Submitted by “Novell’s long journey from NetWare to Linux is finally complete. On Oct. 8, Novell released Open Enterprise Server 2 to its customers worldwide. Shortly after acquiring SUSE and its enterprise-focused Linux distribution, Novell announced that its follow-on to NetWare 6.5 would ship as a set of network services that could run atop the NetWare and the Linux kernel, OES 1.0. OES, which began shipping in April 2005, was the first major step in Novell moving NetWare’s services from its native operating system to Linux. Now, with OES 2.0, the NetWare operating system kernel, NetWare 6.5 SP7, is still there if you run it, but it runs on top of the Xen hypervisor. You can also run the NetWare services, or a para-virtualized instance of NetWare, on top of Xen with the SLES 10 SP 1 kernel. So, if you’re wedded to NetWare and its way of doing things, you don’t have to wave good-bye to it.”
Novell has an excellent suite of tools for managing a network. eDirectory and Zenworks Tools (iPrint, etc) are great. Work great, and are very robust. I miss using them (currently in a 100% MS shop..)
One question though: Did they finally lose the Netware client? That thing was always buggy (because of MS’s ever changing hidden API more than anything)…
No, the client is still around. There are ways to get around using the client, but it’s still probably the most used “gateway” for NCP connection. I think the Vista client was just released, along with the next version of the Linux client. I haven’t had any time to muck around with anything new on this front in the past three months unfortunately.
I’m hoping to set up a small testing lab within a few weeks and install OES2 on top of SLES, OES2 NetWare as a Xen virtual machine, and a bare metal NetWare 6.5.7 and see how they interact with each other. I wonder what my boss would say if I dropped a $3000 PO for a ThinkPad with a VT enabled Intel cpu and 4GB of ram for a “test lab”?
I guess my other question would be: Is the client still buggy? or has the Microsoft-Novell collaboration (among other things) helped them?
Yes, the client still has some bugs apparently, judging by the constant flow of 4.9X SPX updates that have been coming out for a couple of years now. It’s fairly stable from what I’ve experienced(outside of a severe 4.92 SP2 show stopping issue), but we don’t change much from the default configuration during install, so my evidence is anecdotal.
I seem to remember reading a blog posting explaining why the Vista client was taking so long to come out; basically Microsoft was changing so much right up until Vista went gold, Novell had to basically play catch up after Vista was made public.
If I had to guess, I’d say the deal has had zero impact on the Novell client.
My only hope is that it’s better than the original OES. That one was a dead dog.
While I have a lot of love for Novell as a whole the new (IBM) execs are killing the company. Support sucks and is expensive. What they offer small business sucks compared to what they did offer. Mostly they still seem unstable and the rumor that they are working to devalue the company so IBM can buy them up is starting to sound true.
At any rate I’m looking forward to trying out OES 2.
Devaluate so IBM can buy them? Who the hell would want that! Even if it means buying for less, the bought company would have loose all it’s reputation and client base, which IMHO worths more than the company owning itself. IBM has no interest in buying a linux distro, it would have already enough expertise to start over a new one. But they doesn’t want to add a new player in the field. Big Blue as long understood that they need to limit Linux fragmentation for it to get on MS.
Novell as a lot of IP assets, a good Linux distro and a handsome of clients. Yes it could interest IBM and IMHO I’d say it would be profitable to anyone to see a stock for stock buy-out. IBM could simply swallow Novell, offering a mix of cash and stock options. Think about it, IBM buying Novell means : Big Blue Linux (aka SuSe), IBM owning Unix, and IBM putting it’s enormous work force behind a distro. Suddenly Linux future seems even more brighter (as long as OS/2 manager stay far away LOL). [EDIT] Ho and btw, if there is rumors of a possible bid by IBM, Novell values will be far from dropping![/EDIT]
Edited 2007-10-09 20:18
The rumor started when Novell started putting ex-IBM execs on the board and as company execs. It didn’t help when they started killing off all the small business products. After that they started liquidizing assets; killing off projects; laying off workforce; etc. While those things don’t really mean much in the practical sense they do give fuel to the conspiracy crowd.
As for devaluation. If IBM were in the market to buy up Novell one would think that getting IBM loyal people on the board and in the company seats would be smart. Next getting the stock price to drop would be genius. And to do it all without the general public really knowing it would be outstanding. Sure the stock would rise, but if they got it low enough they could still buy it cheap.
