Best Practices for Separating Database Security from Application Tier - oracle

In the Oracle documentation, it is repeatedly mentioned that it is preferable for all users of an application to be database users instead of just records in some kind of 'user' table for numerous reasons. For example, at docs.oracle.com/cd/B28359_01/network.111/b28531/… -- "Where possible, you should build applications in which application users are database users. In this way, you can leverage the intrinsic security mechanisms of the database."
I am reasonably familiar with what they call the One Big Application User pattern, but I have been suitably convinced that this is not the best way. However, it seems that creating a database user for every web user is not particularly possible either, so my question is, what are some best practices for keeping the security definition of your data separate from the application that accesses it?
Some ideas:
A secure interface in between the database and the application. Application never knows about the database.
Create views that utilize application context to automatically filter out rows that the application user cannot read, and only grant the application user access to the view.
Similarly create procedures that utilize application context to only allow authorized data access/manipulation.
I am currently using Oracle 11g XE, and I'd like to avoid solutions that involve additional Oracle products like their Internet Directory or Middleware.

Related

Multi app architecture and databases

I have just developped a Human Ressource WebApp in Laravel.
So in this app, I have my user database.
Now, I have 2 old apps, that I'm about to rewrite / Migrate to Laravel.
The first one is to manage employee payments.
The second app is for monitoring my salesforce in field.
So, Basically, I will basically rewrite them.
My Question is about architecture. I will develop those 2 apps separating APIs in Lumen / Laravel for each app, individually, and then writing clients in Angular2.
I will use OAuth2.0 to set authorization and scopes between apps.
Now, My indecision resides on Database part.
Should I keep database separated, include database in the same schema, or how should I deal with my databases.
Today, I have 1 DB per app, but also, I have redundancy in users: Each App works with the same users, and so duplicate this info.
Right now, When User is created in App1, it calls APIs to create it in app2 and app3. This is not so nice, IMHO.
So, I think I would like to have a single User Database, but I don't really have experience of that:
should I extract User's info in a central database for all users, should
should I keep User's info duplicated between my DB?
if I keep it in 3 differents DB, how should I make JOINs, transactions, Foreign Keys, etc.
I know there is not 1 answer, it depends of what I want, but this is also my problem, I don't know which question should I be asking to make my decision???
It is important to mention that I don't have so much traffic, max 1000 users, and they all belongs to my enterprise, it will not increase to much. So, really, make a scallable app should not be my priority right now.
Any Advice will be appreciated!!!
Since you have shared databases between your apps, the common practice is to have it in the same Laravel/Lumen app.
The way you can achieve this is separating the apps inside route groups, namespacing the controllers hierarchies of each app and declaring inside each model the connection you will be using for this model.
This way you can use the same DB connections for all of your apps, and share the same data without the need of duplicating it.

Is it a bad practice to create a large number of users in an Oracle database?

I want to design a database for a student portal. I want to create a student role, so for every student I need to create a user and assign the users's role. It seems like many users will be created. Could there be any problem such that large resources will be needed.
The number of user accounts is not the issue - you can create as many as you think you need. The issue that is of more concern is how many concurrent user sessions - users logged into the database at the same time - will there be? Each user needs memory to do their work and this must be allocated. If you have more simultaneous user session than memory can accommodate, this can be a problem. Read more here and here. If you will have a huge number of concurrent users, consider a shared server connection setup, rather than the default dedicated connection.
Another thing you need to be aware of is the financial cost implications depending on whether you have a per user licence.
However as others have already mentioned applications usually connect to the database using a single set of credentials, end users are not usually defined as database users, and therefore you would not be able to use the database defined roles to differentiate between the permissions.
You could likely have two database defined roles DBA or SYSADMIN, and PORTAL_APP.
The end user roles would need to be handled within the portal application, and this should be consistent on the front-end and the back-end of the application.

