System level backup directory structure

System level data is backed up to the remote system and organized in a unique directory structure.

The following syntax shows the general directory structure for system level backup data:

For system level data associated with internal IBM Workload Deployer functions:
```
/<backup_location_path>/<rack_ID>/5.0.0.0/backupStorage/<iwd_artifact>/<timestamp>
```
In the directory path, /5.0.0.0/ represents the internal version of IBM Workload Deployer. Note that there is also a separate /backupStorage/ subdirectory in the path.
The /<iwd_artifact>/ subdirectory represents one of the following folder names:
- docroot
- facets
- globalcontent
- key
- scripts
- security
- templates
- vault
- virtual
Under each of these folders are timestamp folders containing the system level data from each backup operation.

In addition to these /<iwd_artifact>/ folders a backup.json file is also present.
For Platform System Manager data:
```
/<backup_location_path>/<rack_ID>/backupStorage/<timestamp>
```
The <timestamp> folders contain the full baseline system level backups of Platform System Manager data for each backup operation, written into the following subdirectories:
- databases
- key
- TPM
- V7000
In addition to these folders, the files.tar.gz.encrypted file is present.
For pattern type data:
```
/<backup_location_path>/<rack_ID>/backupStorage/ptype/<ptype_name>/<timestamp>
```
The /<ptype_name>/ subdirectory represents folders for each pattern type, including these examples and others:
- Application_Pattern_Type_for_Java_1.0.1.1
- Foundation_Pattern_Type_2.0.1.2
- Foundation_Pattern_Type_2.1.0.0
- Guardium_Pattern_Type_1.0.0.0.orig
- IBM_Endpoint_Manager_Pattern_Type_1.0.0.1
- RedHat_OS_Update_Service_Pattern_1.0.0.1
- System_Monitoring_Pattern_1.0.1.2
- Virtual_System_Pattern_Type_1.0.0.0
Under each of these folders are <timestamp> folders containing the system level data from each backup operation.

In addition to these /<ptype_name>/ folders a ptype.json file is also present.

For volume group data:

/<backup_location_path>/<rack_ID>/backupStorage/volumegroup/<timestamp>

Converting Java timestamps to readable values

The following example shows how you can convert a Java timestamp to a readable value:

import datetime

def convert_java_millis(java_time_millis):
  ds = datetime.datetime.fromtimestamp(int(str(java_time_millis)[:10]))
       if java_time_millis else None
  ds = ds.replace(hour=ds.hour,minute=ds.minute,second=ds.second,
       microsecond=int(str(java_time_millis)[:10]) * 1000)
  return ds

dt=convert_java_millis('1234567890123')

str(dt)
'2009-02-13 18:31:30.123000'