At any rate it just speculation (mostly on the part of people who are unhappy with the direction Novell is going). Why IBM would want Novell I couldn’t really say. What they would do if they did purchase them I couldn’t really say.
For Novell to have a future, OES basically had to be an amazing competitor to the OS that is taking away Netware’s market share – Windows 2003. Put simply, it’s not. Keep in mind that the company who Novell is in bed with is producing the OS that has been destroying their market share for some time now. Novell should just have focused on making something good.
The graphical tools that should be available to go head-to-head with Windows 2003 are simply not there, and there has been an awful lot of crap thrown around as to which toolkit to use (Mono obviously isn’t helping), better integration of existing Netware tools into the Linux distribution and open sourcing a lot of Netware tools and the OS to enable that integration to happen and get people using Netware stuff again. What they’ve done is produced something that people who have bought into Red Hat and Windows can’t see anything compelling in, and they’ve completely disaffected their existing Netware customers. Well done Novell.
Novell’s traditional culture hasn’t helped, as we have seen with the takeovers of Ximian and Suse and the arguments about what to use and what not to use. A decision should have come from the top down, and it didn’t. Just ask Chris Stone, John Vigeant of Xensource, Alan Nugent and a lot of other good people from Suse. A company doesn’t lose people like that unless there’s something seriously wrong.
It looks as though there’s another round of layoffs going on at the moment, there’s still lacklustre growth and revenue from any new business, and the knee-jerk reaction from Novell’s executives have only speeded things up.
What graphical tools are you talking about? What does Mono have to do with it? Is there a real benefit to open sourcing NetWare services, or more specifically NCP and eDirectory and dealing with having to remove third party code and encryption techniques(such as RSA bits)?
I’m not going to sit here and defend Novell’s corporate culture. I have no experience to be able to. Nor am I going to defend their management decisions because I don’t agree with all of them.
However, I don’t see how all this directly impacts OES2 and specifically OES2 Linux and whether it is a solid performer in the server room. Apparently it’s feature equal(or better) than OES2 NetWare.
I reserve judgment on the product for myself until I can actually find the time to install it and begin testing.
Edited 2007-10-10 12:49
What graphical tools are you talking about?
Exactly my point. Equivalents that can go head-to-head against its biggest competitor – Windows 2003.
What does Mono have to do with it?
Because Mono is now supposedly their official development environment for producing such tools. I don’t really see it helping them much to produce the aforementioned, and badly needed, tools.
Is there a real benefit to open sourcing NetWare services…
Novell are the ones trying to move to Linux. Their Netware customers didn’t ask for it, because Netware did what it was good at – being a network and file sharing OS. They just wanted Netware to be made better. In moving to Linux, Novell really needed to convince their Netware customers that there were benefits to be had, and they needed to put a lot of work into making sure that what was good about Netware could be transferred as seamlessly as possible to their Linux replacement. If you are able to open source those then it helps in getting them integrated into a Linux environment, and it also allows more people to use and communicate with Netware, which is quite important considering how Netware usage has declined.
However, I don’t see how all this directly impacts OES2 and specifically OES2 Linux and whether it is a solid performer in the server room.
It has everything to do with that with respect to its competition.
Apparently it’s feature equal(or better) than OES2 NetWare.
That’s yet another problem. There should not be a Netware and Linux version of OES. It should be one product. Arguably, SLES and OES shouldn’t be two separate products either. There should be one, as we have with RHEL and Windows 2003 with minor variations.
I gather you mean what does Novell have in the way of tools like Microsoft’s MMC? iManager would be the closest thing, but that’s missing some things, mainly groupwise plugins. I’ve voiced my opinions directly to Novell about that. As for mono, I have thought of that. But, if they’re going to stick with iManager, they just need to get the rest of their products merged into it. Basically, pick something and stick with it. I’ll agree that they lag behind Microsoft in this regard.
