unix

I've read a number of books on Samba starting with one written in 1998 by John Blair. Samba books are commonly written as references, borrowing heavily from other books and man pages. They are also often written from a developer's point of view. "Samba-3 by Example" is different: it is written from a system administrator's point of view. You the reader become the sysadmin of a growing company, and learn how to grow your Samba solution with the company. Each chapter has an "Implementation" section with good hands-on examples. Also, this book gets in to more advanced topics such as integrating Kerberos security into Samba, the black art of migrating data from Windows servers to Linux/UNIX servers running Samba, and the new "passdb backend" smb.conf parameter. Finally, it addresses Samba-3 which is now standard with most Linux distributions as pre-built RPMs. I would definitely recommend this book if you are going to buy a book on Samba.

John Terpstra had done a great job showing how SAMBA can be used in real world scenarios. The scenarios are representative of both the environments and customer needs that I have had to address as a consultant. More importantly, it was great to read a book from the Open Source community that was not a diatribe against Microsoft, but rather focused on how one gets Windows and UNIX/Linux working seamlessly in a heterogeneous environment.

The open and honest discussion around interoperability with Windows versus replacing Windows is quite informative. The book did not contain hyperbolic discussions around the technical or ethical benefits of Open Source. This is a book that I would recommend to friends from either the Windows world or FLOSS space (Free/Libre/Open Source Software).

Chapter 10, "Active Directory, Kerberos, and Security," was especially helpful and informative. I particularly appreciated the examples showing both the UNIX and Windows approaches to various problems. The thorough drilldown on the technical underpinnings and explanation of the historical issues gives a good perspective. This chapter helped me understand SAMBA coexistence/interoperability in Active Directory based environments. Having details on security, performance, locking, and the real world conundrums that arise made it possible to translate the books scenarios to my specific needs.

The book is easy to read and clearly details the steps required when setting up your SAMBA environment. The book includes some great references, and lots of accurate, step by step guidance. Overall, I would highly recommend this book to anyone trying to set up and use SAMBA.

The author of this book is a myopic Nix evangelist.

In the real world there are many reasons for Nix and Windows to Co-exist in a variety of relationships. There is also strong argument that Microsoft's semi-propriety file sharing protocol is better than NFS, regardless of one's feelings on that argument Windows supports the MS protocol and only recently can you download support for NFS without paying additional licensing fees to MS or a third party vendor, and this free support for NFS is only for WinXP and 2000/2003 Server. In any multiplatform environment the protocol of choice is going to be the MS protocol, and the way Nix implements it is through SAMBA and the heroic reverse engineering efforts of the SAMBA team.

The reality of SAMBA is that it is a reverse engineered product with a very limited budget. For those of us coming from the MS side, SAMBA is a bear to administer. My dominant view of reality is different from the author -- the author presumes Nix on the Server Side and Windows on the Desktop, with any Microsoft Servers as an unwanted nuisance. My view is of either very mixed environments (which may also have a strong Mac presence) or of Windows dominated environments where Nix makes its entrance because of either being better suited to some specific task or allowing some significant licensing fees to be avoided.

The choice and structure of the examples was pretty much chosen in light of the author's bias and lacked the full depth that could have been reached by also working examples from a Windows centric view.

An issue that I have found inadequately addressed all around is that in Server 2003 Microsoft introduced new "security features" which pretty much break interoperability with SAMBA. One of the important goals of Samba 3 was to fix what Microsoft broke, the author gave only one useful example of making Samba 3 work with Server 2003. The author omitted how to disable those features which make interoperability with Samba difficult. And there are many reasons to choose having an easier time making things work over the enhanced security provided by the new features, if only as a temporary measure to reach a point where things work while resolving the issues that prevent it from working in the more secure configuration.