Bradley Manning's Surprising Statement to the Court Details Why He Made His Historic Wikileak
Continued from previous page
The organic D6-A machines we brought with us into the field on our deployment were Dell [missed word] laptops and the TPE D6-A machines were Alienware brand laptops. The [M90?] D6-A laptops were the preferred machine to use as they were slightly faster and had fewer problems with dust and temperature than the theater provided Alienware laptops. I used several D6-A machines during the deployment due to various technical problems with the laptops.
With these issues several analysts lost information, but I never lost information due to the multiple backups I created. I attempted to backup as much relevant information as possible. I would save the information so that I or another analyst could quickly access it whenever a machine crashed, SIPRnet connectivity was down or I forgot where the data was stored.
When backing up information I would do one or all of the following things based on my training:
[(1)] Physical back up. I tried to keep physical back up copies of information on paper so that the information could be grabbed quickly. Also, it was easier to brief from hard copies of research and HUMINT reports.
(2) Local drive back up. I tried to sort out information I deemed relevant and keep complete copies of the information on each of the computers I used in the Temporary Sensitive Compartmented Information Facility, or T-SCIF, including my primary and secondary D6-A machines. This was stored under my user profile on the desktop.
[(3)] Shared drive backup. Each analyst had access to a ‘T’ drive — what we called ‘T’ drive shared across the SIPRnet. It allowed others to access information that was stored on it. S6 operated the ‘T’ drive.
[(4)] Compact disk rewritable or CD-RW back up. For larger datasets I saved the information onto a re-writable disk, labeled the disks and stored them in the conference room of the T-SCIF. This redundancy permitted us to not worry about information loss. If the system crashed, I could easily pull the information from a secondary computer, the ‘T’ drive or one of the CD-RWs.
If another analysts wanted to access my data, but I was unavailable, she could find my published products directory on the ‘T’ drive or on the CD-RWs. I sorted all of my products or research by date, time and group, and I updated the information on each of the storage methods to ensure that the latest information was available to them.
During the deployment I had several of the D6-A machines crash on me. Whenever one of the computer crashed, I usually lost information but the redundancy method ensured my ability to quickly restore old backup data and add my current information to the machine when it was repaired or replaced.
I stored the backup CD-RW with larger datasets in the conference room of the T-SCIF or next to my workstation. I marked the CD-RWs based on the classification level and its content. Unclassified CD-RWs were only labeled with the content type and not marked with classification markings. Early on in the deployment, I only saved and stored the SigActs that were within or near operational environment.
Later I thought it would be easier to just to save all of the SigActs onto a CD-RW. The process would not take very long to complete and so I downloaded the SigActs from CIDNE-I onto a CD-RW. After finishing with CIDNE-I, I did the same with CIDNE-A. By retrieving the CIDNE-I and CIDNE-A SigActs, I was able to retrieve the information whenever I needed it, and not rely upon the unreliable and slow SIPRnet connectivity needed to pull. Instead, I could just find the CD-RW and open up a pre-loaded spreadsheet.