Flylib.com

Books Software

 
 
 

Answers to Chapter Review Questions

     

Answers to Chapter Review Questions

A1:

The CIFS Server software simply allows the machine to run some additional server daemons and hence does not require a reboot. CIFS Client adds support for the CIFS filesystem type. This requires the installation of a kernel module. It is the installation of the kernel module that forces a reboot when installing the CIFS Client software.

A2:

The IO error is seen because CIFS requires each user be authenticated before being able to access the filesystem in question. The current user obviously has not been authenticated.

A3:

The cifsmount “P command requires a plain-text password on the command line. This can potentially be seen by everyone on the system using the ps command.

A4:

The main reason is that a Windows administrator could inadvertently add a user called root into the Windows domain. When an HP-UX root user attempts to log in to an HP-UX machine, there is the possibility that he will be authenticated by the Windows machine possibly with a different password than the real root password. This could be seen as a major security breach, not only for the root account but also for any other account with a name with the HP-UX password file and within the Windows domain.

A5:

A CIFS "user map" is a text-file reference on a CIFS Client referenced via the /etc/opt/ cifsclient /pam/smb.conf configuration file. A CIFS "user map" allows a UNIX administrator to map a UNIX username onto a potentially different username within a Windows domain/workgroup.

     

Chapter TWENTY ONE. An Introduction to LDAP

Chapter Syllabus

21.1 >Introducing the Lightweight Directory Access Protocol

21.2 LDAP-UX Integration Products

21.3 Step-by-Step Guide to LDAP-UX Client Services

21.4 Next Steps

Originally, UNIX account and configuration information was stored in a series of text files. As the need to share this information across systems increased, the first widely accepted product named Yellow Pages, and later renamed to Network Information Service (NIS), was developed by Sun Microsystems. NIS provides network-wide management of many UNIX configuration files (e.g., /etc/passwd , /etc/ group , /etc/services ). An NIS master server generates maps based on the configuration files and transfers copies to slave servers. On NIS client systems, operations reading the configuration file are redirected to send a request across the network to retrieve the information from an NIS server. While providing a high degree of backward compatibility with file-based configuration, NIS has limitations in scale and security that prevent it from being easily deployed in enterprise environments. NIS does not support delta-based updates, causing entire maps to be transferred to all the slave servers. These maps are transferred across the network unencrypted. The underlying database used by NIS servers can support a limited number of entries, requiring administrators to break up the data by creating multiple NIS domains. Despite these shortcomings, NIS is widely used today across a variety of UNIX platforms. NIS+ was introduced as a successor to NIS to provide greater scalability and security. While succeeding to some extent, NIS+ has not achieved the level of acceptance of NIS. UNIX administrators have reported that the level of complexity in administering NIS+ often outweighs the benefits. With the arrival of more general-purpose directories, the potential of a more powerful generic directory has supplanted the NIS model in the imagination of the UNIX community. NIS+ suffers from lack of interoperability and, therefore, it is missing the much-needed flexibility in hybrid environments. Many organizations including HP are moving support away from NIS+ in favor of LDAP because they see LDAP as the long- term solution. The acceptance of LDAP and the deployment of LDAP directories in many enterprises have created a need for existing UNIX clients from a variety of vendors to access data stored in an LDAP directory.

LDAP directories can play many roles in an enterprise, one of them being a naming service for POSIX systems. Specifically, LDAP can provide a scalable replacement to an NIS-based architecture. Just as NIS became a de-facto standard, a schema defined by RFC 2307 provides a standard way to represent POSIX naming information (NIS databases) in an LDAP directory. Aside from scalability, LDAP directories (often with the help of a meta-directory) offer the promise to integrate many disparate applications such as HP-UX account information and a Human Resources database, thus consolidating data and administration. For example, a name change in an HR database could change the finger information in the HP-UX account. And LDAP directories appear to be the backbone for future security mechanisms, such as a public-key infrastructure. Many Web browsers can interface directly with an LDAP directory in order to obtain user credentials as well as the necessary public keys to allow symmetric and asymmetric cryptography to take place.

In this chapter, we establish a cookbook approach to setting up an LDAP directory containing the information that we would normally store in the text files we mentioned above, e.g., /etc/passwd , /etc/group , /etc/services , and so on. This is a basic introduction just to show you where to start. Beyond this introduction, we are looking at a much more detailed understanding of products such as Netscape Directory Services. That is a whole different certification and product track that some of us might want to investigate. Essentially, the object of this introduction is to demonstrate how to store user- related information in an LDAP directory. This information could be accessed by many different pieces of software ranging from the login process to email to Internet browsers. We start with an introduction to LDAP and then get straight into using the software.