Tuesday, September 23, 2008

Re: [HACKERS] Proposal of SE-PostgreSQL patches (for CommitFest:Sep)

KaiGai Kohei wrote:
> > [1] Make a consensus that different security mechanisms have differences
> > in its decision making, its gulanuality and its scope
> >
> > I think it is the most straightforward answer.
> > As operating system doing, DAC and MAC based access controls should be
> > independently applied on accesses from users, and this model is widely
> > accepted.
> > These facilities can also have different results, gulanualities and scopes.
> >
> >
> > [2] Make a new implementation of OS-independent fine grained access control
> >
> > If it is really really necessary, I may try to implement a new separated
> > fine-grained access control mechanism due to the CommitFest:Nov.
> > However, we don't have enough days to develop one more new feature from
> > the scratch by the deadline.
> I reconsidered the above two options have no differences fundamentally.
> In other word, making a new enhanced security implementation based on
> requirements also means making a consensus various security mechanism
> can have its individual rules including guranuality of access controls.
> So, I'll decide to try to implement "fine-grained-only" security
> mechanism also, because someone have such a requirememt.
> However, its schedule is extremely severe, if is has to be submitted
> due to the deadline of CommitFest:Nov.
> It is my hope to concentrate development of SE-PostgreSQL in v8.4
> development cycle, and I think the above "fine-grained-only" one
> should be pushed to v8.5 cycle.

Well, those might be your priorities, but I don't think they are the
community's priorities.

I think the community's priorities are to add security at the SQL
level, and then we can see clearly what SE-PostgreSQL requires. This
has been discussed before so it should not come as a surprise.

What you can do is to do things in this order:

1) Add SE-PostgreSQL capabilities that layer over existing Postgres
SQL-level permissions
2) Implement "fine-grained" permissions at the SQL level
3) Add SE-PostgreSQL capabilities for "fine-grained" permissions

Perhaps you can only get #1 done for 8.4; I don't know, but I knew
months ago that #2 had to be done before #3, and I think that was

Bruce Momjian <bruce@momjian.us> http://momjian.us
EnterpriseDB http://enterprisedb.com

+ If your life is a hard drive, Christ can be your backup. +

Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:

No comments: