Slashdot Mirror
Tivoli Thinks About Linux
An anonymous reader wrote in to say "In this story
Tivoli refers to the enterprise management software and a business unit of
IBM. Tivoli is not only considering porting the gateway portion
of the Tivoli Framework (the gateway allows management of workstations,
windows, netware, os/2, as/400, etc.), but making it the
only gateway platform supported!!! Thus every
Tivoli implementation would require Linux. It would be important to note that
currently many flavors of Unix and WinNT are the supported gateway platforms for Tivoli Framework 3.6 and there is no support for Linux at all right now. "
As nice as it would be to have an application exclusively for Linux, I think it is a mistake to shutout other systems. Linux should be adopted because it is the BEST platform for the application, not because it is the only platform. If Linux isn't forced to compete I think that it will only be a matter of time before quality of the OS begins to suffer and then everybody will suffer.
Posted by Largo_3:
I'm a Tivoli Enterprise Consultant, and the Linux version of Framework already exists. However, it will not be released until Tivoli has a proper support division setup to support Linux. Linux will however be a supported Endpoint very soon regardless of when the Framework part comes out, also - using common sense, when Framework for Linux is released, Linux will then also be - a managed node.
nice post and all, but its hardly a rumour, I've see the Tivoli Framework for Linux myself.
Rodney Caston
Yes they do. However we are a seperate company. We can make and choose are own business decisions. While *parterning* with IBM is not gaurenteed. I don't see how an agreement with their PC devision couldn't hurt.
Well, then that could be technically classified as a "BAD" thing.
I guess the unspoken question is:
Does RedHat want to be known as the Microsoft of Linux Distributions?
I use RedHat 6.0, having only started my journey with RH 5.0. I like it, it does what I need it to do. The minute Robert Young begins to comb his thinning hair like Bill Gates, he will get a noogie.
The real problem is getting all the suits to trust the all the guys in t-shirts. Isn't there a HOW-TO on this?
Just as we don't want suits to force us to use a particular OS, we can't expect to force suits to use our OS. (It is ours, BTW, speaking of everyone.) They will use what gives them an advantage. They guage advantage in dollars. We use Linux because it gives us an advantage. We guage that in control.
Now if a suit CHOOSES to use our OS, then they are beginning to see that money isn't the only way to guage advantage. I like those businesses.
So there you are.
"Classic UFO's
I would certainly think that if a company was going to install something like this that they probably also have some other Unix boxes around too. If they have Unix admins, they have people who can work on the Linux boxes.
It certainly won't be any of these . Despite IBM talking the talk very well, I'd like to see more consistent action. I'm thrilled with all the s/w stuff they've done, but a hardware strategy would be nice, too....
(and yes, I'm well aware of VA Research, Penguin and other fine companies etc.)
IBM kicks HP's butt once again! ;)
Aww, Christ man. Please take it from someone who knows -- you DO NOT want to get involved with CA. They are worse than M$, they liberally reinterpret signed contracts and you have to sue them to make them come around (three times so far), their products are OK but not great and largely inflexible ... and CA support is a trip. And they are not cheap, not at all.
Maybe it is easier for people who expect abuse to cope with them, but I came into my position completely by accident (my father was playing golf with his buddy the CFO and was talking about his odd daughter, the one who liked computers and had finished college at 21 but liked working in menial data center positions). I was never broken down like some of the older tape operators who spent thirty years afraid about their jobs and have never quite recovered. I have never experienced stuff like the things that CA does before in my life. I admit that I am used to being listened to. My background did not predispose me to be servile (NMH and Williams, fifth gen NMH [in as much as one can be, given the history and the merger] both sides and seventh Williams both sides), to be kind (not a snob, just aware of social realities). And I have been told that I tend to speak out of turn. A lot. For most of my 27 years. Furthermore, I believe that respect and trust (such as you would accord to a vendor upon whom the health of your data center and thus the health of your Fortune 500 company depends) is very much a quid pro quo situation, my patience with some of the less savory neighbors and childhood "friends" having worn pretty thin (i.e., just because someone has the background and the education does not mean that they are nice, and often they can be pretty damned weird if their parents are worse, vide a lot of my friends at NMH and my friends at college who had gone to Exeter, Kent, and so on)(especially Deerfield -- either geniuses or insane). At this point, you have to earn respect from me. Nonetheless, all of my background predisposed me to expect that a company with great responsibility would do a good job (and having had some pretty solid relationships with IBM over AIX and MVS) because of a feeling of obligation and responsibility (especially if you were paying them a hell of a lot for the task). I was completely unprepared for CA.
