Slashdot Mirror
Windows 2000 to provoke domain game
According to this article found on PC Week, Mircosoft Windows 2000 implements DDNS (Dynamic Domain Name System) in a way that makes it extremely difficult for administrators to
integrate the operating system upgrade with Unix systems, which use the older, static DNS. I would like to ask if someone here could explain what is the difference between Static DNS and Dynamic DNS, and why it's not implement almost at all unices, including Linux. I smell a fight here between Unix Admins and NT/2000 Admins in some corporates. Am I wrong?
It's okay to use an excerpt from the article, as long as you use quotation marks. Otherwise, it's just plagiarism.
You've obviously never seen a major corporation deal with a crack. No one but the admins know about it, unless enough damage was done to be obvious to everyone, or a web page got tagged. Slashdotters hear about less than 1% of corporate security problems.
you are allowed to use parts of copyrighted material in review, critique, or parody. I'd say this prolly falls under review. Besides, zdnet i'm sure is happy everytime one of their articles gets posted to slashdot, more hits, more banners loaded, more ad money.
-matt
What is it with slashdot that attracts moron posters?
I've seen a few intelligent comments here, along with a whole slew of "Help me! Help me! Microsoft is out to get me, those evil dirty bastards!"
Sheesh. Assign your root DNS to your Unix machines, and delegate the Win2K DDNS to a subdomain. It's that simple...
They can coexist. How the hell do you think the internet works except for delegation of DNS duties to thousands of different machines and DNS implementations?
The Microsoft DNS implementation is compliant with BIND 8. It may or not allow dynamically allocated Unix machines, but it most certainly responds to DNS lookups from Unix machines, and it most certainly will use a Unix machine as an authorative DNS for a different domain. I'll bet it even implements some level of security to prevent a machine from overwriting the DNS record for a server... in fact I'm going to go experiment with this right now.
Sheesh, what a bunch of maroons. This is a non issue, the article was FUD, get over it.
Pardon, Mr. NT Admin?
What's this I hear?
YOU talking about SECURITY?
Haven't been paying attention to recent news, eh? Remember the 7-second crack? Used a trojan installed on an employee's NT box to fetch their password though they were connecting via SSH.
But no, a hacker can't do any harm an NT box. Riiiight.
An NT admin would be just as wise to apply patches as we are -- there's no less need. Except Microsoft distributes patches but rarely, so you CAN'T. Personally, I'd rather have the option to spend the time to be secure. You don't even have the choice.
Honestly I don't know how you m$ employees sleep at night you must have no concience.
"While we will eventually support a standard, the IETF is having problems coming up with final draft."
Crap why do you lie? You know as well as I do that M$ has no intention of supporting the standard. You will give some lame excuse like you did with your HTML standard. Why can't you ship W2K which supports the current standards and then implement the new standards when they get approved. Read the haloween docs it is the stated intent of m$ to break standards.
RE JAVA.
It does not matter if Java is a language or platform you dolt!. You signed a contract and then violated it with malicious intent. M$ INTENDED to break java. m$ signed a contract they knew they were going to break. Read the DOJ transcripts, real the depositions before you go sprouting off on lame excuses.
M$ lies, m$ cheats, m$ steals. You my friend are an instrument of unethical people. Clean up your karma before it's too late.
War is necrophilia.
I agree with the concept of Office being the "killer ap" that drives M$. I recently was working with the CTO of one of the largest Cable companies (as in leased line) in Asia, who told me that if Office or something that could guarantee 100% compatability with M$Office (both in read/write and user interface) were available he would switch all 20,000 (yes the number is correct) of his corporate desktops to Linux immediately. This by the way is comming from a man who has linux on his desktop and notebook with Staroffice. Currently Staroffice comes close but it doesn't do what is needed. Porting of Doc and Excell files is still a little iffy. It comes close and doesn't lose too much data but it seems to usually require a cleanup, which drives secerataries and bookeepers nuts, and costs companies $$$ in lost productivity. Look at what M$ did to Lotus, Harvard Graphics, Dbase, and Multimate (Yes these were once the dominate force in corporate software) by giving the user 100% compatability with their current DB of information and archives in one package with 1 licensing fee they gained total dominance of the market. Dynamic DNS won't sell servers in and of itself. Most good sysadmins have already created a work around for this problem. What drives M$ is Office and it's lack of competition.
I'm sorry, I'm to tired to be witty at the moment so this message will have to do.
Hopefully some of the linux zealots will take your advice and do some research about NT before they start making wild claims that it crashes every 4 hours.
M$ lies, m$ cheats and m$ steals I can live with. It' the way how M$ sucks that kills me.
"Be nice, veer left, and never stop thinking" Iain Banks - Walking On Glass
Uhh, NetBSD and FreeBSD already have very useable USB drivers, and support for USB keyboards and mice. And yes Virginia, Linux isn't the only Open Source OS. That is if you consider the GPL open.
The revolution will be mocked
Correct me if I'm wrong, but hasn't BIND 8.x had this capability for some time? What is the difference in M$'s implementation? Are they "extending the specs" the way they did HTML?
---
Stephen L. Palmer
http://midearth.org
Just another BOFH.
At work (not my URL) even the NT admins were annoyed by MS' behavior, trying to ram Win2K DNS down our throats. So the NT guys--to their credit--decided to go with MetaIP from Checkpoint. The one thing I'd like to see from MetaIP would be a little less proprietary approach, then I could endorse them. They talk of a "one-time conversion" of DNS files from human-readable text to some funky proprietary format. As anyone who has administered DNS or mail will attest, you do NOT want your info to be a binary blob that you can't decipher if you start having problems. Anyway, at least Checkpoint pays lip service to standards, and since Checkpoint is an Israeli company, over time they'll be inclined to favor Linux for its technical elegance, accessibility, etc.
RFC status means nothing.
A document becomes an RFC by:
a) being written
b) being sent to the IETF
c) waiting in a 100-deep queue for some time
d) getting assigned a number
RFC-ness doesn't guarantee that it is official doctrine, only that "hey, here's the spec, get it at your local site."
There are stronger levels of IETF document for official blessings.