NetWare, begining with version 5.1 I believe, has been slowly separated out from its services or rather the services from it. Running NCP or eDirectory is no longer platform specific. So all you’re left with is the operating environment. I don’t think NetWare has had a large scale rebuild in a decade or so. There is no bare metal 64-bit version currently, nor will there be ever. NetWare tends to struggle going over addressing over 4gb of memory. Third party support has been drying up since the mid 90s. Nobody wants to go through the pain of writing NLMs. It’s not economically feasible. The same goes for drivers. Everything I’ve been lead to believe, it’s a huge undertaking involving a heavy rewrite of the OS for little or no benefit over running those same services on the Linux kernel. We can’t make the assumption that anyone would actually want to hack against NetWare if it was every open sourced anyway.
In terms of sales and market share, yes. In terms of pure performance and feature set, you have to take the two products out of the politics and put them head to head.
It will be one product. The issue is, you can’t just up and do away with NetWare in one movement. You have to continue to support it and provide a migration path. This is a phasing out process and OES2 is just next step of that process. Moreover, OES2 has been turned into an add on package to SLES10 rather than a separate operating system for lack of a better descriptor.
I’m getting off point now. To sum it all up, my main point is that NetWare is a dead end in terms of development and Linux is the best option as a successor. Rewriting NetWare to support new hardware would be a very time consuming and expensive process. Open sourcing NetWare would be a very time consuming and expensive process. In my opinion, Novell would be wasting their money doing either. Open sourcing the services however, that’s a whole different story…
I’ll agree that they lag behind Microsoft in this regard.
They log a long way behind, and there’s no reason they should because moving to Linux should have given them what they needed.
There is no bare metal 64-bit version currently, nor will there be ever.
Basically, Novell couldn’t afford to maintain it or keep up with the hardware world. We all understand that.
The same goes for drivers. Everything I’ve been lead to believe, it’s a huge undertaking involving a heavy rewrite of the OS for little or no benefit over running those same services on the Linux kernel.
That’s exactly what they needed to do if they wanted their customers to go with them. Not everything is going to work, and the concept of NLMs would have to go the journey, but everything else needed to fit and Novell needed to get everyone so excited about it that they would want to go along. That’s also what I’m hinting at when I talk about graphical tools.
In terms of pure performance and feature set, you have to take the two products out of the politics and put them head to head.
Same difference. Windows Server has been eating share at the expense of Netware for years, and in terms of features Novell needed to arrest it. They havent.
It will be one product.
It isn’t though, and it shouldn’t have been until they released it.
The issue is, you can’t just up and do away with NetWare in one movement. You have to continue to support it and provide a migration path.
That’s Novell’s problem, and they’ve done it very badly. In fact, they don’t have any kind of adequate migration path. Mostly, it’s left up to customers all in the name of choice. People don’t want that – they want a network operating system and they want Novell to tell them what they are selling.
This is a phasing out process and OES2 is just next step of that process.
Phasing out shouldn’t be necessary. At the moment, many people are caught in limbo between one and the other. Virtualisation provides Netware with something of a future for people who need all of it, but OES should have been released with all the tools necessary for people to migrate from Netware to Linux and have it be a drop-in replacement. Many jumped ship when that didn’t happen.
Moreover, OES2 has been turned into an add on package to SLES10 rather than a separate operating system for lack of a better descriptor.
That’s good, yes, but the two are named completely differently, and really, there should still just be one product.
To sum it all up, my main point is that NetWare is a dead end in terms of development and Linux is the best option as a successor.
I think everyone can understand that, but Novell’s customers don’t care and simply want a network operating system to do what Netware did with quite a bit extra. Frankly, Novell have failed to provide it.
Open sourcing NetWare would be a very time consuming and expensive process.
I’m not saying they should open source all of it – just the parts that they can get up and running on Linux.
How many jumped ship because of that and who? Novell’s entire reason for running 2 kernels parallel with similar feature sets was because the customers wanted it. You can’t rip and replace like that. If you really want to drive customers away, then having OES services sitting on the NetWare kernel only one version and then eliminating that and going with the Linux kernel only on the next version would basically kill your File/Print business. Migrations paths are not supposed to be static rip and replace scenarios that are equal across the board. They need to provide as many options as possible for the varying situations one customer might have compared to the next. To do so is ignoring your customers and what they might want.
That’s just the point, they are not one product. OES proprietary services sit on top of SLES, but it provides radically different services than basic SLES. Not everyone needs those services. Otherwise you’re basically using the analogy that Exchange 2007 should come with Windows Server 2003 because it direct extends the server and the directory.