I have a weird amount of leeway due to the way I got here, so despite the analyst position I can essentially work around the data center manager, and they have been absolutely delighted to let me take a piece out of CA's corporate rear end on a regular basis. In fact, they think that it is cool, because no matter how much the CA sales slimeballs complain about me to upper management while smiling at me in the halls, I really can't get fired. At this point, coping with CA is probably 40% of my entire job.
CA is a bad company, just as bad in terms of ethics as Microsoft (I believe, only having interned there, not actually worked there, and having been very unimpressed). Wait for Tivoli -- it functions the same, just with more configurability, the price is about the same, the support is dramatically better, and you actually get what you pay for, on time, at the agreed price.
Bitter, yes. But more disappointed.
Pardon my ignorance, but what exactly is it? I know it's a network management software of some sort but beyond that I'm clueless.
___
If you think big enough, you'll never have to do it.
In Austin it is pronounced:
Beer Bash
I am looking at setting up a small doctor's office with billing and I was just thinking how nice it would be to be able to use ADSM ("look -- each day, put a new tape in, pull the old one out, except when the light is blinking, OK?"). Great. As a couple of posters have pointed out, ADSM can't be too far away.
/., Medical Manager (www.medman.net) has ported their SCO setup to Linux. Works like a charm. Fast a hell. And it isn't SCO ;)
Anyway, as this came up in ask
Just though that I should make a note of this.
Sing it, sister!!!
Big corps are being challenged and unless they are willing to mutate (quickly), the small furry critters running around at their feet are going to suck the juices right out of their eggs.
Hmm, the Internet as being analogous to the Chicxulub Mass Extinction Event? Yup.
"Classic UFO's
Is that necessarily a good thing? It'd be nice to see freedom of OS choice prevail over freedom of Linux distribution choice for their application. I'm all for Linux, don't get me wrong, but I don't think forcing people into using it (or anything) is a good thing.
Don't want to turn into a you-know-whosoft, you know..
I think while this is a GREAT thing for Linux, I have to agree that it should not be the ONLY platform for implemantation. Freedom of choice is what Linux has been about and Tivoli should follow suit. Right now we run it on AIX and it kicks butt! But, it would sure be nice to have a Linux portion added. It's great to see an Enterprise level application moving in this direction.
"Klaatu, verada, necktie!" -Ash
Linux is free, remember? They can bundle Linux with Tivoli if folks want. Think of it as a very sophisticated runtime environment.
You're certainly not forcing people to buy it, so comparisons with a certain large Washington state software company are non sequitor.
-- Alastair
Not quite the same thing. There are plenty of apps that run exclusively on certain operating systems. Tivoli should have the same freedom to innovate and choose a particular OS to target and support, especially if that strategy gives them an advantage:
From the article -
"If the only gateway we supported was Linux, it would be cheap, said chief technology officer Tim Bishop. "It would make configuration testing much easier and it would be easier to support."
I hardly think that Tivoli has a monopoly on network management applications. Whereas it can be argued, indeed, is being argued, that another company has such a monopoly on operating systems.
Besides, if Linus became like Bill, we'd just give him a noogie and tell him to stop it.
John Hebert
"Classic UFO's
I really, really can't wait. Linux NetView.
...
If Tivoli is supporting the general management stuff, that means ADSM for Linux. That means that the Linux could run the ADSM and spool the data for NT, UNIXes, and so on. That means a bulletproof management platform as part of a package. Do y'all realize how many awful copies of HPhUX OpenView has sold?
That means that Linux would get serious jukebox and tape vault support, probably stuff for the big EMC coffins. Because it needs to be watched. That means that 3Com and Madge and Fore and Intel will be releasing networking tools that work with Linux and interface with the Tivoli tools.
That means that SAP with DB2 or Informix (or Oracle or Sysbase if you must) will be able to pull some serious numbers, especially with the JFS coming in one descrition or another. And all of this will interface with the management tools, and will bring in batching stuff, financial stuff, and possibly a Linux Workload Manager (ohhhhhh -- kids -- you have no idea how sweet that would be). That means more serious database tools for Linux, because the large storage would be there, with the backup, with the databases, all managed. That means SAS tools. That means clustering and more VMs.
That means that I might get a chance to work with Linux/SAP/DB2 combos backed up with ADSM on DLT libraries, terabyte storage on EMCs and SAS tools to mine it, all controlled with the Tivoli stuff reducing the 04:00 pages, and all making Linux the simple and clear choice for hard core ERP. All with a workload manager like MVS.
