Deconstructing The PetaBox


     The day before shipping, everyone chipped in and helped take the PetaBox rack apart and do some last-minute QA.  First the nodes were taken out of the rack, then they were opened up and their secondary IDE bus cables turned around (/dev/hdc and /dev/hdd needed to be swapped in most machines).  Then the systems were put back together and turned on, their BIOS settings checked for correct boot order (USB-HDD first, then LAN boot, then HDD0), then rebooted to a "login:" prompt just to make sure that everything would work okay.  Finally they went into boxes for shipping.
 

 
Please click on a thumbnail to view larger image.

 

Bruce Baumgart, hard at work.  Bruce will be taking over PetaBox software development and system administration from me so that I can go back to working in the Data Repository group at the Internet Archive.

Bruce again.  Hi, Bruce!

A PetaBox in pieces.  What a mess!
    

John and Joerg, hard at work.

John, taping up a box.

Smile, Joerg!  Joerg is the senior system administrator at the Internet Archive.  He helped a lot with the baseline PetaBox configuration.  The Debian distribution we used was Joerg's preferred production platform.
    

Who is this mysterious figure?

John Barry is the manager of Operations at the Archive.  I owe him my thanks for putting me into the "software/sysadmin" slot for the PetaBox project.  He also figured out the shipping and other necessary details for the PetaBox.

It's Bruce again!  If his hands look a little blurry, it's because he works with such inhuman speed.
    

Luca, Joerg, Brad, and Bruce, making an assembly line.

Igor, myself, and Bruce, getting the last dozen or so boxes ready to go.


.
.
.
.
.