Dynamic dns takes automatically care of nameip mappings when ip's change due to for example dhcp. Static dns systems will need to run update scripts for this or the mappings have to be updated manually. (Or, as in the company I work for, names change when ip's change.) The whole thing could be done with a dhcp server that knows how to update the name info of bind. The friendly people at Redmond have some sort of system available for this. Note, though, this is nothing new; it's available for current NT's if you look hard enough.
I'm fairly certain my school(USC) implements this, and it honestly works fine w/ the win9x/nt dhcp clients, but all the linux clients i've tried to use have screwed it up.
Hopefully when a more complete linux dhcp client is working the problems will be solved.
Does anyone else find it ironic that someone named "HeUnique" would copy a headline word for word from another publication?
BTW, you probably want to change that before someone sues.
---
Put Hemos through English 101!
"An armed society is a polite society" -- Robert Heinlein
Linux MAPI Server!
http://www.openone.com/software/MailOne/
(Exchange Migration HOWTO coming soon)
Alright. I'm getting sick and tired of listening to all of these MS conspiracy freaks blow 'facts' out of their a$$es.
For the record, get it straight. MICROSOFT DID NOT INVENT DDNS! In my (not so) humble opinin, this is a great move! Finally we are getting rid of WINS (which was TRULY a Microsoft-only thing) and replacing it with a decent 'standard'.
Stop looking for reasons to berate Microsoft, especially when the lot of you haven't even tried to check on the facts. I have to be one of the few people here who knows what WINS was, and to realize that it deserved all of the negative feelings that DDNS is getting.
Get a life. Go read Linux-Advocacy-HOWTO. Stop being a bunch of conspiracy-driven punks.
Only on slashdot can a posting be rated "Score -1, Insightful".
Yes, this could be a risk. To address this risk, you are allowed to limit who you accept updates from (in both BIND and WIN2K DDNS).
A Win2K DHCP server can act as a proxy for its clients so that registration of both A and PTR records occurs via the DHCP server, NOT the DHCP client.
Most installations that I've seen only accept updates from the DHCP server, not the individual clients.
Dynamic DNS, if Microsoft is following the emerging RFC (yeah, right) give you the ability to automatically update your DNS tables if a machine's IP address changes. So, for example, if your machines are on DHCP, and their lease runs out and they get a new IP address, the DNS server will be updated to reflect this new address so that other clients will be able to resolve it's address.
DHCP Distribution: Version 3.0
Current Version: 3.0b1pl0
Version 3 of the ISC DHCP Distribution adds conditional behaviour, address pools with access control, and client classing. An interim implementation of dynamic DNS updates for the server only is included, but is not supported. The README file contains information about how to enable this - it is not compiled into the DHCP server by default.
Features in upcoming releases, starting with 3.1, will include the final asynchronous Dynamic DNS Support, DHCPv4 16-bit option codes, asynchronous DNS query resolution, DHCP Authentication, and support for a DHCP Interserver Protocol and live querying and update of the DHCP database. I don't see why they say it doesn't exist on UNIX. There are also perl scripts that do the job.
I'm working with DDNS both at home and at work using both Unix (Proprietary or Linux) and Win2K. They interoperate fine.
The only issues I've seen are with IXFR implementations (incremental zone transfers) and some "noise" data for some subzones. The workaround is that you can delegate the "noise" zones back over to a Win2K box until the BIND 8.2.1 code is fixed.
The REAL PROBLEM as documented in the story about Boe...oh, the "large aerospace firm" is that many large enterprises segment their IT structure along operating system lines rather than functional lines. It is much more efficient to LOSE operating system religion and use the "appropriate tool" for a job.
The DNS folks where I'm consulting use both Solaris and Win2K systems as nameservers. Solaris hosts the root namespace and the IP management tools. Win2K hosts the Active Directory Integrated delegated zones. The same folks in THE NETWORK GROUP (a functional split not an OS-centric split) manage all of these zones. There is no pissing contest over OS machismo. If more companies were to split their IT into functional areas, rather than OS empires, they might see a better result.
I'll get off my soapbox now. Just my two cents.
And how many languages do you know? The world doesn't revolve around the US and Britain. Hell, if you're going to start getting technical on spelling and grammar, you're just being pedantic. How many people do actually speak proper English anymore anyway? Learning the language is a good thing, and more than likely the person is. Just don't dock him for not being a master yet, and learn some compassion. Being a tight wad doesn't make you friends.
Blah I can't get my sig to work, it won't fit.
Let's see. This one is easy to debunk. I've made an expert system, I've made a distributed system, hell, I've even made a music composition program. All worked cross platform, without JDirect stuff. I disagree that Java sucks ass without Microsoft. Java's pretty cool. Even if I feel it's a bit slow, You can write faster Java if you pay attention to what you are doing. I'm not some grandious visionary, I just write the code to fit the specs of the people paying the checks.
Now, please mark this as Off Topic, and let's get back to DDNS.
--Rae
Blah I can't get my sig to work, it won't fit.
So I'm a corporate IT manager. I've had the misfortune to hire an NT bigot and a Unix weenie as sysadmins for their prospective domains. Both refuse to work on each other's systems, and both demand control of DNS.
They've swallowed the FUD about DDNS in this article, ignored the fact that's it's substantially a technical non-issue, and now I have both of them in my office shouting at each other, both demanding control.
What do I do?
Yep. Sack 'em both, and get two (or one?) admins who are prepared to work on both systems and do what it takes to get the job done. The company will be a better place without weenies, OS bigots, or prima donnas.
Eagles may soar, but weasels don't get sucked into jet engines.
This is very open to interpretation. Linux has more security issues, because the code is able for review by anyone.
Beside the fact that I disagree with you, perhaps the reason that NT has less "security issues" is because the code is not open for such review.
If the code was at least open for REVIEW (not development), at least a lot of unresolved bugs that are going to pop up sooner or later, and take big hits with them. At least if the code is there for review, an admin could take steps to prevent something from getting exploited, even if it doesn't actually FIX the problem.
I'm a full advocate for open source, but when security is an issue, the more you can see the better.
-Erik-
but why change it when, as the old expression goes "a picture says a thousand words"?
or, at least I _think_ I got the old expression right
Insert mind here.
Well, are you using Visual Studio 6.0, IE 5.0 and Office 2000? I believe his point was that order very much matters on installing these things because some installations rewrite system files without checking to see if it's a later version, and without prompting you to replace them. My big problem is that they even replace these files through app installs. If you're just using NT with SP5, you can get it to work just fine. Try adding in a few random other microsoft products, though.
