Backup and Restore Procedures

Overview

A robust backup strategy is essential for protecting your Sakai LMS data and ensuring business continuity. This guide covers backup procedures, restoration processes, and disaster recovery planning.

Backup Strategy

Components to Backup

A complete Sakai backup includes:

Database - All Sakai data (users, courses, content metadata)
File System - Content hosting files, attachments, resources
Configuration Files - sakai.properties, Tomcat configuration
Custom Code - Local customizations, plugins, skins

Backup Schedule

Component	Frequency	Retention
Database (Full)	Daily	30 days
Database (Incremental)	Hourly	7 days
File System	Daily	30 days
Configuration	Weekly	90 days
Point-in-time	Every 5 minutes	24 hours

Database Backup

MySQL/MariaDB Backup

Full Backup

Create a complete database backup:

#!/bin/bash
# Sakai MySQL backup script

BACKUP_DIR="/backup/sakai/mysql"
DATE=$(date +%Y%m%d_%H%M%S)
DB_NAME="sakai"
DB_USER="sakai"
DB_PASS="your_password"

# Create backup directory
mkdir -p $BACKUP_DIR

# Full backup with compression
mysqldump -u $DB_USER -p$DB_PASS \
  --single-transaction \
  --routines \
  --triggers \
  --events \
  --hex-blob \
  $DB_NAME | gzip > $BACKUP_DIR/sakai_full_$DATE.sql.gz

# Verify backup
if [ $? -eq 0 ]; then
  echo "Backup completed: sakai_full_$DATE.sql.gz"
else
  echo "Backup failed!" >&2
  exit 1
fi

Use --single-transaction for InnoDB tables to ensure consistent backups without locking tables. This is critical for production systems.

Binary Log Backup (Point-in-Time Recovery)

Enable binary logging in MySQL configuration:

# /etc/mysql/my.cnf
[mysqld]
log_bin = /var/log/mysql/mysql-bin.log
expire_logs_days = 7
max_binlog_size = 100M
server-id = 1

Backup binary logs:

#!/bin/bash
BINLOG_DIR="/var/log/mysql"
BACKUP_DIR="/backup/sakai/binlogs"
DATE=$(date +%Y%m%d)

mysqladmin flush-logs
rsync -av --remove-source-files $BINLOG_DIR/mysql-bin.* $BACKUP_DIR/$DATE/

Oracle Backup

RMAN Backup

-- Full database backup
RMAN> BACKUP DATABASE PLUS ARCHIVELOG;

-- Incremental backup
RMAN> BACKUP INCREMENTAL LEVEL 1 DATABASE;

-- Validate backup
RMAN> RESTORE DATABASE VALIDATE;

Export Backup

#!/bin/bash
expdp sakai/password@SAKAIDB \
  directory=DATA_PUMP_DIR \
  dumpfile=sakai_full_%U.dmp \
  logfile=sakai_export.log \
  full=yes \
  parallel=4 \
  compression=all

PostgreSQL Backup

#!/bin/bash
BACKUP_DIR="/backup/sakai/postgres"
DATE=$(date +%Y%m%d_%H%M%S)

pg_dump -U sakai -d sakai -F c -f $BACKUP_DIR/sakai_$DATE.dump

# With compression
pg_dump -U sakai -d sakai | gzip > $BACKUP_DIR/sakai_$DATE.sql.gz

File System Backup

Content Hosting Backup

Sakai stores files in the content hosting directory (typically /var/sakai/content):

#!/bin/bash
CONTENT_DIR="/var/sakai/content"
BACKUP_DIR="/backup/sakai/content"
DATE=$(date +%Y%m%d)

# Incremental backup using rsync
rsync -avz --delete \
  --link-dest=$BACKUP_DIR/latest \
  $CONTENT_DIR/ \
  $BACKUP_DIR/$DATE/

# Update latest symlink
ln -snf $BACKUP_DIR/$DATE $BACKUP_DIR/latest

Content hosting can be very large (hundreds of GB to TB). Use incremental backups and consider deduplication to save storage space.

Configuration Backup

Backup critical configuration files:

#!/bin/bash
BACKUP_DIR="/backup/sakai/config"
DATE=$(date +%Y%m%d)
TOMCAT_HOME="/opt/tomcat"
SAKAI_HOME="/var/sakai"

tar -czf $BACKUP_DIR/config_$DATE.tar.gz \
  $SAKAI_HOME/sakai.properties \
  $TOMCAT_HOME/conf/ \
  $TOMCAT_HOME/bin/setenv.sh \
  /etc/httpd/conf.d/sakai.conf

Automated Backup Script

Complete Backup Solution

#!/bin/bash
# Comprehensive Sakai backup script

set -e

# Configuration
BACKUP_ROOT="/backup/sakai"
DATE=$(date +%Y%m%d_%H%M%S)
RETENTION_DAYS=30
DB_NAME="sakai"
DB_USER="sakai"
DB_PASS="$(cat /root/.mysql_password)"
CONTENT_DIR="/var/sakai/content"
LOG_FILE="$BACKUP_ROOT/backup.log"

