Software Engineering discussion

I so looked forward to this chapter, and I was so let down. First, this was more architecture than code. Then, of all the potentially interesting topics surrounding software for a space mission... they pick ground-based file sharing?? And, they chose EJBs for their implementation, and EJBs give me hives!

This chapter didn't bother me too much. I can certainly see your point of view and being let down.

My point of view of NASA is from museums and 1960s technology on display. It was nice to see NASA archetecture lives in the modern age.

I noticed one section that mentioned "token" sharing. I wasn't sure if that was an IBM network thing, or if they were using the token for something different.

They mention two types of tokens. The user token is provided to a user by the authentication subsystem after they are authenticated. It typically consists of some contents known to the file system that are encrypted by the authentication subsystem with a key that the file system knows. So when a user presents the token to the file system, the file system decrypts it and checks that the contents are as expected. This is a quick way to verify that a file system request is coming from an authenticated user without having to run through the more expensive authentication code, and potentially involving the user each time. In short, it is a proxy for an authenticated user. Kerberos calls this idea a ticket/authenticator.

The second kind of token is a file token. The purpose of this is not clear to me, but I think it stores file system state pulled out from the file access bean so that the bean can be implemented in a stateless fashion.