ELMA365 backup should be done for:
- PostgreSQL;
- MongoDB;
- S3 Object Storage.
For a consistent state of the backup copy, disable the operation load. You can do it with the following commands:
kubectl scale deploy --replicas=0 --all -n default
kubectl -n default patch daemonset billing -p '{"spec": {"template": {"spec": {"nodeSelector": {"non-existing": "true"}}}}}'
Then wait for all the services to stop and run the backup.
After you have created a backup copy, run all the services using the following commands:
kubectl scale deploy --replicas=1 --all -n default
kubectl -n default patch daemonset billing --type json -p='[{"op": "remove", "path": "/spec/template/spec/nodeSelector/non-existing"}]'
Before recovering datab, you also have to stop all the application services completely and then recreate the PostgreSQL and MongoDB databses. To learn more see Install PostgreSQL and Install MongoDB.
Then recover the databases and restart the application services.
PostgreSQL Dump
To dump the PostgreSQL database, use the standard utility function:
pg_dump $postgresURL -C -c --if-exists -F directory -f $postgresDIR
MongoDB Dump
To dump the MongoDB database, use the standard utility function:
mongodump --uri $mongoURL --gzip --quiet --out $mongoDIR
S3 Dump
To copy the files, you can use any utility that connects to the S3 database, for example the mc:
mc mirror $source $destination
Restore PostgreSQL
To restore the PostgreSQL database, enter the following command:
pg_restore -d $postgresURL -j 1 -F directory $postgresDIR
Restore MongoDB
To restore the MongoDB database, enter the following command:
mongorestore --uri $mongoURL --dir $mongoDIR --drop --gzip --preserveUUID --excludeCollection=head.settings.view
Copy S3 files
To copy the files, use the command:
mc cp $backup $destination
Found a typo? Highlight the text, press ctrl + enter and notify us