Guys, this is the entering wedge. This is the beginning of the end. I predict that in five years Microsoft will be 1/20th the size it is today and will be acquired by someone else for legacy service contract support. Perhaps IBM will get their revenge after all.
The next big thing will be when one of the large UNIX companies (I am betting Compaq) opens up their UNIX code to the world under GPL or close and decides to contribute heavily to Linux to cut their development costs. Then another (I am betting one of the smaller ones, like DG), then another (HP?), and then the largely NT big iron companies will start to come over (NCR, anyone).
Savor the sweet taste of cold revenge
So, does this mean they'll stop shipping their product with perl4 and use perl5 instead?
"We are not tolerant people. We prefer drastically effective solutions"
Tivoli have too many large customers that rely on that 'can't keep it up to save your life' os: NT.
---
Ryan Wilhelm
Lotus Notes Administrator
Executive Risk, Inc.
I agree that this is great for us penguin heads and the os, and it's a good stab in the side of the evil empire (M$), but what the H311 is big blau doing here. Market it as a price-point advantage to go with the Linux version at the same time continue support on other platforms.
I'm currently logged in as my redundant backup account as my primary failed over.
I think they would have difficulty trying to get everyone to use just Linux, in companies and deperatments where they currently have no Linux installations, and hence so sys-admin. This is for enterprise level stuff, so this is important.
At the moment, all they seem definate about is adding Linux support, to their current list. Migrating to Linux-only seems more like wishful thinking in the short/medium term.
(gripe mode on) If you're going to post stories like this, why couldn't you post one I submitted the other day - Sun licensing their JavaServer Pages/servlet source code to the Apache Group, and letting Apache distribute that source code under the Apache license. see the Jakarta home page and this JavaWorld article(grip mode off)
Well, this would definitely be a good thing for Linux. This would really help dent NT's appeal and could help make people see the benefits of free software.
This has been addressed before, but here goes. Other OS makers love to discredit Linux's "free" appeal because they claim that "the OS is a minimal cost in terms of the big picture." How valid a statement is that? It merely jumps the issue. Linux is free so you pay nothing for it. Is it the only cost in a computing solution? Of course not.
Big or small computing, the OS is going to factor in. In the PC market, the OS is the only major componant that seems to increase its price. How can we call OS price insignificant? On a PC, you can eliminate somewhere near $100 if you don't add Win98. On large multi-user systems, the cost of liscencing and per-user are zero under Linux. How can that be insignificant?
-Clump
Um, could you send me your e-mail address? I think we have the beginnings of a business plan here...
John Hebert
"Classic UFO's
Tivoli is a very expensive package, which usually means that it is used in large shops that have large operating budgets. I have no hard stats, but I'd be willing to bet that it is these shops that have higher beauracracy and unwillingness to commit to an "unsupported" OS for mission critical apps yet.
I'd be very happy to just see Tivoli support Linux as an endpoint / managed node. Still an interesting development in Linux's preogress to world domination!!!
It should be clear why I'm being an AC... :) ...how does one pronounce "Tivoli"? It doesn't seem to be in the NHD. I've not used it, but our company is looking for network management software, and I'm curious how to pronounce it...
THIS COMMENT IS ENTIRELY OF MY OWN OPINION AND DOES REFLECT THE VIEWS OF MY EMPLOYER, TIVOLI SYSTEMS.
If you read the acticle carefully, you'll notice that Tivoli never stated that they would support Linux Gateways as the only platform. They said it would be cost effective if they did.
Tivoli will not drop its entire OS support line for Gateways in favor of linux. Nobody in their right mind would. There is too much money to be made. However, I can see the need and usefullness of supplying pre-built gateways running linux to customers. When a new customer(or existing one) is considering the layout of a 3.6 Tivoli Implementation, machines designated as gateways are necessary for the implementation. Often new machines need to purchased. Customers always ask what kind of machine do I need to buy? Unix or NT? Sun or AIX? By having the ability to supply the customer with a machine pre-configured as a gateway, which happens to run on the linux, this is a cost-effective solution that we COULD OFFER to our customers.
However, since Tivoli Systems is a software company, not a hardware company, I don't see in the forseeable future Tivoli shipping a pre-configured linux box with the Tivoli Gateway installed. However, I can see Tivoli Systems partering with a company who would supply those machines. This partner would be responsible for the Hardware and OS support while Tivoli Systems would support the Tivoli Software.
..a little trouble with your CAPS LOCK key, dude?
Hmm... a good point. What makes it acceptable to support only one OS and not only one distribution?
