[GUIDE] How Backup Server Stores Data
NovaBACKUP xSP / Remote Workforce
This article refers to how Data is stored on the Backup/Storage server when using the xSP or Remote Workforce product line.
When a backup server receives a backup job from the client, it creates a folder with that user name in the specified datastore. If the User backing up is named "FileServer01", a folder is created in the Primary Storage location for that user called c:\UserData\FileServer01. After this, a sub folder is created using a Unique Identifier which is a reference to the machine name. For example. On this file server lets assume there are 3 different physical hard drives and one plugin such as the "System State" module. The tree structure where the data is stored would look like this. (all actual data below it is encrypted)
If i were to have 2 different computers using the same account, there would be another unique identifier with the corresponding data on it. Assuming i also had a SQL server machine using this account also, the structure might look like this.
As you can see there are 2 different computers using this account (2001 and 2002). Going down further in the tree you would be able to see the actual data backed up (encrypted and compressed by unique numbers that are referenced in the database)
What happens when backup client renames its folders on the client?
Since all data that is stored on the backup server is kept in its original location format, you need to pay close attention to if the client makes any major changes to its file structure. For example. If a file server had 10TB of data located in D:\data1 and the end user decides to rename that folder D:\data2, as far as the Backup Server knows, that is an ENTIRELY new folder and directory. It has no way of determining that the files are the same and just the root folder changed. That means the backup server will received the request to backup data and tell the client is has no data under d:\data2 and will tell the client to re-upload its 10 terabytes all over again to the new directory.
The same situation occurs for any of the plugin modules as well too.
What happens when you rename a machine?
The same situation will occur if the machine name is changed. In the above example, 2001 and 2002 represent two different machines tied to the same account. If the end client were to rename their machine, as far as the backup server knows it is a different machine connecting on the same account and generates a new ID to represent the new machine. It will not have any record of previous backup information from that machine and request a backup of the entire data set again.