description: Accessing the OVH Public cloud archive with the automated backup tool backupninja using its duplicity backend
date: 2022-03-03
---
## motivation
OVH provides a cheap way for longterm storing backup data, the Public Cloud Archive (PCA). As this is for write-once data (cold data store), it needs to be used together with a little bit more expensive hot data store for meta data when used for sequential backups. The backup solution [duplicity](https://duplicity.gitlab.io/duplicity-web/) is capable of using such multi backends. [backupninja](https://0xacab.org/liberate/backupninja) provides a nice solution for a standardized configuration of duplicity and running it automatically.
## Prerequisites
### On system
- Enable email sending for backup status mails by [installing a mta](/msmtp-on-debian/).
-`/etc/backupninja.conf`: Here you can set the time of day when the backup should be run, and to whom to send status emails. This file is pre-installed and quite self-explaining
- There is a full example for duplicity configuration in `/usr/share/doc/backupninja/examples/example.dup` find a slightly pre configured version [here](/texts/20_ovh_pca.dup)
- adapt the `url` keys to something meaningful ex. `backedupserver1_cold` and `backedupserver2_hot`. This will be the names of the Cloud Archives resp. Object Stores
- change `PCA_TENANTID` and `SWIFT_TENANTID` to noted `OS_TENANT_ID`
- change `PCA_USERNAME` and `SWIFT_USERNAME` to noted `OS_USERNAME`