Have you used OES2 yet? Have you thoroughly tested it to make that conclusion? I don’t know about you, but have a load of NetWare servers 30 feet away from me right now. I want to be able to run the OES services on new hardware which is becoming more and more difficult on the NetWare path. I also can’t go in to my server room and rip and replace all my NetWare servers with OES2 Linux boxes. My next(as in time) best option moving foward seems to be OES2 Linux at this point. Although to be fair, I have not tested it myself, so my opinion on what the next best option is based on what I have read, what I have heard from people who have gotten their hands on it already, and the realities of supported hardware with NetWare.
Why? I’m not saying it’s a terrible idea, I’m just curious as to what benefit would it be for them? Everything they need from NetWare has been ported to Linux already.
Edited 2007-10-10 19:44
How many jumped ship because of that and who?
Well, I’m afraid you’ve been out of the loop. Customers do not want a choice of two kernels – they want to know what Novell is selling. Customers have been jumping ship for quite some time, hence Novell’s somewhat dire financial results and a new round of layoffs apparently.
Novell’s entire reason for running 2 kernels parallel with similar feature sets was because the customers wanted it.
No, they didn’t. Novell went and did that of their own accord, and they thought that’s what customers wanted. You have to get the meaning from what customers say, not what they actually. Rather like women! Who in their right mind thinks that customers want to choose a kernel?
You can’t rip and replace like that. If you really want to drive customers away, then having OES services sitting on the NetWare kernel only one version and then eliminating that and going with the Linux kernel only on the next version would basically kill your File/Print business.
Novell are doing that fine, I’m afraid. The key here is decisiveness. Novell should have announced that they were moving away from Netware, but should have come up with a clear migration plan and tools and a clear incentive for customers to move to the new Linux offering so that it was as damn near a drop-in replacement as possible – with tons of added goodies to keep them. There’s simply no reason for any Netware admin today to move to the Linux version of OES, simply because it’s different, there’s nothing compelling to move to it for (other than Novell can’t keep up with hardware support for Netware, which is not a customer’s problem) and it’s simply a Linux version of something that does what Netware does, except arguably worse in his eyes with nothing extra. Many organisations are simply moving to Windows servers completely to manage their networks.
That’s just the point, they are not one product. OES proprietary services sit on top of SLES…
They are one product, and they should both have the same name. It’s an OS. Why is the OES stuff proprietary anyway? I thought Novell was an open source company (which causes yet more confusion for people)?
Otherwise you’re basically using the analogy that Exchange 2007 should come with Windows Server 2003…
Errrr, no, because one’s a mail server and one’s an OS.
Have you used OES2 yet? Have you thoroughly tested it to make that conclusion?
We have one big Netware using client, and they’ve already been making the shift to Windows servers to replace what Netware is doing. Once Netware support goes completely, apart from when it is being run in a VM, then so will Netware – and they won’t be moving to OES Linux. It’s just too much hassle, and there is little incentive to do it.
I want to be able to run the OES services on new hardware which is becoming more and more difficult on the NetWare path.
You need an awful lot more incentive than the ability to run Netware services on new hardware if you’re going to move to something new. If you’re moving to new hardware then you might as well just move to Windows, or Red Hat, and that’s the view many companies are taking.
I also can’t go in to my server room and rip and replace all my NetWare servers with OES2 Linux boxes.
That’s about the size of it. That’s exactly what Novell should be helping you to do – as painlessly as possible, with lots of goodies to make the whole process worthwhile.
Why? I’m not saying it’s a terrible idea, I’m just curious as to what benefit would it be for them?
That sounds just like Novell themselves. The open source company………that isn’t. Because they need to prolong the life of their OS and their services in another operating system, that operating system is open source, and one of the benefits of Novell using Linux and open source software is shared development and Netware services usage increasing.
I know Windows has been dominating NetWare market share since the late 90s before Linux was a blip on the radar, and it continues to this day. Neither your nor I really know if a dual kernel approach during a phase out of the NetWare kernel has any impact whatsoever on that outside of anecdotal evidence.
Well, probably the customer with an installed base of NetWare servers, who need(ed) time to plan a proper migration and get their staff properly trained. You can’t get all customers on board with a new product 100%. Microsoft can’t even do that with moving users off of XP onto Vista. It just isn’t realistic. Some customers welcome change, some resist. You can’t alienate one group. I know I want the choice to run either in our server room.
