Chapter 5. Backing up a Core

5.1. Overview

You can back up a Core by following this procedure. After completing the procedure and storing the backup data safely, you can restore a Core to the state at the time of backup using the Restoring a Core Backup procedure.

5.2. Requirements

  • A self-managed Core installation on an OpenShift platform
  • A local installation of the oc binary
  • The oc binary has a logged in user on the platform you wish to back up
  • The oc binary has a logged in user with permission to run the oc get pc command

5.3. Expected Storage Requirements

Other factors that have an impact on how much storage is required for backups include:

  • how often you backup
  • what compression is used
  • the length of time you store the backups

5.4. What Data is Backed Up

You must back up the following items to back up a Core:

  • Mongodb Replica Set data
  • MySQL data
  • Nagios historical data
  • Core metrics files
  • Git files
  • Core SCM files

5.5. Backing up the Mongodb Data

Back up the Mongodb data using the mongodump command in combination with the oc exec command: 

oc exec `oc get po --selector='deploymentconfig=mongodb-1' --template="{{(index .items 0).metadata.name}}"` bash -- -c '/opt/rh/rh-mongodb32/root/usr/bin/mongodump -u admin -p ${MONGODB_ADMIN_PASSWORD} --gzip --archive' > ./core7_mongodb.gz

5.6. Backing up the MySQL Data

Back up the MySQL data using the mysqldump command in combination with the oc exec command: 

oc exec `oc get po --selector='deploymentconfig=mysql' --template="{{(index .items 0).metadata.name}}"` bash -- -c '/opt/rh/mysql55/root/usr/bin/mysqldump -h mysql -u ${MYSQL_USER} -p${MYSQL_PASSWORD} --all-databases' > mysql-backup.sql

5.7. Backing up Filesystem Data

Back up the Nagios, metrics, git and scm files by copying the files on the associated persistent volumes:

  1. Determine the Volume names that need to be backed up, by entering the following command: 

    oc get pvc

    The output of this command displays the volume name, for example: 

    NAME                      STATUS    VOLUME    CAPACITY   ACCESSMODES   AGE
git-data                  Bound     pv5g32    5Gi        RWO           20h
metrics-backup-log-data   Bound     pv5g39    5Gi        RWO           20h
mongodb-claim-1           Bound     pv25g95   25Gi       RWO           20h
mysql                     Bound     pv5g173   5Gi        RWO           20h
nagios-claim-1            Bound     pv1g18    1Gi        RWO           2m
scm-data                  Bound     pv25g45   25Gi       RWO           20h

    This shows that only the pv5g32, pv5g39, pv1g18 and pv25g45 persistent volumes need to be backed up, corresponding to the persistent volumes named git-data, metrics-backup-log-data, nagios-claim-1 and scm-data.

  2. Determine the location of each of the filesystems that you want to back up using the oc describe pv command. For example to determine the location of the filesystems associated with the pv5g32 persistent volume in step 1, enter the following: 

    oc describe pv pv5g32

    The output of this command displays the filesystem location, for example: 

    Name:		pv5g32
Labels:		<none>
Status:		Bound
Claim:		core7/git-data
Reclaim Policy:	Delete
Access Modes:	RWO
Capacity:	5Gi
Message:
Source:
    Type:	HostPath (bare host directory volume)
    Path:	/home/vagrant/exp5g32

  3. Create an archive of the files. For example, to archive the files for the git-data volume in the steps above, enter the following command: 

    cd /home/vagrant/exp5g32 && tar -zcf ~/core_backup_2016_09_26_1300_001/core7_git-data.tar.gz .

5.8. Backup Frequency

Red Hat recommends backing up at least once per day, but you might decide to back up critical data more frequently.