Thing is, I can answer that one. The differences between distributions are sufficiently small and user-configurable (if the app says glibc 2.0 on a kernel not prior to 2.0.2 with libraries of this-and-such versions being installed, that's easy to do without changing distro). Heck, my RedHat 5.2 is practically Mandrake... If I'd gone another route it could be very similar to any other RPM-based distro.
So, should I be supported? What about someone who bought another RPM-based distro and modified it to be like Red Hat? Why even make a distinction?
These issues don't exist across multiple OSs... there are very real issues there (the nonportability of binary code and OS-specific bugs/features being among them) -- and you can't tell someone running HPUX, as you can someone running Slackware, to just use a Linux 2.2 kernel.
Could you send me an e-mail at my work address? I started at CA recently; I'll try to find out for you.
Jeff
True, not all installations require a gateway server...yet. Tivoli introduced the lcfd with 3.2 and implemented it with 3.6. This is the transition phase, soon only endpoints will be supported for monitors and TME modules (for notes, adsm, etc.). Managed nodes also take more resources on a system and increase the chances of a corba database corruption (not to mention the time it takes to check and backup the database).
Keep in mind that if Tivoli will use Linux as a gateway server, then it must be a managed node. This makes it much easier to use open source network management software and integrate it with the Tivoli product. For example, mrtg and cheops could easily send information for TEC events. IMHO, this is more important than the gateway functionality.
I posit Tivoli's large customers relying on NT are just looking for a reason to drop NT from the TME infrastructure (TMR and Gateways). I take that back...no one runs a TMR on NT (well, for more than a week, then they switch to any unix version available).
The code is there, Tivoli is just waiting on a customer who wants to use it on linux so that an announcement can be made. This is how Tivoli was built...by customers asking for features.
This is not specifically true. The biggest advantage to using Endpoints is that you can have more than 200 of them in one TMR, in addition to the smaller footprint. Looking at the big picture, very soon there won't be many things that will need Managed Nodes. There are man services that right now today in Tivoli 3.6 require that Endpoint be installed on the Managed Node.
Still and all, Tivoli rocks, Linux or no. One issue that I'd be concerned about is that while I would like my gateways (technically, gateway proxy hosts) to be appliances, we still have that PC-hardware is hard to manage remotely in the datacenter (the serial-port console issue from the other day).
It's IT management software for large high end enterprise situations... Take a look at their webpage (cited in the article).
Corporate customers pay for solutions and g couldn't care less if you ship them a network appliance and it always works. From what I gathered in the story that was the general idea -- making a Linux network appliance. This is no different than you getting Cisco router running the Cisco router OS on it. You can't put NT on one of those, either.
That's probably the reason they'd decide to drop NT.
What problems are you having with Tivoli? It looks to me like you have nothing really to say, but a one-liner's worth of FUD.
On the other hand, if we compare to proprietary Unix OSs, then the software costs are much lower. Unix software is generally considered expensive - this is often true for a single user, or small number of users, but often it becomes very cheap (per user) for an enterprise setup. For example, the RRP for Solaris workstation is about $400, though you can get it much cheaper than this - free for non-commercial usage. That's obviously more than Windows. However, getting an infinite user license for Solaris costs $1500 or something, meaning if you have 1000 users in your company, it costs $1.50 for each one for the OS. NT workstation is $250 (or something like that) per user, and you don't really get volume discounts.
Putting it on a different level... for the high-end/datacenter level, the hardware could easily cost several million. Compared to this, a couple of hundred for the cost of the OS is nothing. Besides which, the cost of the proprietary OS will be included in the cost of the hardware, effectively for free. (I believe you can get Sun hardware without the OS, which does give you a small discount - the cost of the CDs, books and other media, which is about $20 or something)
On another level... the cost of migrating can be pretty huge - first, you need to re-train your sys-admin (or chuck out the old ones, and get in some new) which can be pretty painful. Ditto if you have software developers. Then you need to re-train your staff, port your specialist software (if applies), etc etc. You also have to figure in the cost of downtime while this is all going on. This is why people are still using 'ancient' mainframes - because a) the risk of moving is too much, and b) it probably does what you need already anyway.
someone moderate this down as it's obviously just a troll
"We are not tolerant people. We prefer drastically effective solutions"
It is important to remember that they are talking about only Linux just on the middle server. This box is responsible for communication between the agents on the end-point managed machines (servers doing the actual work or workstations used by employees) and the software used by the people responsible for managing the network. In other words, Linux will be run on a box that has a very singular purpose, data collecting and playing middle man and runs only software from Tivoli. Usually no one sits at or tries to run another application on it.
