db-derby-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Rick Hillegas (JIRA)" <j...@apache.org>
Subject [jira] Commented: (DERBY-2109) System privileges
Date Mon, 21 Jan 2008 18:22:35 GMT

    [ https://issues.apache.org/jira/browse/DERBY-2109?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12561086#action_12561086

Rick Hillegas commented on DERBY-2109:

Thanks for the continued discussion, Dan. I agree that we seem to be talking past one another.

>> 2) AuthorizationID -- this is the SQL concept of identity. This is case-insensitive
unless double-quoted.
>We seem to be using the same terms for slightly different concepts.
>I believe I'm using authorization identifier in the way that the SQL standard defines
it. This would be a case-sensitive value that defines a unique identity in the sql system
(database). Thus an authorization identifier is never double >quoted, contradicting your
statement 2).

I think the relevant section of the SQL spec is part 2 section 5.4 (names and identifiers).
In that section <authorization identifier> can resolve to be a <user identifier>
which in turn can resolve to an <identifier> which in turn can resolve to a <regular
identifier> or a <delimited identifier>. The <regular identifier> is uppercased
to a normalized value regardless of its original casing and the <delimited identifer>
is quoted and is not uppercased. This is what I mean by "This is case-insensitive unless double-quoted".

I think the disconnect here is that the authorization identifier can be referenced in two
ways. When it is used in a SQL statement and resolves to a <regular identifier>, it
is normalized to uppercase and so, as I would put it, it is case-insensitive. When it is referenced
in the form stored in the system catalogs, it is case-sensitive.

>A UserName is a representation of an authorization identifier using the SQL rules for
regular and delimited identifiers and thus has the case folding rules etc.

The term UserName is a word I invented for this discussion. But it pretty much corresponds
to userName as used in the javadoc for UserAuthenticator.

>In SQL this behaviour, mapping UserName to authorization identifier is defined by the

This is where I am not tracking you. The term UserName does not occur in the SQL standard
as far as I know. It is a term which I thought I invented in order to bring some precision
to this discussion. But I don't seem to have succeeded. Can you refer me to the section of
the SQL standard which defines this term (or a related term)? That will help me propose some
new language which we can agree on.

>In Java code and Java property files Derby chose to use SQL identifier rules (ie. UserName)
as the representation of an authorization identifier. Looking back this maybe was a poor choice,
a direct representation of the authorization i>dentifier might have been better, (ie. jdbc:derby:cs;user=fred
and jdbc:derby:cs;user=FRED would connect as different authorization ids, today they map to
the same authorization id). So if we want to be consistent with other Java us>es, the policy
file should probably use UserName, however using authorization identifier might be clearer.
>> 1) UserName -- this is part of the credentials passed to the authentication service.
This could be case-sensitive or case-insensitive depending on the rules of the authentication
>I think the last sentence is incorrect. The rules of UserName are set by Derby, not an
arbitrary implementation of the authentication service.
>As above, the rules for how UserName map to a unique identity follows the rules of SQL
identifiers, and thus UserNames of FRED and fred always map to the same unique identity FRED.
However, I'm not sure you agree with this, sinc>e you are promoting authorization identifier
as a SQL only concept. That may be a valid approach, but I think you need to clearly state
the rules for that, and then we can discuss if having two different models (one for database
and o>ne for system) adds any benefit or adds complication. The rules would need to state
how UserName maps to a unique identity in a system context. 

Ha! Once we can agree on some terms, I can take another crack at stating the rules.


> System privileges
> -----------------
>                 Key: DERBY-2109
>                 URL: https://issues.apache.org/jira/browse/DERBY-2109
>             Project: Derby
>          Issue Type: New Feature
>          Components: Security
>    Affects Versions:
>            Reporter: Rick Hillegas
>            Assignee: Martin Zaun
>         Attachments: DERBY-2109-02.diff, DERBY-2109-02.stat, derby-2109-03-javadoc-see-tags.diff,
DERBY-2109-04.diff, DERBY-2109-04.stat, DERBY-2109-05and06.diff, DERBY-2109-05and06.stat,
DERBY-2109-07.diff, DERBY-2109-07.stat, DERBY-2109-08.diff, DERBY-2109-08.stat, DERBY-2109-08_addendum.diff,
DERBY-2109-08_addendum.stat, SystemPrivilegesBehaviour.html, systemPrivs.html, systemPrivs.html,
systemPrivs.html, systemPrivs.html
> Add mechanisms for controlling system-level privileges in Derby. See the related email
discussion at http://article.gmane.org/gmane.comp.apache.db.derby.devel/33151.
> The 10.2 GRANT/REVOKE work was a big step forward in making Derby more  secure in a client/server
configuration. I'd like to plug more client/server security holes in 10.3. In particular,
I'd like to focus on  authorization issues which the ANSI spec doesn't address.
> Here are the important issues which came out of the email discussion.
> Missing privileges that are above the level of a single database:
> - Create Database
> - Shutdown all databases
> - Shutdown System
> Missing privileges specific to a particular database:
> - Shutdown that Database
> - Encrypt that database
> - Upgrade database
> - Create (in that Database) Java Plugins (currently  Functions/Procedures, but someday
Aggregates and VTIs)
> Note that 10.2 gave us GRANT/REVOKE control over the following  database-specific issues,
via granting execute privilege to system  procedures:
> Jar Handling
> Backup Routines
> Admin Routines
> Import/Export
> Property Handling
> Check Table
> In addition, since 10.0, the privilege of connecting to a database has been controlled
by two properties (derby.database.fullAccessUsers and derby.database.defaultConnectionMode)
as described in the security section of the Developer's Guide (see http://db.apache.org/derby/docs/10.2/devguide/cdevcsecure865818.html).

This message is automatically generated by JIRA.
You can reply to this email to add a comment to the issue online.

View raw message