Novell customers have known since OES1 came out that NetWare’s development was on the decline. It was not only an obvious scenario, Novell outright said it. You’re off base on your comment about hardware support. If you want to continue to run NCP/eDirectory/NSS volumes/iPrint and so on, hardware support it very much is a customer problem in addition to Novell’s. Beyond that, if a hardware vendor won’t make drivers available for NetWare, what can Novell do? Moreover, third party support for Linux is far and away better than NetWare, especially in the enterprise sector. If those aren’t clear reasons for a customer to move, then clearly IT isn’t critical to them.
The mail server is a service, and the other is a series of services. eDirectory is not an OS, it’s a service. iPrint is not an OS, it’s a service. iFolder is not an OS, it’s a service. SLES is the OS, NetWare is the OS, OES is the services coupled with the OS.
What do they use NetWare for then?
I don’t need any more incentive than that. If you’re using those services I mentioned above, then you want to have and OS that will run those services and be supported on current hardware. You can’t tell me that moving to OES Linux is harder or more of a hassle than ripping up your directory services and replacing it with Active Directory, replacing your entire mail system with Exchange, moving all of your printers to a wholly different management suite. RedHat doesn’t even have proper equivalents in many of these service areas. You’re incorrectly trivializing the costs of these processes. Migrations from NetWare(or from any other server OS including Windows) is not driven by a “hassle” factor. It’s driven by service needs. People move to Windows to run Exchange, NOT because Linux is too hard. If that’s all it takes to move from NetWare, I can’t honestly tell you why you(as in a customer) were even using NetWare in the first place because it couldn’t have been for much.
I can’t disagree with that. But you didn’t answer the question. Where is the benefit? Are you assuming that the open sourcing of eDirectory would automatically drive development for it? If Hula was any indication, just because you open source something, doesn’t mean there is any interest in it out there in the community.
To sum up, you don’t seem to put much value in the services that run on OES NetWare or OES Linux and in turn seem to extrapolate that general feeling onto customers as a whole. You also discount 3rd party support and hardware support. You discount the time it takes to plan a migration, secure the funds, secure the hardware, and train your staff. You also say that Novell should rip and replace NetWare and that there is no incentive to move. Well, if you want to continue to run Novell’s services, clearly OES Linux is your best choice since NetWare is going to be EOL’ed.
Neither your nor I really know if a dual kernel approach during a phase out of the NetWare kernel has any impact whatsoever on that outside of anecdotal evidence.
It’s the easiest thing in the world to say ‘no evidence’, but confusing customers over something that is pointless doesn’t exactly help – let’s put it that way.
Well, probably the customer with an installed base of NetWare servers, who need(ed) time to plan a proper migration and get their staff properly trained.
There is no way you need a choice of two kernels to do that. This is what people, and certainly Novell, don’t understand. This is a Novell problem, not a customer’s.
Some customers welcome change, some resist. You can’t alienate one group.
Novell listened too much to the people who were vocal about change. If they’d made sure that OES Linux was a much a drop-in replacement for Netware as possible, and then put the legwork into making it a serious proposition against Windows 2003, any resistance would have melted away.
I know I want the choice to run either in our server room.
Why? It gives you no benefit whatsoever, and just proliferates the numbers of systems you have and things to go wrong.
I don’t need any more incentive than that. If you’re using those services I mentioned above, then you want to have and OS that will run those services and be supported on current hardware.
So you’re going to go through the upheaval of moving to something new, that doesn’t do very much more than what you were using previously, just to run it on new hardware? I’m afraid you’re in a minority.
You can’t tell me that moving to OES Linux is harder or more of a hassle than ripping up your directory services and replacing it with Active Directory…
That transition has been happening for years, since AD is largely a pre-requisite for Windows desktops. Novell just didn’t take heed.
Migrations from NetWare(or from any other server OS including Windows) is not driven by a “hassle” factor.
Listen. Migrations from Netware have been happening for years. What Novell have done is simply speeded that process up by creating hassle that customers see as unnecessary.
It’s driven by service needs. People move to Windows to run Exchange, NOT because Linux is too hard.
Novell have made Netware too hard, that’s what it boils down to. However, the above is also true, and Novell just haven’t come up with the applications, management tools features and services to make moving from Netware to OES Linux worthwhile.
