Slashdot Mirror
Inside Apple's Leopard Server OS
An anonymous reader writes "Mac expert John Welch, author of the widely read OS X versus Vista comparison, delves into Apple's Leopard Server OS. He and Information week have on offer a deep dive into what's known so far about OS X Server 10.5, which will be showcased at Apple's Worldwide Developers Conference in June. Welch weighs in on Leopard's iCal, Wiki, file, Quicktime, and mail services, along with Xgrid 2, Open Directory 4, and 64-bit capabilities. What does it all add up to? His assessment: Apple probably isn't aiming at 'big' enterprises; just the same, Leopard Server is shaping up to be a great SMB (small and mid-sized business) product. Welch writes: 'For about a thousand bucks on existing hardware, or for the cost of an Xserve, you get a really solid server, able to support Web services, collaboration, groupware, IM, and file services. You can run it with its own directory service, or as part of an Active Directory implementation out of the box. It provides some features that due to pricing and/or setup requirements, have traditionally been reserved for big enterprises — in particular clustering of both email and calendaring servers.'"
Tell me how fast it copies files...
Thanks
--Bill G.
Sure, but it wouln't provide any obvious reasons for Vista-bashing. Where's the fun in that?
The one thing that has really helped MS in the enterprise has been that the sell an entire solution that all works together. Windows desktops sign into Windows Active Directory run by Windows Servers. Outlook connects to Exchange running on Windows servers with Kerberos AD logins. Office and Sharepoint getting along to create and maintain intranet content.
Apple has made huge inroads with solving the desktop issues of running Unix on the desktop. For the most part though I have seen either Linux or MS solutions on the server for file sharing and web serving and NIS/NFS and such. Even on the mac I would imagine that Entourage connecting to an Exchange server makes up a large portion of the Enterprise mail community.
If Apple can provide a cheaper end-to-end solution from the server to the desktop with LDAP directories, email, calendering, intranet etc - all preloaded on their server hardware and ready to go - then they have a real winner. Hell the cheaper licencing costs they can offer from basing on open-source can help subsidize their higher hardware margins to make this a comparable, if not cheaper, solution compared to something MS from the likes of Dell or HP.
If they really wanted to twist the knife in they should release some client software/drivers for Windows that make it just as easy to connect that to their servers and services as Macs to accomodate the need for having some PCs in a newly mac office.
Now is the time to do this as companies are faced with upgrading to Vista on the desktop, a new version of Office, and soon a new server platform. Most of this means new hardware purchases anyway. They might be able to just swoop in and offer a complete solution the likes of which linux has been unable to - all bundled with and guaranteed/supported on their own hardare as well.
I manage the modest network of a facility with 900 Macs and 120 WinPcs, 1800 users, 3 mail servers, 2 webservers, etc. All servers except one anti-virus server and one Linux mail server, are Mac Xserve running Tiger 10.4 server. All the computers are in a single OpenDirectory domain with all users, including Windows users, authenticating off OpenDirectory. All have roaming home directories which users can access from Macs or PC, managed printer and workstation access... The fact is, we use OS=X server because we have easier fine grain control over desktop applications and resource acces than we do from Windows servers. We find the Mac servers very adequate for our needs. The management software is superb and much more logical than Server 2003 and older and far more coherent and centralized than I have seen so far Linux or BSD
It takes one full time person to run the network and maintain the network and computers, largely due to Apple's remote desktop and net boot reimaging services and the fact that all the Macs can run off two install images: one for Intel, one for PPC. The WinPCs take roughly twice the maintenance time per cpu than the Macs largely be cause of differing drivers.
Oh.. The facility is a public k-12 school. Mac server are also very common in colleges, law offices, large and medium printing companies, greeting card companies, scientific research, and newspapers.
I think that there is a very flawed perception in your argument. You are correct that Apple does not reveal "everything" about new products before unveiling, and with some products (especially new ones) they are absolutely secretive.
But I think it is a common mistake in industry to think that you can do better planning based on the information from any vendor other than Apple (in the Steve Jobs Era). If you take Windows Vista vs. 10.5 as an example:
Microsoft has been touting features of Vista for years now, but if you take a look at the list of those features, and the ones that businesses were planning on building on, you would have been completely mislead as recently as 9 months ago. WinFS (database based file system) was arguably the killer feature that everyone was planning on. And we don't know when and if that will be delivered. And if you are really one of those planners who needs to know the future, then you would know that this feature was originally on the plan for Cario, which was Windows 95.
So Microsoft has been giving out information all along, but you can't rely on that information at all. Sure they have had a beta program going for quite some time... but we are talking about long-term planning here. The people who make those plans do not have time or inclination to play with those betas.
Now Apple on the other hand: I was at WWDC last year, and so got to see a lot of the new API's that Apple was working on, and I got to see a lot of the demonstrations of technologies that will be in 10.5. There are a whole number of technical-level details that Apple gave out, the type of things that are very important for programmers, and systems integrators. I got a great idea of how 10.5 will fit into my employer's network (even better than 10.4).
I didn't get to see the wiz-bang super-secret features that are still secrets, but to be honest, those aren't things I have to plan for until 9 months after 10.5 comes out anyways. The things I need to know to do my planning or programming Apple has made available to me (granted not for free), and the stuff that it would be cool to know, but I don't need to know to get my job done they still have behind the curtain.
And the stuff I saw I know will make it into 10.5 (unless the specifically told me it was on the bubble). Apple has a great track record with that. The stuff they didn't know if they could pull off correctly was excluded from public view. In my mind that helps me make the right decisions, rather than lead me to false expectations.
If they get ZFS working, as is the rumour, and it's not buggy, then I think they may have a killer feature on their hands.
Even Microsoft shops might be inclined to test this out for a NAS box. One of the big reasons why people by netapp boxes is for the snapshot and snapmirror capabilities. With ZFS, OSX would have very similar capabilities for a lot less $$.