# Logging function
log() {
  echo "[$(date +'%Y-%m-%d %H:%M:%S')] $1" | tee -a $LOG_FILE
}

log "Starting Sakai backup"

# 1. Database backup
log "Backing up database..."
mkdir -p $BACKUP_ROOT/database
mysqldump -u $DB_USER -p$DB_PASS \
  --single-transaction \
  --routines --triggers --events \
  $DB_NAME | gzip > $BACKUP_ROOT/database/sakai_$DATE.sql.gz

if [ $? -eq 0 ]; then
  log "Database backup completed"
else
  log "ERROR: Database backup failed"
  exit 1
fi

# 2. Content hosting backup
log "Backing up content hosting..."
mkdir -p $BACKUP_ROOT/content/$DATE
rsync -avz --delete \
  --link-dest=$BACKUP_ROOT/content/latest \
  $CONTENT_DIR/ \
  $BACKUP_ROOT/content/$DATE/

ln -snf $BACKUP_ROOT/content/$DATE $BACKUP_ROOT/content/latest
log "Content backup completed"

# 3. Configuration backup
log "Backing up configuration..."
mkdir -p $BACKUP_ROOT/config
tar -czf $BACKUP_ROOT/config/config_$DATE.tar.gz \
  /var/sakai/sakai.properties \
  /opt/tomcat/conf/ \
  /opt/tomcat/bin/setenv.sh

log "Configuration backup completed"

# 4. Cleanup old backups
log "Cleaning up old backups..."
find $BACKUP_ROOT/database -name "*.sql.gz" -mtime +$RETENTION_DAYS -delete
find $BACKUP_ROOT/config -name "*.tar.gz" -mtime +$RETENTION_DAYS -delete
find $BACKUP_ROOT/content -maxdepth 1 -type d -mtime +$RETENTION_DAYS -exec rm -rf {} \;

log "Backup completed successfully"

# 5. Send backup status notification
# mail -s "Sakai Backup Status" [email protected] < $LOG_FILE

Schedule with Cron

# crontab -e
# Daily backup at 2 AM
0 2 * * * /usr/local/bin/sakai-backup.sh

# Hourly incremental database backup
0 * * * * /usr/local/bin/sakai-incremental-backup.sh

Disaster Recovery Testing

Test Restore Procedure

Create a test environment separate from production
Restore database from latest backup
Restore content hosting files
Verify application starts successfully
Test critical functionality (login, content access, submissions)
Document any issues and recovery time

Test your restore procedures at least quarterly. Untested backups are not backups.

Restore Procedures

Database Restore

MySQL/MariaDB Full Restore

#!/bin/bash
BACKUP_FILE="/backup/sakai/database/sakai_20260304_020000.sql.gz"
DB_NAME="sakai"
DB_USER="sakai"
DB_PASS="your_password"

# Stop Tomcat
systemctl stop tomcat

# Drop and recreate database
mysql -u root -p << EOF
DROP DATABASE IF EXISTS $DB_NAME;
CREATE DATABASE $DB_NAME CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci;
GRANT ALL ON $DB_NAME.* TO '$DB_USER'@'localhost';
FLUSH PRIVILEGES;
EOF

# Restore from backup
zcat $BACKUP_FILE | mysql -u $DB_USER -p$DB_PASS $DB_NAME

# Start Tomcat
systemctl start tomcat

Point-in-Time Recovery (MySQL)

#!/bin/bash
# Restore to specific timestamp

FULL_BACKUP="/backup/sakai/database/sakai_20260304_020000.sql.gz"
BINLOG_DIR="/backup/sakai/binlogs/20260304"
TARGET_TIME="2026-03-04 14:30:00"

# 1. Restore full backup
zcat $FULL_BACKUP | mysql -u sakai -p sakai

# 2. Apply binary logs up to target time
mysqlbinlog --stop-datetime="$TARGET_TIME" \
  $BINLOG_DIR/mysql-bin.* | mysql -u sakai -p sakai

Oracle Restore

-- RMAN restore
RMAN> SHUTDOWN IMMEDIATE;
RMAN> STARTUP MOUNT;
RMAN> RESTORE DATABASE;
RMAN> RECOVER DATABASE;
RMAN> ALTER DATABASE OPEN;

Content Hosting Restore

#!/bin/bash
BACKUP_DIR="/backup/sakai/content/20260304"
CONTENT_DIR="/var/sakai/content"

# Stop Tomcat
systemctl stop tomcat

# Backup current content (safety)
mv $CONTENT_DIR $CONTENT_DIR.pre-restore

# Restore from backup
rsync -avz $BACKUP_DIR/ $CONTENT_DIR/

# Fix permissions
chown -R tomcat:tomcat $CONTENT_DIR

# Start Tomcat
systemctl start tomcat

Configuration Restore

#!/bin/bash
CONFIG_BACKUP="/backup/sakai/config/config_20260304.tar.gz"

# Extract configuration
tar -xzf $CONFIG_BACKUP -C /

# Restart services
systemctl restart tomcat

Special Considerations

Site Statistics Database

If using a separate statistics database:

# Backup site statistics
mysqldump -u stats_user -p \
  --single-transaction \
  sakai_stats | gzip > /backup/sakai/stats_$DATE.sql.gz

You may choose not to restore statistics in disaster recovery scenarios as they are not critical for system operation.