Installations are permutable. Order matters.
Raelin
nPr vs nCr
Blah I can't get my sig to work, it won't fit.
man stat
/* Time of last file status change */
/* Times measured in seconds since */
/* 00:00:00 UTC, Jan. 1, 1970 */
time_t st_ctime;
and further down
...
st_ctime Time when file status was last changed. Changed by
the following functions: chmod(), chown(),
creat(), link(2), mknod(), pipe(), unlink(2),
utime(), and write().
Yes, creating a hard link to a file, chowning it, or chmoding it will change its ctime. creation time my eye. Oh well slashdot doesn't respect pre tags anymore, deal with the formatting.
I've finally had it: until slashdot gets article moderation, I am not coming back.
Oh please, that is such utter piffle.
Nobody can set anything up these days on NT with the click of a mouse, you need MCSEs, service packs, hotfixes, HUGE NT manuals, etc.
I added 100 IP addresses to an NT box recently and it took more than one mouse click to do it.
Chris "Ng" Jones
cmsj@tenshu.net
www.tenshu.net
"The NT kernel comes formt the UNIX family...microsoft just took the kernel developed it to there own means...in fact winnt is posix compatible..."
Huh? NTs design is influenced by a number of things, including early versions of OS/2, VMS and Mach, but it really isn't any of those things, and it certainly isn't a monolithic kernel.
The POSIX API support you mention is separate from the OS core, so, for that matter, is the Win32 API.
Why not? Could it be that metadata can't ever be associated with an inode? tch, just too bad, isn't it.
I've finally had it: until slashdot gets article moderation, I am not coming back.
"in fact winnt is posix compatible I am a unix guy but nt has its place too"
Yes, I agree. At my work I have a very nice oak desk. For fear of wrecking the wood, I use the NT4 CD as a coaster. It's very effective.
It's been working great for over 4 months. Who says NT is worthless.
My studio - www.graylands.ca
1. They are keeping up with their students, and just keeping a record of what MAC addresses are whose, that way if you do anything illegal, they can say "It was this guy".
2. They are giving you a static IP (good thing) which is the way to go. That way, you get the benifits of DHCP, and the benifits of a static IP. So, is your IP the same all the time? Or does it change?
. when in danger or in doubt, run in circles scream and shout --Robert Heinlein
First posting sucks,..
Let me be the FIRST to announce,..
"LAST POST!"
Ok,. now no one else make comments please...
My studio - www.graylands.ca
In a very very very long while.
How we know is more important than what we know.
MCP Magazine published a similar piece in the 9/99 issue. For various reasons, they claimed that NT's DNS as superior to Unix. However, the uneducated author of that piece was really contrasting BIND 4 and BIND 8. I sent MCPMag a nasty-gram for that. Linux has had a BIND 8 compliant DNS out for a couple of years already. Unfortunately, many of the major Unices don't have a BIND 8 DNS out. AIX 4.3 being one exception to that. IBM was fairly aggressive in keeping up with the latest RFCs governing DNS/DDNS and DHCP. The key difference between static DNS and DDNS is that DDNS allows zone files to be updated via special nsupdate packets. Update packets are sent to the DNS by either a DHCP/BOOTP server or by each node. No current MS OS supports this latter type of DDNS registration, but their are 3rd party tools to make it happen. One bummer about Linux is that its DHCP server does not yet support DDNS updates. Here's one area where NT's DNS/DDNS really stinks: you are forced to use the MMC GUI tool to admin it. Yucko! IMNSHO, vi & perl are the ONLY tools for DNS. =)
accountability is about saying "you know.. if I post this stupid comment someone might form an opinion about me" and then maybe reconsidering.
How we know is more important than what we know.
Cool, Astroturf has reached Slashdot...
I don't know what experience you have on UNIX boxen, but I've used both UNIX workstations and NT workstations, and I can tell you you are full of shit. NT is a productivity destroyer, as the Windows interface just isn't designed to get work done. It may have been designed not to scare Joe Blow, with the dancing paperclips and the flying sheets of paper, but it certainly hasn't been designed to let people do what they want to do.
Hell, even the bloody Macintosh is better in that respect, because at least it has a good graphical interface. Windows is just an ugly, unholy mess built on top of an unstable kernel.
"Be nice, veer left, and never stop thinking" Iain Banks - Walking On Glass
I like Linux because of the plethora of modifications I can make to it, and the amount of customization I can make to the UI. I also really appreciate the online documentation, which is in a sane and easy to use format. Once I figured out how to use it, I fell in love.
So, what computing paradigm to you love? I probably haven't tried it, but if you would help me get it installed, I'll be happy to give it a try.
. when in danger or in doubt, run in circles scream and shout --Robert Heinlein
I am running Office 97, VS 6.0 and IE 5 ( it is my developent machine - server cause I need to test stuff with MS SQL 6.5 - which by the wy is installed too - and all of this in the IBM thinkpad -128 MB ram ) So far works perfectly ...
And also Linux on another partition - works fine too with exception that video drivers were so painfully slow I had to get AcceleratedX...
Practice what you preach there, pal. I've got no idea what you are talking about and neither do you it seems. NT4/2k stores the DNS data in a binary format, have fun admining that with edit.com.
>>Are unix people all anti-capitalists?
you know that is obviously false. i think his point was that microsoft is often associated with having fabulous amounts of money, perhaps more than people should be giving them. (oh god, i'm just setting myself up for an anti-trust debate, aren't i?)
>>Micros~1 - Hello... I've used win95/98/nt and I've *never* had to type in an 8.3 file name.
if you use any of your older 16-bit apps in windows, you will encounter these silly abbreviations (eg pictur~1.jpg) all the time. however, i usually run into these things in DOS, which turns long filenames into a nightmare, not a blessing.
>>I think linux geeks are just bitter about MS's dominance.
the complaints about ~1 filenames have nothing to do with MS's dominance. just because you've never had to struggle against its ugly side does not mean it's not a bad system. people complain because it sucks.
