[
https://issues.apache.org/jira/browse/DERBY-3200?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Kim Haase updated DERBY-3200:
-----------------------------
Attachment: AuthExampleEmbedded.java
DERBY-3200-3.zip
DERBY-3200-3.diff
Thanks again for all your help, Dag. I've revised the other programs along the
same lines you suggested in your version of the embedded example. The biggest
change, I think, is in the simpler client examples: previously they only closed
the connection between the first and second programs, but now they shut down
the database as the other programs do.
Hope the comments in the programs are now helpful.
I'm attaching DERBY-3200-3.diff and DERBY-3200-3.zip with the changes. I'll
attach the new versions of the source files too. They've all changed
significantly.
> Developer's Guide: Add examples showing use of SQL authorization with user
> authentication
> -----------------------------------------------------------------------------------------
>
> Key: DERBY-3200
> URL: https://issues.apache.org/jira/browse/DERBY-3200
> Project: Derby
> Issue Type: Improvement
> Components: Documentation
> Reporter: Kim Haase
> Assignee: Kim Haase
> Priority: Minor
> Attachments: auth2.log, AuthExampleClientSQLAuth1.java,
> AuthExampleClientSQLAuth1.java, AuthExampleClientSQLAuth1.java,
> AuthExampleClientSQLAuth1.java, AuthExampleClientSQLAuth1.java,
> AuthExampleClientSQLAuth1.java, AuthExampleClientSQLAuth2.java,
> AuthExampleClientSQLAuth2.java, AuthExampleClientSQLAuth2.java,
> AuthExampleClientSQLAuth2.java, AuthExampleClientSQLAuth2.java,
> AuthExampleClientSQLAuth2.java, AuthExampleEmbedded-dhw.java,
> AuthExampleEmbedded.java, AuthExampleEmbedded_dhw.java,
> AuthExampleEmbeddedSQLAuth.java, AuthExampleEmbeddedSQLAuth.java,
> AuthExampleEmbeddedSQLAuth.java, AuthExampleEmbeddedSQLAuth.java,
> DERBY-3200-2.diff, DERBY-3200-2.zip, DERBY-3200-3.diff, DERBY-3200-3.zip,
> DERBY-3200.diff, DERBY-3200.stat, DERBY-3200.zip,
> rdevcsecuresqlauthembeddedex.dita, sqlauthclient.txt,
> sqlauthclientshutdown.txt, sqlauthembedded.txt, sqlauthembedded.txt
>
>
> This is the followup to DERBY-1823 that Francois Orsini suggested.
> I've been experimenting and reading the Developer's Guide section on SQL
> authorization (User authorizations, cdevcsecure36595).
> It appears that the only use of SQL authorization mode is to restrict user
> access, not to expand it.
> For example, if you set the default connection mode to noAccess, a user with
> fullAccess can't grant any privileges to a user with noAccess. And presumably
> if the default connection mode is readOnlyAccess, a user with fullAccess
> can't grant any privileges beyond SELECT, which the user has anyway.
> Only if the default connection mode is fullAccess is SQL authorization mode
> meaningful. That means that a fullAccess user can use GRANT to restrict
> another user's privileges on a particular database that the user owns.
> I'm running into a problem at the end, though. At the beginning of the
> program, as nobody in particular, I was able to create several users, some of
> them with full access. But at the end of the program, it seems that even a
> user with full access isn't allowed to turn off those database properties:
> Message: User 'MARY' does not have execute permission on PROCEDURE
> 'SYSCS_UTIL'.'SYSCS_SET_DATABASE_PROPERTY'.
> This seems a bit extreme. I know that with SQL authorization on, "the ability
> to read from or write to database objects is further restricted to the owner
> of the database objects." But the ability to execute built-in system
> procedures? Can I log in as SYSCS_UTIL? How?
> I realize that having access to SYSCS_SET_DATABASE_PROPERTY would allow me to
> in effect delete myself -- but that's essentially what I do at the end of the
> program that sets derby.connection.requireAuthentication but not
> derby.database.sqlAuthorization.
> The documentation does say that once you have turned on SQL authorization,
> you can't turn it off. But it doesn't say that you can't turn anything else
> off, either!
> I'll attach the program I've been using. Most of the stacktraces are
> expected, but I'm stumped by that last one.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
