Backing Up and Restoring Continuum

Last updated
Save as PDF

Overview

This article describes the process of backing up the data, storing it off-server, and restoring it on an as needed basis.

Continuum consists of two areas that should be backed up on a regular basis: MongoDB database and a configuration directory on the file system.

Backup

The backup process requires access to the Continuum server. The following commands can be used or customized for use to create a backup tar file.

# save the date in a variable
export THISDATE=$(date +%Y-%m-%d_%H-%M)

# create a temporary directory
mkdir /tmp/continuum_backup_${THISDATE}
cd /tmp

# create a dump of the continuum mongo db
mongodump --db continuum --out continuum_backup_${THISDATE}/continuum_mongo

cp /etc/continuum/continuum.yaml continuum_backup_${THISDATE}/continuum.yaml
cp /etc/continuum/service.conf continuum_backup_${THISDATE}/service.conf

# tar and gzip all the directories and files
tar -czf continuum_backup_${THISDATE}.tar.gz continuum_backup_${THISDATE}

# remove the directory
rm -rf continuum_backup_${THISDATE}

At this point you will want to get the file off the Continuum server using sftp or scp or some other means such as uploading to object storage like Amazon S3. Then remove the file from the Continuum disk using the following command:

rm continuum_backup_${THISDATE}.tar.gz

This backup process can be used in a bash script and placed into a cron job or can be run from a scheduled Continuum Task. It is important to understand what an acceptable risk may be with regards to how often you run the backup: once a day, once an hour, etc.

This backup process is in it's simplest form, there may be better options as far as performing snapshots or specific third party database backup technologies that your enterprise be already have at your disposal or may want to explore.

Offsite Storage

VersionOne does not provide any specific guidance on which method of offsite storage is used, whether that is a cloud-based object storage (e.g. Amazon S3), an ftp server or otherwise. We do recommend that you keep several interactions of backups and test your restore process.

Restoring a Backup

Make sure you have the same version of MongoDB on the server from which you took the backup and the server where you restore the backup. If not, upgrade the MongoDB before you restore the backup.

After choosing the backup version to restore, get the tar.gz file downloaded to the Continuum server. If you are recovering from a total and complete failure, you will need to first perform a new install on a new machine. Once the install has completed and the backup file is on the server, perform the following commands.

The first command sets the THISDATE environment variable to the date / time string part of the backup file name as a matter of convenience in the following commands.

export THISDATE=2016-01-29_04-13

# stop the Continuum services
ctm-stop-services

# untar the file
tar -xzf continuum_backup_${THISDATE}.tar.gz

# restore the mongodb
mongorestore --drop --db continuum continuum_backup_${THISDATE}/continuum_mongo/continuum

# restore the config files
cp continuum_backup_${THISDATE}/continuum.yaml /etc/continuum/continuum.yaml
cp continuum_backup_${THISDATE}/service.conf /etc/continuum/service.conf

# start the services
ctm-start-services