>> Macrostupid - Surely the people who were fooled into running these macros are to blame.
the people that passed on that virus are mostly newbies, and don't know how to wield the power of macros anyway! unfortunately, there are millions of newbies out there, and microsoft pushes their products into their hands more feverishly than any other demographic. if you don't believe me, i'm sure the talking paperclip can convince you otherwise.
>>I get 15+ days uptime all the time on my machine.
well cool, as long as we're sharing our experiences i might as well give mine! i've got win98 on a top of the line dell machine that i'm using right now. this whole summer it has not had an uptime of more than 5 days. hell, i don't even push this machine! i use it to browse slashdot and chat mostly. what causes it to crash after 5 days for no good reason? beats me. buggy coding i guess. oh sure, it made it to 8 days *once* but who wants to use a computer that has 194 meg of allocated memory with no apps running?
and this is windows *98*, supposedly fixing those truckloads of bugs. when i had win95 on a diff computer (that i sold, hehe) i couldn't even make it through the *day* without it crashing. usually crashed 3 times a day. are you telling me this is 'acceptable'?
"...The DEFAULT authentication for Win2K is kerberos."
There I was addressing the current roaming user, using a non-Microsoft platform. I have a problem
with DDNS in general, not just Microsoft's.
I don't think dynamic DNS solves the roaming
machine problem because of the TTL and security
issues. The problem it does help solve is plug-
and-play -- you can fire up 100 w2k boxes on a
network using a promiscuous DHCP server or even
the client-autoconfigured range and they will
all get registered in the DNS based on the
network settings on the *client* end. Just like
the Macintosh Chooser. How useful that is depends
on what you plan to do with the information and
how scalable it needs to be. In our environment,
out-of-band end-user access through a secure web
server is better.
To answer your question, as far as I've seen, the
w2k DDNS client does not do kerberos or any other
form of strong auth.
Internet Explorer does not and for the forseeable future will not do kerberos (according to the lead
NT5 security engineer when I was in Redmond). The
NT5/w2k version does some SID token-passing with
proprietary headers, not SASL. You can't really
fault MS for this, though, because there are no
standards for kerberos over http. CMU keeps trying
to get people to kerberise web applications, but
Stanford gave up and went s/ident and even MIT
makes x509 client certs for users rather than
force kerberos on an unwilling application.
File & print service does do kerberos and can be
configured to refuse to negotiate non-kerberised
connections if you know that all clients and all
servers on your network are guaranteed to be
running w2k. In the real world, I expect WINS and
legacy NT domains to last through 2010. We still
have 2 key production NT 3.51 servers because the
commercial off-the-shelf application they run is
not reliable under NT 4.0. There are more
architectural differences between NT4 and w2k than
between NT3 and NT4.
> *last I looked the RFC wasn't final yet, MS has
> been updating the W2K code to follow the RFC.
> drafts.
Or vice versa.
Assuming you configure BIND 8.2 to accept DDNS
requests with no authentication whatsoever.
If you do that, you deserve what you get.
Yeah, and did you notice that msvcrt.dll changed in between VC 6.0 and VC 6.0 SP2? That was to fix
bugs. On my machine, installing VC 6.0 broke at
least 2 non-MS applications. Go take a look at
bug reports.
Man, I had to take care of a dozen NT boxes
loaded with development tools. I know more about ways to destroy this systems by a wrong sequence in applying patches, fixes and errata than I ever wanted. Our Linux boxes are order of magnitude easier to maintain. And I am no UNIX fan. I like goog GUI and IDE's. It just a fact that
UNIX style is much more stable for development use.
Original poster insisted that MS enviroment is stable. I think he is full of shit.
AcceleratedX rocks, BTW...
<^>_<(ô ô)>_<^>
I remember reading somewhere that although ipv6 is 128 bit addresses (iirc), only the last 64 bits are used for actuall address... thr rest are country, state, etc... (not sure how it's split up in this part) but you automacically keep your ip and routers can figure out from the first part of the addy if they've got to drop packets to your backyard or to china... I remember that it promised to make routing much simpler, iirc...
Always the chance I could be wrong, of course.
This sig left intentionally blank.
Ya I figured this might happen. Sentence and word optimizations can confuse people.
I was talking about the concept of DDNS vs SDNS.
The concept behind DDNS is that a device should always have the same name but the IP can change.
The concept behind SDNS is that a device should always have the same IP but the name can change.
That is what I meant by bindings.
ayottesoftware.com
If you need a GUI to do productive work, I feel sorry for you. I rarely use the GUI functions of windoze, and even then I use them only when a CLI solution doesn't readily exist. Point and click is for dry firing your gun, not for using your computer.
Just because you're paranoid doesn't mean they aren't out to get you
I just sat in on a several hour AD presentation done by MS. You can tie the DDNS in Win2k to Bind 8.x servers.
There's another way of doing that of course, but it's evil as hell. :) :)
If you don't have any REAL subnets / LANs then you can tell the ridges that all conference rooms are really in all the vLANs of the organisations, and stuff will just magically work. I suspect the behind-the-scenes cost in data traffic is horrific, but I don't care
This also lets you grab a server, complete with UPS, and run over to another building with it, and hardly anyone notices
Or so I'm told (Do you read this stuff Tim?)
I think dynamic DNS is a solution in search of a
E XT=935866614.1926103089&hitnum=9
problem.
You say, "Its nice to be able to connect w/ a laptop anywhere on a 100+ subnet network and get the same domain name to resolve everytime."
Why?
How many people besides you regularly connect to
a server running on your laptop?
Are you sure you control the TTL on your DNS
server, every DNS server used by every client
that talks to you, and every server you talk to?
What do you do when a remote site's TCP wrappers
refuse access because they cached your old PTR
record?
What assurances do you have that someone can't
spoof your dynamic name and steal credentials? If
you think you're authenticated by MAC address,
try ifconfig eth0 hw de:ad:be:ef:01:23 (doesn't
work with all enet cards, but does with the common
ones). If you use kerberos, x509, or ssh host keys
and you actually bother to verify them, then you
have less of a problem, but many common services,
like unencrypted web pages, have no end-to-end
server verification protocol. Interestingly
enough, Microsoft's NT domain protocols do not
strongly authenticate the server to the client.
If an attacker puts himself at the server's IP
address and generates a nonrandom nonce, you lose.
