Slashdot Mirror
Where are the 'Modern' Directory Services?
MarcQuadra asks: "I've been a Linux user since 1998, and I admin Mac OS X machines at work, but I have yet to find a distribution that comes out-of-the-box with modern directory services. Sure, there are guides to kerberize and set up OpenLDAP, but before I can start pushing Linux as an alternative at work I'll need a few things. Are there any distributions out there that can auto-mount SMB shares as home directories without heavy modification? How about a distro that's based on OpenLDAP and can easily be configured with LDAP-enabled SAMBA and Kerberos? Am I missing something, or is this not a priority with the community at-large?"
Sounds like you want Windows and Active Directory.
LDAP, Kerberos, Samba and all the things that come with that are critical to Linux's survival now. Linux will either live or Die on its ability to use LDAP, Kerberos, SSL and Samba.
LDAP is Linux's ultimate ability that permiates everything Linux can do and makes the many peices of Linux whole. Only the greatest of Linux Users cann use LDAP.
The thing is, its too damn hard, too damn difficult, and there is not enough documentation and configuration too;s for LDAP out there. I've spent three years on LDAP - I know.
Expensive, insecure, closed. Choose 3.
Yes having a setup for LDAP with SAMBA tied in would be a plus, you have to consider why it hasen't happened yet.
Only fairly large shops NEED that and they only need to set it up once. The existing howtos appear to be addressing that need well enough that it has not become a big enough itch for anyone to scratch. Again, because once you know enough about it to write the wizards to make setting it all up easy, you have your site done and will probably will never need to do it again. So until a distro vendor sees it as a big enough selling feature to undertake the work I doubt it will happen.
Democrat delenda est
The venerable 4.4BSD automounter (am-utils) is nice for auto-mounting nfs. nis isn't ideal but works, and can do much more than just throw passwd around. In fact, I'd not use it for the passwd stuff, but just announce amd maps with it.
samba is quite useful, even if I still have to look at its new 3.* features. LDAP is somehow the obvious directory choice, even if it is clearly not ideal. Maybe that is because all others are even less-than-ideal, or just not open and/or sane enough. RADIUS is often only used by (I)SPs and the like, but could be used in the local network, too. And of course there's kerberos.
The only real problem is lack of vision (because there's so many ways to do it, and every company needs something different, maybe?) and, as already remarked, the combination of all the HOWTOs into something more closely knit together.
But the parts are all there, no doubt about that. So far it's only been the commercial sector that's been doing the integration and/or building their own solution.
The money you spend on new hardware will be far less than what you'll spend in time and trouble getting a half-assed Linux solution together.
You want Mac OS X Server. Trust me on this.
Because 'the people upstairs' who make purchasing decisions are dead-set on x86 hardware in the server room.
They are wrong. Explain this to them. That's part of your job.
Also, there's perfectly good x86 hardware in there now, I'd rather use itr than pay Apple for new metal.
Given that this "perfectly good x86 hardware" is absolutely incapable of doing what you want it to do without a massive investment of time and effort, it seems obvious to me that it's not "perfectly good" at all, is it?
Run the numbers. You will find that buying an Xserve will cost you much less than trying to make your jury-rigged solution work.
Grab a copy of Open Enterprise Server from Novell. Its in open beta and is basicly what you are asking for. It may be more than your asking for actually as they offer lots more services than you need.
I have had a chance to play with it, Its Suse with Netware services on it basicly. NDS is probably the nicest directory out there and it has LDAP built into it so you can connect other Linux distros into it if you don't want to just run OES.
They have made Samba talk to NDS so you create user objects in NDS and it works through out the system. They plan on replacing Netware with OES so its well polished.
OK, a turnkey alternative to AD is highly desirable, but doesn't solve the whole puzzle.
What is needed is for OSS applications to be tightly integrated into this environment.
Microsoft's biggest selling point is integration of it's applications with each other and AD. That's what enterprise customers want(and need) to hear, and are willing to spend $$$ on.
Dunno, they've been in business quite a bit longer than any other major Linux supporter, excepting IBM. I don't think Novell will be disappearing any time soon.
Weapons of Mass Analysis
Well I'm not sure about how much you've got invested in PC's already, but I think OSX is more of an investment. Microsoft and Linux require faster and faster hardware every year, while OSX gets faster and faster on the same hardware. Assuming this trend continues, this could reduce your upgrade cycle quite a bit.
Yes, a forest should never be necessary in theory, but that's not reality. You end up merging with other companies that have their systems setup differently and the only easy solution is to join your trees into a forest.
Or you have different divisions that insist on their own autonomy, and each sets up their own domain. You can imagine that this would be common in university environments.
The thing in contention here is "demand". Now, OK, frex; IE has 90% of the market, Firefox less than 10%. A conventional view says that IE is in considerable more demand than Firefox (or Opera). Now, allright, I can accept that, but I don't agree with it. The bottom line is that no one (or very few) actually want IE but they have it and don't want another browser enough to learn how to download and install (or are not permitted to... or...). Given that you had to choose and download a browser would the ratio of 90/8/2 (IE/Firefox/Opera) be the same? I sincerely and very strongly doubt that that is the case. IE is crap in comparison to either of the others mentioned. So when people talk about "demand" or "market demand" they are not talking about demand in the english use of the word at all. They are talking about usage figures not how much one product is valued/wanted/desired over another. If the "market" was on equal standing the situation would be very different.
So, what I mean when I say there is no "demand" for MS products is that no one really likes them. No one really wants them. And if there was something that was not harder for them to deal with and they had a real choice they would abandon MS gleefuly and rapidly.
I'm actually quite sick of the pro-anti-Microsoft war and don't particularly care much about it, but that isn't going to make me abandon the truth of things. MS is a bag of worms, Linux was developed from a terminal emulator and shows it, UNIX (although my favourite) is thirty year old concepts overlaid with patches and extensions usually badly implemented. It is _all_ crap. Live with it.
Anyway, it will all pass. MS has most likely had its day in the Sun. It's optimal strategy for long term survival now would be to fund say, twenty guys to work on Hurd (and maybe another 20 for EROS too). To stay ahead and set directions, to truly open just about everything except the UI. In the end it is only the UI - the user experience - that is important. So, right now, MS has sufficent resources to fund as much of the OSS movement as it wants. If it (MS) funded say 1/3 of the current OSS developers, how could it not stay in front? Wouldn't worldviews suddenly change?
Zero Sum (don't amount to much). [root@localhost]