I can’t disagree with that. But you didn’t answer the question. Where is the benefit? Are you assuming that the open sourcing of eDirectory would automatically drive development for it?
The question has been answered umpteen times over. Novell moved to using Linux and open source software because they simply couldn’t afford to maintain and push forwards Netware any more, and it follows that they needed to follow through with that with the rest of their software. Successfully integrated with open source projects in the Linux and open source world would get people using Netware again – at least in part.
If Hula was any indication, just because you open source something, doesn’t mean there is any interest in it out there in the community.
Hula was a disaster because it was simply Netmail re-branded and code dumped because they had no further use for it. There was also little interest when compared with the hype because there are an awful lot of groupware solutions in the open source world such as Zimbra, eGroupware, Open Groupware and Kolab. Quite why Novell are flogging the dead horse called Groupwise one can only guess.
If they’d come up with something a bit different that others could have got interested in, that might have made things different. As it is, Hula was essentially useless to people.
To sum up, you don’t seem to put much value in the services that run on OES NetWare or OES Linux and in turn seem to extrapolate that general feeling onto customers as a whole.
Nope. I put quite a lot of value in those services, but obviously Novell doesn’t and many of their customers who are jumping ship feel the same way. Novell have made a half-baked and botched half-move from one platform to another.
You also discount 3rd party support and hardware support. You discount the time it takes to plan a migration, secure the funds, secure the hardware, and train your staff.
Unless what you’re moving to has some tangible benefit, doing any of that stuff is pointless. If I were a Netware customer I would just run Netware as-is until the hardware had fallen over, and then I’d replace it with something else. OES isn’t particularly compelling because it doesn’t fix some of the shortcomings of Netware, and that’s the way many customers are looking at it.
You forget that Netware usage has been declining for some time and people have been migrating away. There are specific reasons for that. Novell’s bottom line is evidence of that.
Well, if you want to continue to run Novell’s services, clearly OES Linux is your best choice since NetWare is going to be EOL’ed.
I think that final sentence adequately sums up what I’ve been saying. People are being herded into moving to OES Linux because Netware is heading towards the end of its life. There’s no compelling benefit to anyone in that, other than the ability to continue to run Netware services and Novell software. That isn’t much of an incentive.
And this is the real crux of our disagreement. I believe it is an incentive, you don’t. I prefer to stay away from Microsoft in the server room for most things. It’s a personal preference. I’m fully aware that NetWare to Windows migrations have been going on for years. That happens for a variety of reasons. Maybe “the boss” decides since everyone else runs windows, we should too. Maybe a new CIO comes in and only knows the Windows world and wants to make a change to what he/she knows. Maybe it’s purely political. Maybe Microsoft cut them a deal because they run Windows on the desktop. Maybe the business wants to use Exchange because it will sync well with their PDAs, or they want SharePoint. It could be a variety of reasons not wholly attributable to Novell’s know shortcomings.
You can’t take your perceptions of Novell’s services and make that fact. Here’s an example from my point of view. I have no idea what benefit anyone would have running Solaris over let’s say RHEL. Is that to say there isn’t? I don’t know that. I’m sure plenty of Solaris admins would put my head of a pike if I came out and said that continuing to run Solaris instead of RHEL has no benefits or that there are no incentives to run Solaris anymore.
I see the value in running Novell services, you don’t. Until we see eye to eye on that, I don’t think we’ll have a common ground on this.
One more thing, before this gets buried and we both forget about this and move on, I’ve enjoyed the discussion.
And this is the real crux of our disagreement. I believe it is an incentive, you don’t.
That’s absolutely fair enough. For Novell’s sake, I just hope there are enough people like you who see value there. Sadly, I just can’t see it in others.
I prefer to stay away from Microsoft in the server room for most things. It’s a personal preference.
Don’t we all? I suppose I’m just frustrated that Novell have had everything they needed to ensure that Windows stayed out – both in a technical, and a political and management sense.
I see the value in running Novell services, you don’t. Until we see eye to eye on that, I don’t think we’ll have a common ground on this.
Oh, I do see value in Netware’s services. Netware is an absolutely first-rate network OS, as are its services. I just want Novell to…………build on them.
One more thing, before this gets buried and we both forget about this and move on, I’ve enjoyed the discussion.
So have I. I would imagine many such discussions go on.