Microsoft considered strongly authenticating
DDNS to be too hard (and nonexportable), so they
basically trust whatever you put in the Network
Control Panel (or a packet manufactured with
smblib) as long as the name has not already been
taken. Taken names can probably be freed up with
the same sort of games people play to take over
IRC channels. Bzzt! Game Over.
Microsoft says it plans to get rid of WINS, but
the initial implementation brings all the
instability and insecurity of WINS to DNS. No
thanks. The non-Microsoft solutions tend not to
be much better at this time.
Out-of-band authentication like MyIP or the old
ml.org web page works, but that ain't DDNS,
that's end-user access to static DNS... which
can be a good thing. We provide something similar
for our students.
In case deja URLs aren't permanent, search for "WINS" in comp.os.ms-wendows.networking.win95 during January 1996.
http://x22.deja.com/getdoc.xp?AN=135549278&CONT
whatever man...
politics is politics...
posix is posix...
and personally i am not surprised that you feel
way and i am not surprised MS has taken another
step...it's just unfortunate..oh well...
>If anyone deserves bashing here, it is PCWeek and/or this particular >reporter,not MS.
Get real. You don't actually think that MS didn't approve of this article before PCWeek released it,do you?
You MS PR flacks are really quite stupid,you know.
How come all of these MS praising posters are all Anonymous Cowards?? As for the comment that "[Anyone making the] assumption that a Sys Admin that runs MS products is ignorant, is nothing more than tunnel-vision and narrow mindedness", the plain hard facts are that anyone voluntarily using an MS Product for mission critical (otherwise phased as 'important') server applications is a little daft as the tendancy for MS products (in the vast majority) to be: 1. Crashware 2. Bloated 3. Slow/Inefficient 4. Insecure is notorious. While not every MS System Admin does so of their own free will, I would have to agree that anyone that claims that NT is a better solution to everything else is being a little ignorant of the facts. (For actual references, just refer to the many past slashdot articles and posts on similar subjects. This topic is getting old... I guess MS marketing really does get to some people...
This feature of ISC DHCPD and the DDNS features
of w2k are totally unrelated.
The ISC DHCPD 3.1 feature referenced, and the
patches to 2.0 which have been around for over a
year, does this:
When a Windows 95/98/NT client, or a UNIX or any
other client configured to send option 12, is
assigned an IP address, the ISC DHCP server
connects to the DNS server on the client's
behalf to update its entry.
This allows you to secure your DNS server to
accept (possibly DNSSec'd) updates from your
DHCP server only.
The Microsoft DDNS solution does this:
After a Windows 2000 client has been assigned an
address by the DHCP server, it contacts the DNS
server directly to update its entry.
The Microsoft solution requires your DNS server to
accept updates directly from your clients. The
Microsoft solution does not attempt to support
Win95/98/NT clients at all.
As for the comment that
"[Anyone making the] assumption that a Sys Admin that runs MS products is ignorant, is nothing more than tunnel-vision and narrow mindedness",
the plain hard facts are that anyone voluntarily using an MS Product for mission critical (otherwise phased as 'important') server applications is a little daft as the tendancy for MS products (in the vast majority) to be:
1. Crashware 2. Bloated 3. Slow/Inefficient 4. Insecure
is notorious. While not every MS System Admin does so of their own free will, I would have to agree that anyone that claims that NT is a better solution to everything else is being a little ignorant of the facts.
(For actual references, just refer to the many past slashdot articles and posts on similar subjects. This topic is getting old...
I guess MS marketing really does get to some people...
As for the comment that
"[Anyone making the] assumption that a Sys Admin that runs MS products is ignorant, is nothing more than tunnel-vision and narrow mindedness",
the plain hard facts are that anyone voluntarily using an MS Product for mission critical (otherwise phased as 'important') server applications is a little daft as the tendancy for MS products (in the vast majority) to be:
1. Crashware 2. Bloated 3. Slow/Inefficient 4. Insecure
is notorious. While not every MS System Admin chooses their server platform of their own free will, I would have to agree that anyone that claims that NT is a better solution to everything else is being a little ignorant of the facts.
(For actual references, just refer to the many past slashdot articles and posts on similar subjects. This topic is getting old...
I guess MS marketing really does get to some, dare I say it, *ignorant* people...
"Khttpd is not in the kernel."
Its in 2.3.15
Paul Laufer
This is my current project, so here is my take on what micros~1 is doing.
:-) At the end is an opinion, which is not a clue, but can be ignored or countered as you see fit.
/etc/hostname.
/etc/hostname in option 61 of the DHCPREQUEST, and get that code into every major package out there. Then the FUDders will not be able to do any more than superficial damage.
First, some background as to what Dynamic DNS truly is, because its obvious most of the slashdotters are posting without a clue. Here's a clue, and its free, as in free software
What is Dynamin DNS?
DynDNS is result of putting together several RFC documented techniques in a quite nifty way. Start with DNS [rfc1034 & 1035], add DHCP [1531, 1532, 1533, 1534] and tie the two together with Incremental Zone Transfers and Notify [rfc 1995 & 1996], and call it DynDNS [rfc 2136 & 2137].
Read rfcs 1995 & 1996 for a discussion on why full zone transfers [AXFR] are a bad thing (for bandwidth consumption), and see the elegant solution proposed with the incremental zone transfer [IXFR] extension. This is the basis for updating a primary name server with a new RR containing the hostname & IP pair (and IP->hostname reverse pair). You can also use this mechanism to remove a RR when the host is no longer associated with that address. There is also a discussion of security so that only pre-programmed IP addresses can do IXFRs, and allows extensions for fully authenticated updates when someone gets around to writing the code someday.
Read rfc 2132 to understand how a DHCP client does a DHCPREQUEST to a dhcp server, and how it can pass its hostname inside of option 61, client identifier. This is what win9x currently does with its client code, but only a patched version of some dhcp clients for linux do this.
Now, to put it all together.
A machine [win or linux] with a dhcp client boots up, broadcasts a bootp request (the transport mechanism for dhcp) with a DHCPDISCOVER message. A dhcp server on the network responds with its local address in a broadcast (because the client has no IP address at this point, all traffic must be broadcasts), and then the client broadcasts a DHCPREQUEST to that specific server. Contained in the REQUEST packet is option 61, containing the hostname of the machine. In win9x, this is what is entered in the network control panel "computer name" field, in *nix it the contents of
Then there is a whole bunch of communication between the dhcp server and client so they both agree on things (go read the rfcs, or sniff some packets off the wire, or both) with the end result the dhcp server now has given the client a lease on an IP address for a certain amount of time.