Migrating Single-Tenant to Multi-Tenant application

We are upgrading a single-tenant application to a multi-tenant (separate DB for each client) SaaS application. My responsibility is to design the tenant provisioning part. I am thinking of the following steps.
While our admin will install our software, one DB should always be get created.Argument for this step is that Admin will install the software only when at least a tenant comes up. So, creation of a single db as part of installation of our software should make sense.
The common meta data will be stored in a common place. We are thinking of using ZooKeepar to store the common meta data.
Each tenant will be given a subdomain. For example www..samplaTenent.com . This tenant_id will be extracted from the URL in each request and will be used to identify the particular tenant and respective DB.
If more tenants need to be provisioned, the admin will create a new DB instance through user interface.
Technology used:
Java 6
Struts Framework
MS SQL
Now, I would like to get expert opinion about the proposed solutions. Is there any other way by which we can build a robust, secure and scale-able SaaS application?
Thanks.
IMHO, I would suggest that you should have a sibgle code base that can handle the multi-tenancy aspects be it configuration of db, settings per tenant etc.. I am not aware of the code base that you have now. Hence, you can make the provisioning part as follows,
You provision a new tenant by getting all the prerequisite info. You can generate the URL for this tenant automatically or allow the tenant to choose one in the latter a validation has to be made for the URL and the tenant combination.
On this step completion, you can have some scripts at the server that clones a master code base and rolls out a new tenant specific db. So things are automated.
The next step will be to store the tenant db and the app details in your common or shared db so that when a tenant logs in, you can redirect him to his URL. The db details will be required if you want to handle all of the tenant and his user authentication from the main app.
The main perceived advantage of having a single code base is that the customization and configuration can be in one place with ease of code maintenance and bug fixing or enhancing. Though this may not work for all scenarios, this is what I feel more suitable option on the long run.
Please share your thoughts or post additional queries that you may have on multi-tenancy.

Apex "security module"?

My manager keeps talking about how I will be "developing" an Application Express "security module", however from what she told me we need to have, I don't see what there would be to develop, seeing as Apex already has authorization/groups which allow for various groups of people to see various content.
Is there something that I am missing? What does she mean by a "module", or is it just general wording?
APEX provides several different ways to authenticate users. One approach is to use the "Application Express" authentication scheme and just to create APEX users. Another approach is to use the "Database Account" authentication scheme and to create Oracle users. A third option is to create a custom authentication scheme and to implement your own user management functionality.
Application Express authentication tends to be the easiest to deploy for a small application but tends to get unwieldy over time. It's hard, for example, to give an application administrator the ability to create APEX accounts. You can't tie an APEX account in to a single sign-on solution. It's not easy to integrate with the permission management systems that other applications use. If you're deploying an application in a large company, the last thing the security department needs is one more place where they need to create user accounts, manage privileges, de-activate accounts when someone leaves or changes roles, etc.
Database authentication tends to be more scalable than APEX authentication since Oracle database account provisioning is likely already part of your organization's authentication and authorization infrastructure. On the other hand, that still means that you're creating an Oracle database user for every user you want to create in your application which probably involves a call to a DBA (technically, you could create database users from your application, but most DBAs are going to be concerned about the security implications of that). If you intend to create an internet-facing application with tens of thousands of users, database accounts may get unwieldy.
I'd wager that the vast majority of medium to large-scale APEX applications use a custom authentication scheme. That may involve creating a USER table where you store the username & the hash of the password or a query against an LDAP/ AD repository. That sort of approach provides the most flexibility since you can code whatever you'd like into the authentication system. You can hook into whatever custom authentication/ single sign-on solution the organization happens to use. It probably makes creating new users from within the application much easier (obviously depending on how the authentication system is designed).
My assumption is that your manager is expecting that you'll be writing a custom authentication scheme for your APEX applications.

How to manage user/permissions in an environment web/PL/SQL correctly?

My team will develop an internal (known users) application that has an architecture based on Java as front-end and PL/SQL as back-end. So, currently we are thinking in a better solution to manage the user/permissions, and we have two options:
Each user has their own database account, granted with the permissions. Currently the legacy system use this approach and I don't like it because it manages permissions based on database objects' granularity. So, I believe it is a bad choice to have a database connection per user. Can you see more cons here?
Build some tables at database to store the users and theirs permissions/profiles and build a PL/SQL procedure to do the login, generating a token and include a parameter to all others PL/SQL to verify this token and then authorize (or not) the execution.
So, you can ask me: why not just manage your permissions in your web-application? Answer: Those PL/SQL are already done and are used by all legacy systems, and this web-application should behave according it (ie. User permissions should be managed by the PL/SQL and its granularity based in please.)
How do you proceed in this case?
I think using the database's built-in mechanism is always to be preferred over rolling our own. And that applies to logging in users as much as anything else.
The biggest single advantage of dedicated user accounts is that we can link a given session with a named user. Well, yes, duh. But the point is, doing thinks like auditing user activity or tracing a performance issue in some process is way more difficult in web applications with generic accounts.
To address your main objection, we don't have to manage database privileges at the user level. That's why we have roles. For normal users, a role will provide sufficient privileges.
So:
define a set of roles which match the various business jobs your application serves.
grant system and object permissions to those roles; remember that roles can be additive (i.e. we can grant privileges on a role to another role).
grant roles to the users.
Find out more.

Resources