.

11.4. Recovering from a Full Disk

The following procedure describes how to resolve the situation when a logical volume (LV) with the Pulp database on it has no free space.

To recover from a full disk

  1. Let running Pulp tasks finish but do not trigger any new ones as they can fail due to the full disk.

  2. Ensure that the LV with the /var/lib/pulp directory on it has sufficient free space. Here are some ways to achieve that:

    1. Remove orphaned content: 

      # foreman-rake katello:delete_orphaned_content RAILS_ENV=production

      This is run weekly so it will not free much space.

    2. Change the download policy from Immediate to On Demand for as many repositories as possible and remove already downloaded packages. See the Red Hat Knowledgebase solution How to change syncing policy for Repositories on Satellite from "Immediate" to "On-Demand" on the Red Hat Customer Portal for instructions.

    3. Grow the file system on the LV with the /var/lib/pulp directory on it. For more information, see Growing a File System on a Logical Volume in the Red Hat Enterprise Linux 7 Logical Volume Manager Administration Guide.

      Note

      If you use an untypical file system (other than for example ext3, ext4, or xfs), you might need to unmount the file system so that it is not in use. In that case, complete the following steps:

      1. Stop the satellite-maintain services: 

        # satellite-maintain service stop
      2. Grow the file system on the LV.

      3. Start the satellite-maintain services: 

        # satellite-maintain service start
  3. If some Pulp tasks failed due to the full disk, run them again.

11.5. Managing Packages on the Base Operating System of Satellite or Capsule

To install and update packages on the Satellite or Capsule base operating system, you must enter the satellite-maintain packages command.

Satellite prevents users from installing and updating packages with yum because yum might also update the packages related to Satellite or Capsule and result in system inconsistency.

Important

The satellite-maintain packages command restarts some services on the operating system where you run it because it runs the satellite-installer command after installing packages.

Procedure

  • To install packages on Satellite or Capsule, enter the following command: 

    # satellite-maintain packages install package_1 package_2

  • To update specific packages on Satellite or Capsule, enter the following command: 

    # satellite-maintain packages update package_1 package_2

  • To update all packages on Satellite or Capsule, enter the following command: 

    # satellite-maintain packages update

Using yum to Check for Package Updates

If you want to check for updates using yum, enter the command to install and update packages manually and then you can use yum to check for updates: 

# satellite-maintain packages unlock
# yum check update
# satellite-maintain packages lock

Updating packages individually can lead to package inconsistencies in Satellite or Capsule. For more information about updating packages in Satellite, see Updating Satellite Server.

Enabling yum for Satellite or Capsule Package Management

If you want to install and update packages on your system using yum directly and control the stability of the system yourself, enter the following command: 

# satellite-maintain packages unlock

Restoring Package Management to the Default Settings

If you want to restore the default settings and enable Satellite or Capsule to prevent users from installing and updating packages with yum and ensure the stability of the system, enter the following command: 

# satellite-maintain packages lock

11.6. Reclaiming MongoDB Space

The MongoDB database can use a large amount of disk space especially in heavily loaded deployments. Use this procedure to reclaim some of this disk space on Satellite.

Prerequisites

Procedure

  1. Stop Pulp services: 

    # satellite-maintain service stop --only \
pulp_celerybeat.service,pulp_resource_manager.service,pulp_streamer.service,pulp_workers.service,httpd

  2. Access the MongoDB shell: 

    # mongo pulp_database

  3. Check the amount of disk space used by MongoDB before a repair: 

    > db.stats()
  4. Ensure that you have free disk space equal to the size of your current MongoDB database plus 2 GB. If the volume containing the MongoDB database lacks sufficient space, you can mount a separate volume and use that for the repair.

  5. Enter the repair command. Note that the repair command blocks all other operations and can take a long time to complete, depending on the size of the database. 

    > db.repairDatabase()

  6. Check the amount of disk space used by MongoDB after a repair: 

    > db.stats()

  7. Exit the MongoDB shell: 

    > exit

  8. Start Pulp services: 

    # satellite-maintain service start

11.7. Reclaiming PostgreSQL Space

The PostgreSQL database can use a large amount of disk space especially in heavily loaded deployments. Use this procedure to reclaim some of this disk space on Satellite.

Procedure

  1. Stop all services, except for the postgresql service: 

    # satellite-maintain service stop --exclude postgresql

  2. Switch to the postgres user and reclaim space on the database: 

    # su - postgres -c 'vacuumdb --full --dbname=foreman'

  3. Start the other services when the vacuum completes: 

    # satellite-maintain service start