Now comes the DynDNS bit.
The dhcp server now communicates to the primary name server with an IXFR message, sending a RR containing an A record (and a PTR to the reverse DNS server) with the any and all information that might be contained in a RR, and the TTL is set to one half of the lease time given to the client. If the name and IP address are not currently in the DNS database, they are added. If they already exist, the IXFR message is refused, and the DHCP server must change the name to something unique. This is one mechanism to prevent overwriting your important servers addresses with bogus info.
What micros~1 is doing.
From what I can tell from some presentations I have seen, and playing with win2k beta, they have tied their DynDNS into ActiveDirectory as an attempt to shut out the *nix/OSS implementations until they get a foothold in the corporate door. I can't tell exactly what they are doing until I get a lab testbed set up and see if they interact correctly with BIND 8.2.1 or other rfc2136 compliant systems (someone mentioned cisco's registrar product, its real nice, and real expensive, and not based on any bind code). There is something going on with rfc 2052 defining directory servers on the internet, but I only read enough of it to give me a headache.
Static vs. Dynamic
M$ strategy is to put all IP addresses into AD, making the entire network a big, dynamic mess. As a network guy, I want all the important services to have static IP addresses. This means servers, DNS machines, router ports, mail servers, and anything else that should be stable.
M$ considers servers to be unstable (based on BSoDs and regular reboots), so they want the IP addresses to be dynamic. That's a bad way of thinking.
The article in ZD is actually correct on a lot of things. There are already battles going on between the ultra-reliable thinking *nix admins and the reboots-are-good ninnies who have realised they can't make M$s win2k work in a unix based world.
The only solution is for the OSS community to make a standard implementation of dhcp client, one that by default passes
the AC
Hemos is like...sci-fi fans;he thinks technology is cool, but he hasn't bothered to understand the science it's based on
Good user interfaces can improve productivity. The Windows user interface is just about the most useless I've seen in my life. The SGI UI is the best I've used so far, followed by the Mac. KDE suffers from trying to look like Windows.
"Be nice, veer left, and never stop thinking" Iain Banks - Walking On Glass
Rob - You may want to change the M$ icon. I think the troll kiddies get aroused whenever they see it.
DDNS is indeed implemented in the Unices - w/o a problem. The current version of Bind (8) supports DDNS and the development version of DHCP supports the DDNS updates.
The difference in the two (Dynamic/Static) is that, as everyone knows, static DNS requires you to know the IP address of the domain name you're recording. In DDNS, the client requests an IP address from a DHCP server, then, as long as the DHCP server is configured to 'know' the client, it recognizes which client is requesting the IP (based on MAC addressing) and informs the DNS server that it is giving a certain IP address to a client for a particular domain name, and the DNS server accepts the information and adjusts its lookup tables accordingly.
I've implemented this in Linux w/o a problem whatsoever - and I know of a school that has implemented it in a Solaris environment.
Its been out there for a LONG time, btw - by that I mean at least 3 yrs. It wasn't pretty, at times, 3 yrs ago - but it was there. Now, it is a very well integrated solution.
Its nice to be able to connect w/ a laptop anywhere on a 100+ subnet network and get the same domain name to resolve everytime :).
Btw - first? :-)
Brice
probably more, since installing NT doesn't automatically make you a sysadmin like installing Linux does (anyone know how to get fetchmail working over ppp with Debian slink?)
Just meant to say that MS has added a useful feature and Linux should do the same.
The situation I described with their MCSE's has to be addressed if Linux is to keep their gains in the server market...
Jim in Tokyo
-- My Weblog.
It's Embrace, Extend and Extinguish all over again, but with a substantially different tactic.
Last I checked, DDNS was already a set standard, albeit a very new one that most Unices don't use yet. So there's nothing inherently evil about including that in Win2000. But, M$ is breaking interoperability with Unix servers to do so, due to the poor design decision of making a lot of their stuff (although with "Active" in its name, you can tell it's going to be insecure/unstable/buggy/all-of-the-above) depend on a standard which isn't mainstream yet, even if it is probably an open one.
Very clever, I must admit. A way to twist Open-Source to their advantage. Nonetheless, I'd say this ought to go into the 2.3 development tree now, so that it'll hopefully be ready before Win2k or at least not long after.
Somehow, I don't think that Ziff-Davis if going to take Slashdot to court over something as simple as the title of an article. 'Sides, they'd fire a warning shot first. (Nastygram from the lawyer.) Even more so, why flip off Slashdot when it brings them so much revenue? Ziff loves Maldo, and so does Davis. (But I don't think they're having a three-way.)
if that's your philosophy, why don't you just change your minimum threshold to be 1. That way you won't see stoopid posts by AC's. I prefer to read the AC postings because every once and a while an AC had something useful to say.
-earl
someone mentions 'Ignorant MS-using sysadmins' and everyone assumes they were the subject...
i would read it as referring to the subset of MS-using sysadmins who are ignorant, not as labeling the entire group as ignorant...
Oh, please Bill! Let me work there!
Honestly, can you really think of another company that has enough power to even think of doing what the Halloween documents suggested. Remember, might != right. Being able to force your customers to buy something does not make a good long-term business plan. Eventually they come after you with pitchforks.
BIND 8 has this, and has for a while. Copying standards and then making them incompatible with the rest of the world is the only way microsoft pushes the envelope.
thank you.
wow, that really beats my linux box that's been up for ~100 days since that power outage...
get your dirty sig off me, you filthy APE!
I'm not sure what the whole issue is here - ISC's BIND supports dynamic updates now. And their DHCP client supports sending the hostname as part of the packet.
In fact, if you look at this link, you'll see that I currently use a perl program to take entries out of my DHCPD lease file, and update my DNS with the new hostnames, DYNAMICALLY!
- Kazin
For those of you who don't know, DDNS is basically DHCP with hostnames built in.
Problem is.. it violates the real standards for DNS.
To do DDNS requires that all upstream servers update excessively; AXFR's are performed on average every *FIVE MINUTES* in DDNS from what I've seen.
