Skip to main content

Back up MySQL to Filebase

Mirror the Postgres recipe using mysqldump.

One-off backup​

mysqldump --single-transaction --quick --routines --triggers \
  -u backup_user -p mydb \
  | gzip -9 \
  | aws --endpoint https://s3.filebase.io s3 cp - \
      s3://my-db-backups/mysql/mydb-$(date +%Y%m%d-%H%M%S).sql.gz

--single-transaction keeps the dump consistent without long table locks (InnoDB tables only).

Restore​

aws --endpoint https://s3.filebase.io s3 cp \
  s3://my-db-backups/mysql/mydb-20260501-120000.sql.gz - \
  | gunzip \
  | mysql -u admin -p mydb

Daily cron job​

/usr/local/bin/mysql-backup.sh
#!/bin/bash
set -euo pipefail

BUCKET="my-db-backups"
DB="mydb"
DATE=$(date +%Y%m%d-%H%M%S)

export AWS_ACCESS_KEY_ID="$FILEBASE_KEY"
export AWS_SECRET_ACCESS_KEY="$FILEBASE_SECRET"

mysqldump --single-transaction --quick --routines --triggers \
  --defaults-file=/etc/mysql/backup.cnf "$DB" \
  | gzip -9 \
  | aws --endpoint https://s3.filebase.io s3 cp - \
      "s3://${BUCKET}/mysql/${DB}-${DATE}.sql.gz"

Use --defaults-file=/etc/mysql/backup.cnf (chmod 600) so the password isn't on the command line.

Per-database backups​

For multi-tenant systems where you want per-database files:

for db in $(mysql -u backup -e "SHOW DATABASES;" -s --skip-column-names | grep -v -E "^(mysql|information_schema|performance_schema|sys)$"); do
  mysqldump --single-transaction "$db" \
    | gzip -9 \
    | aws --endpoint https://s3.filebase.io s3 cp - \
        "s3://my-db-backups/mysql/${db}-$(date +%Y%m%d).sql.gz"
done

Physical / hot backups​

For TB-scale MySQL, use xtrabackup or MySQL Enterprise Backup. Both can stream archives directly into S3-compatible storage.

What's next​