Search Index

If using Elasticsearch for search:

# Elasticsearch snapshot
curl -X PUT "localhost:9200/_snapshot/sakai_backup/snapshot_$DATE?wait_for_completion=true"

# Restore snapshot
curl -X POST "localhost:9200/_snapshot/sakai_backup/snapshot_$DATE/_restore"

Sakai can rebuild the search index from database content, so search index backups are optional.

Content-Review Integration

For Turnitin or other content-review integrations:

# Disable indexing in QA/restore environments
contentreview.enableContentReview=false

When restoring production data to QA/staging environments, disable content-review indexing to prevent duplicate submissions.

Offsite Backup

Cloud Storage Sync

#!/bin/bash
# Sync backups to AWS S3
BACKUP_DIR="/backup/sakai"
S3_BUCKET="s3://my-sakai-backups"

aws s3 sync $BACKUP_DIR $S3_BUCKET \
  --storage-class GLACIER \
  --exclude "*.tmp" \
  --delete

Encryption for Offsite Storage

#!/bin/bash
# Encrypt backup before upload
BACKUP_FILE="sakai_full_$DATE.sql.gz"
GPG_RECIPIENT="[email protected]"

gpg --encrypt --recipient $GPG_RECIPIENT $BACKUP_FILE
aws s3 cp ${BACKUP_FILE}.gpg s3://my-sakai-backups/

Recovery Time Objectives

Scenario	RTO Target	RPO Target
Database corruption	2 hours	5 minutes
Complete server failure	4 hours	1 hour
Data center disaster	24 hours	24 hours
Accidental deletion	1 hour	1 hour

RTO = Recovery Time Objective (time to restore service)
RPO = Recovery Point Objective (acceptable data loss)

Best Practices

3-2-1 Rule

Maintain 3 copies of data, on 2 different media types, with 1 copy offsite.

Test Regularly

Test restore procedures quarterly. Update documentation based on test results.

Monitor Backups

Set up alerts for backup failures. Verify backup integrity automatically.

Document Everything

Maintain runbooks for recovery procedures. Include contact information for vendors.

Troubleshooting

Backup Failures

Problem: Database backup fails with “Lock wait timeout” Solution: Use --single-transaction for InnoDB tables or schedule backups during low-usage periods. Problem: Content backup takes too long Solution: Use incremental backups with rsync or implement block-level snapshots.

Restore Issues

Problem: Restored database has different character encoding Solution: Ensure database is created with correct charset:

CREATE DATABASE sakai CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci;

Problem: Content files have wrong permissions after restore Solution: Fix ownership recursively:

chown -R tomcat:tomcat /var/sakai/content
chmod -R 755 /var/sakai/content

Installation

Configuration

Operations

​Overview

​Backup Strategy

​Components to Backup

​Backup Schedule

​Database Backup

​MySQL/MariaDB Backup

​Full Backup

​Binary Log Backup (Point-in-Time Recovery)

​Oracle Backup

​RMAN Backup

​Export Backup

​PostgreSQL Backup

​File System Backup

​Content Hosting Backup

​Configuration Backup

​Automated Backup Script

​Complete Backup Solution

​Schedule with Cron

​Disaster Recovery Testing

​Test Restore Procedure

​Restore Procedures

​Database Restore

​MySQL/MariaDB Full Restore

​Point-in-Time Recovery (MySQL)

​Oracle Restore

​Content Hosting Restore

​Configuration Restore

​Special Considerations

​Site Statistics Database

​Search Index

​Content-Review Integration

​Offsite Backup

​Cloud Storage Sync

​Encryption for Offsite Storage

​Recovery Time Objectives

​Best Practices

3-2-1 Rule

Test Regularly

Monitor Backups

Document Everything

​Troubleshooting

​Backup Failures

​Restore Issues

​Related Resources

Build docs developers (and LLMs) love

Overview

Backup Strategy

Components to Backup

Backup Schedule

Database Backup

MySQL/MariaDB Backup

Full Backup

Binary Log Backup (Point-in-Time Recovery)

Oracle Backup

RMAN Backup

Export Backup

PostgreSQL Backup

File System Backup

Content Hosting Backup

Configuration Backup

Automated Backup Script

Complete Backup Solution

Schedule with Cron

Disaster Recovery Testing

Test Restore Procedure

Restore Procedures

Database Restore

MySQL/MariaDB Full Restore

Point-in-Time Recovery (MySQL)

Oracle Restore

Content Hosting Restore

Configuration Restore

Special Considerations

Site Statistics Database

Search Index

Content-Review Integration

Offsite Backup

Cloud Storage Sync

Encryption for Offsite Storage

Recovery Time Objectives

Best Practices

Troubleshooting

Backup Failures

Restore Issues

Related Resources