Problem #2; Microsoft doesn't even know what an AXFR is. NT DNS follows standards for lookups, but if you need a secondary DNS server and your primary is NT, well, break out the checkbook. M$ DNS follows ZERO standards in zone transfers, not to mention file format! You *CAN'T* secondary with unix without more headaches than it's worth.
DDNS is nothing more than another Microsoft attempt to gain more control over the internet through 'evolving' standards by blatantly ignoring them.
I pity the fools who believe the hype.
-RISCy Business | Rabid System Administrator and BOFH
your company here.
shelby != ford
Its all a buncha BS. it is 100% possible imho to have a unix box and an NT4SVR box competing for the same "highest" uptime. our WWW (AIX) and DNS (NT) boxes never go down. the real problem here is the stupid NOVELL using sysadmins pulling the power cable on the boxes to restart them because they dont know how to restart a service. just use WHAT WORKS FOR YOU.
hmm, someone forgot to get a babysitter for the script kids tonight..
/* Jesus loves you, but everybody else thinks you're a dork. */
I am only guessing, but I imagine that W2K integrates DNS and DHCP in such a manner that a UNIX DNS server cannot be used (embrace and extend, anyone?). A FQDN would be assigned to a host regardless of its IP address; the DNS server would update the host's IP address dynamically whenever it changes.
There is nothing that says that you need dynamic DNS in order to associate a FQDN to a specific workstation in a DHCP environment. With DHCP, you can reserve an IP address for a specific workstation simply by giving it the workstation's ethernet address. I set up a bunch of X terminals like this at my previous job. Works great. Less filling.
As a rule of thumb, servers (i.e. hosts that need to be accessed via a specific FQDN) ought to have a static IP address anyway, and it is unwise to create dependencies like this (for example, NIS server needs DHCP server in order to boot).
In my opinion, Dynamic DNS is nifty, but if Microsoft is not keep the standard open, then it is useless.
> The Mailing List and Newsgroup just aren't an adequate security model.
If you don't trust a patch floating around a mailing list/newsgroup, fine. They will eventually get looked at by the (trusted) maintainers, who will personally review the patch and likely include it in the standard distribution. It's not as if joe schmoe can magically write some code, post it on a newsgroup, and *bam*, it's in the distro. It doesn't work like that. Code has to go through an EXTENSIVE public review process before it gets merged into the main tree. That's a more than adequate security model, and better than most proprietary software vendors.
If getting patches from an untrusted source in a newsgroup bothers you, then you can wait for them to get reviewed and either be rejected (and the functionality added in some other way), or make their way into the standard distribution. I don't see what's so hard about that.
You obviously haven't actually had any direct experience with the way these projects work.
Berlin-- http://www.berlin-consortium.org
DNA just wants to be free...
I've been lurking for a bit and watching the discussion. Perhaps some of you could tell me what the MS DDNS means for the following implementation:
We're currently installing an Oracle workflow system that relies on LDAP to grab user information from our e-mail server to populate the workflow system directory. The Oracle system is hosted on a Unix box, but most of the user information comes from our e-mail servers, which are all MS Exchange. We also use NetWare.
If the directory services in Win2k are all one-way into the MS directory and we migrate to Win2k, will it prevent our Oracle WF system from pulling user data from the DDNS to populate its own LDAP directory?
Thanks in advance. And if I've phrased the question incorrectly (or cluelessly), please give me a clue.
(Pulling on reflective armor and awaiting response to my first-ever Slashdot post!)
TLR
A man no more knows his destiny than a tea leaf knows the history of the East India Company
WINS is limited to M$ machines. Non-M$ machines will not appear. So reverse WINS != DDNS.
As far as DDNS' usability goes. Keep in mind that DDNS is an option in Win2k - not a requirement. On my win2k advanced server running Release Candidate 1 of the beta code, you can still choose to run the regular DNS service. The advanatages of DDNS will likely make administrators want to move to it.
The1Genius - Littera Scripta Manet
Cisco has had a DNS/DHCP server implementation for a long time
The new "Network Registrar" comes in Solaris and NT (4) flavours.
It has a policy based DHCP server and dynamic DNS functionality.
It basically allows the administrator to map a hostname to a MAC address, independent of the IP Address (hence the need for dynamicaly updating the hostname's IP). I can see only a few instances where this might be required. For 95% of the cases, why not just put a static IP in the DHCP server for a particular host. Saves lots of problems.
I guess Microsoft not only invented the internet, they pioneered DDNS with Windows 2000 as well! (*cough*)
--
Let's not all suck at the same time please
Let's not all suck at the same time please
Why would you do this? Just randomly change the name associated with everyone's static IPs to confuse them? Part of the point of static IP addresses is that you can then use static DNS (specifically, a name-to-ip database without capability for easy rebinding) and still keep the same name corresponding to your device.
Basically, you just said the exact same thing with slightly different terminology, without making any more sense than you did the first time.
Unless you're using some odd definition of DNS, the IP address and name are always bound together. The new idea in dynamic DNS is that this binding can be easily and quickly changed, the purpose being that the changes in IP->name binding will match those in device->IP binding, giving an apparent binding of device->name even though it's really just carefully managed co-incidence.
-----
--
perl -e'$_=shift;die eval' '"$^X $0\047\$_=shift;die eval\047 \047$_\047"' at -e line 1.
If anyone is interested in actually reading them, the RFCs MS is SUPPOSED to be following with this are 2136 and 2052.
Also, no one I know who is testing this out (in the IT consulting firm who will be doing a great deal of this whem it spills out upon the world) is fooling themselves about what a GIANT political battle this could turn into. To avoid this, you will probably see Active Directory Domains handling their own DDNS, and forwarding to existing UNIX infrastructure for all other name resolution if those doing the implementation aren't up to the fight.
...how other systems in the network will resolve to systems in the DDNS zones is supposed to be worked out, (with the use of some crazy zone magic) but I've not seen it work yet.
"I don't think I ain't" -Thompson's Corollary to Descartes
My 2.5 lb Sony Vaio does firewire quite nicely - as we speak, it is pulling video off my Digital Video camera.
Snooty people annoy me.
-- My Weblog.
really now, /can/ integrate, with some work. Im sure once I start playing with Win2k, it'll integrate too. If anything, using DNS lookups for name resolution, intead of WINS should make cross-platform stuff a little easier.