For this reason, supporting one OS at this level is a decent idea. Tivoli will not have test fewer combinations of OS's used. Their management environment is already complex considering the different types of end-point boxes they can handle (workstations to high-end servers and mainframes). Now communication to these end-points will be coming from only one OS and one code tree.
To me, this is kind of like choosing Linux as an embedded OS becuase in most cases it will be serving one purpose and that is all. Nice, closed, and manageable.
Users lose little or nothing because of this. The network management people still get their NT, 98, Solaris, or whatever in front of them on their desk.
Still this approach to integrating Linux is backwards from the normal. Usually, a company would add Linux support on the periphery. In this case that would mean allowing Linux to be a managed end-point. Minimal risks. Instead, they are yanking it in and counting on it.
Before, if their customers had chosen NT to run this server and NT puked on them, the could say that it was the customer's fault because the chose NT. Tivoli has no way out here if Linux fails, except to fix the problems in Linux that are causing the failures, resulting in a better OS for us.
We do a lot of SNMP stuff at work, and I'd
really like to get a top-notch commercial
SNMP manager for Linux. Several months ago
RedHat and CA announced that Unicenter would
be available. I e-mailed both, pleading for
a beta copy or whatever. No luck.
What gives? Inquiring minds want to know.
-- cary
While it's good that Tivoli's gateway will run on Linux, it's VERY bad if it only runs on Linux. Not only is it the very thing we currently protest (programs only available on Windows or Mac), but homogenous networks are scary. If something takes down one system it can take down them all (e.g. recent viruses and worms on Windows), while on hetrogenous networks only some hosts will be vulnerable. I'd rather have some systems run Linux, some run *BSD, some run Solaris, and a few with other, less common systems (MacOS X, BeOS, etc) than all Linux. It's more of an administration hassle to manage many diffirent systems but I feel safer with a hetrognenous network.
with wake on lan NICs and the Tivoli software suite could be the ticket for gateway proxy appliances.
I wonder if a cobalt raq type solution would be better? Configure the TMR for the gateway, plug the appliance into the network and boom, you have a gateway.
IBM will start to run soon.
Step one, more to come. Time is on our side. Linux doesn't rely on quarterly income to keep succeeding, while the alternatives to Linux (except for freeBSD) do. Hmm, come to think of it, now _that_ would be a nice OS dominance war, between the open source UNIXes of Linux and freeBSD. Only good can come of that!
(remember, Lotus Notes/Domino is owned by IBM also, and they are releasing a Linux version of the full R5 server edition this year).
--- Hindsight is 20/20, but walking backwards is not the answer.
From what I am reading, this sounds like an application where they can embed Linux so deeply into the server that nobody needs to know what OS it is running or how. I envision them shipping a single CD with OS and server application (and of course source code for the Linux part of the project- for what that's worth) and it being a plug-and-play solution. They can tailor and fine tune the Linux part of the distribution and make it exactly what they want. I don't see them, in this vein, supporting any currently existing 'distribution' of Linux. It's to their advantage to roll their own. There also will be no reason for them to GPL any of their code.
I think this is not only a great step for Linux, but a great step for the Enterprise Server Industry. The middleman machine is once more going to be a fast as hell one-trick bitty box, but now it will be completely reconfigurable when no longer useful as a gateway.
Summary: Good Thing
--
(sourceCode == freeSpeech)
What if that line read:
"If the only gateway we supported was Red Hat, it would be cheap, said chief technology officer Tim Bishop. "It would make configuration testing much easier and it would be easier to support."??? (Italicized part is edited)
Comments, anyone?
Save Maine's economy: write stuff down. All comments are exclusively my own, not my employer.
This is fantastic news. I worked with Tivoli 3.2 at my previous job and was really wowed by the neat stuff they have. Of course, the first question I asked of the sales-support guy was, "Does it run on Linux?" A lot of the features and optional packages are done (often done better) in other products, but having them all use the TMF framework is incredibly handy at times, and reduces headaches.
:)
However, I'd like to point out that not ALL Tivoli installations require a gateway server; only those that deal with so-called "endpoints" in the Tivoli Lite Client Framework (LCF). The LCF was introduced in an earlier version of TME 10 (the old name for Tivoli Enterprise) and was designed originally to make Win9x machines work better with Tivoli and to allow for a smaller Tivoli footprint on end-user workstations. My understanding is that LCF is gradually becoming the preferred method of connecting Tivoli clients, but servers (both Un*x and NT) work better with the old "managed node" scheme, which doesn't require a gateway.
So if you're not using Tivoli to manage end-user PCs, a gateway is not really required.
Just a minor point of correction