This repo provides a way to perform full backups on databases running on Cloud, storing them in a Cloud Object Storage instance. Currently PostgreSQL full backups are available.
PostgreSQL backups are available as full backups performed by pg_dump or pg_dumpall utilities. The file name format is:
<PGHOST>_<DatabaseName>_<YYYYMMDDHHmmss>.dump.gz
All files are compressed by default.
If DATABASES variable value is 'all', pg_dumpall is used. Otherwise, pg_dump is used for each table defined in DATABASES variable.
# npm install only once
npm install
node index.jsBACKUP_RUNNER: Backup runner module to execute. Currently only 'postgresql' is supported.
DATABASES: comma-delimited list of databases to perform a backup.
COS_ENDPOINT: Endpoint for your Cloud Object Storage instance (i.e. s3.sao01.cloud-object-storage.appdomain.cloud)
COS_APIKEY: Cloud Object Storage API Key to authenticate with. This API Key must have at least Writer permissions for the bucket.
COS_CRN: Cloud Object Storage Instance ID (i.e crn:v1:bluemix:public:cloud-object-storage:global:a/XXXXXXXXXX:XXXXXXX).
COS_\BUCKET_NAME: Bucket name to store your backups.
PAGE_SIZE: Size of sent messages to COS (Default is 5MB).
QUEUE_SIZE: Size of queue of concurrent connections to COS (default is 5).
PGHOST: PostgreSQL host name.
PGPORT: PostgreSQL port to connect to.
PGUSER: PostgreSQL user.
PGPASSWORD: PostgreSQL user's password.
PG_ADDITIONAL_OPTIONS: comma-delimited list of parameters to be passed to PostgreSQL backup command (i.e: -F,custom).