If I were a less secure person I would probably take offense to being called a 'Ignorant MS-using sysadmin'. As it is, I happen to be an MCSE who sysadmins roughly 300 NT servers in a network with over 20,000 nodes and almost 40,000 users. I use Linux for lots of network-odd-jobs, and NT and Unix
To use Windows 2000, a site also will likely have to use DDNS because it is extremely difficult to integrate the operating system and its Active Directory into a network using static DNS.
"I haven't seen anyone try to deploy Windows 2000 without DDNS because that would probably mean trying to deploy Windows 2000 without Active Directory," said John Kretz, a systems integrator at Enlightened Point Consulting Group, in Phoenix.
The problem is that *to use Active Directory*, you appear to have to use NT for DNS. Since Active Directory is a key selling point, and most admins think it is a Good Thing, they want to use it.
In theory, I agree - Novell's networking metaphor worked well for all size shops, large and small, and spread out to WANs respectably well. Now MS seems to be attempting to build some of those features into it's NOS. It's an improvment over their old, horrid system, where you had a nightmare with real world access and information issues.
(obHelp: if you're a person with *nix experience admining NT, get the O'Reilly book that covers the command line versions of their admin utilities, and combine with Perl. With early 4.0, it adds a great deal of flexibility to their point-and-cry interfaces.)
"$30 for the One True Ring. $10 each additional ring!" -- JRR "Bob" Tolkien
Here's a few pieces of information for you, that I got at a briefing by Microsoft's marketdroids a few months ago. Take the following with a grain of salt, since any technical data was filtered through a non-technical marketing person, and then my memory (which isn't the most reliable storage):
According to the briefing, ActiveDirectory uses DNS "like a database" to resolve the names of any Active Directory objects. Dynamic DNS is required, since any changes to the ActiveDirectory name space require updates to the DNS data. We were told that ActiveDirectory wouldn't work without it, and the implication was that Active Directory stored more than just machine names and IP addresses in there.
I asked why use DNS for this, when LDAP (for example) already exists specifically for managing directory information. I got an oblique answer that Active Directory can import information from an LDAP server, but can't update one. Probably the marketdroid didn't understand my question, or know the answer.
I also asked if a RFC-compliant non-Windows Dynamic DNS implementation would work with Win2k and Active Directory, and was told that it would, but that Microsoft very strongly recommended using only their DDNS (anything else is unsupported).
My overall impression is that they're following the RFC's for DDNS, but filling the DNS database up with a lot of records that don't have anything to do with machine-name to IP-address mapping.
Well, in face of the emerging IPv6, I can't stop wondering what the hell can DDNS or even DHCP useful for in the coming times.
Let's see, a common saying is that with IPv6 every one could have a network the size of the current internet withouth exausting the address range, so one can say that it is possible for even ppp conections (usually associated with a user/client) to have a fixed ip address, or even every student in an university.
From a superficial point of view, DDNS and DHCP seem like DOA to me, with the exception of free connections without a fixed login, but even that can be arranged by usage of the CID.
Does anyone agree with me?
Some of the early articles I read suggested that MS DDNS was nothing more than Bill squishing WINS into their DNS services.
To err is human, to blame it on computers is even more human.
Actually, it's much more to do with VAX. One of DEC's head developers joined MS to work on OS/2 not long before M$ dropped it...
(ceci n'est pas un
"...This has been available in the fubar distro since 1978..." "I hacked together something that will do this for some time using a perl script and DR-Dos..." "...M$ is evil, Why doesn't the DoJ just shut them down?"
h tm )
Quit whining. There is too much work to be done. Add the features while maintaining backward compatability and this will cease to be an issue.
Something like DDNS sounds like a good thing. I don't care who implements it. Now when I say 'Implement' - I mean that if you take the setup disks for your favorite OS, install the OS using the defaults, it's set up for you. MS has simply chosen to do this. If Redhat has a brain, they will do the same. It should work without a second thought.
When I dial my ISP, which is how most people encounter dynamic IP's, I have set my Win98 DNS settings to have my machine be sparky.gol.com.
As I see it, I have dialed in and told them my username, so they know that I am sitting on a particular IP address. That address should have my hostname appended to it for the duration of my session.
It is the fault of the software that my friends in the US can't find me by contacting sparky.gol.com. (I use ICQ so people can find my machine. It is the fault of the software that people can't do a simple "finger" command, unless I happen to be logged into a shell somewhere.
I am not a SysAdmin. I am a windows programmer. Like many people, I ran a linux system for fun when I had the space for a second machine.
People are looking at RedHat, because it is the most mainstream of the Linuces in the public's awareness. Now that RedHat is more of a commercial venture, they will need to address things like this, right out of the box. They are making strides because they are a usable server for windows clients. They will lose this market to Win2K if they are not as full-featured in this respect. Sorry, that's how it is. I would love to see linux on my computer at work. I wish I could just use StarOffice and forget about MSOffice. But it's not going to happen this year. I have work to do...
MS has made an unspoken promise to thousands of people around the country - Become a MSCE and you will have a pretty good job. They don't care if you were selling stereos last week. They don't care if you went to college and spent four years playing RPG's on a vax. They 'tell' you that if you get certified, you will be able to afford a couple of kids and a decent car. Every couple of months, you will get a couple of CD's in the mail and you install a service pack. You'll get beta's of new software. Take a day or two and play around with the new software instead of working. As I see it, they are keeping their promise. Are these people going to give that up because the software doesn't comply with some RFC sitting in an obscure URL in Switzerland somewhere? Do these guys know what an RFC is? No. Now picture this: Joe MSCE in Maryland gets a call from his boss who is on travel in Atlanta. The boss can't print from his hotel. Joe tells the boss to connect to the internet and open NetMeeting. Joe uses NetMeeting to call the boss at bigboss.company.com and they share the document. Joe shows the boss how to print. Joe looks like a genius. Joe does not lose is job for "not keeping the printers running," at least in the boss's mind... Get the picture? Now realize that perhaps neither of them know that DDNS made this do-able. (*You* try to get your non-techie boss to run WinIPCFG and tell you his IP. Ten minute phone call at least...) This is how it works in the real world. These guys don't have the luxury of being able to appreciate elegant code. MS has made it easy for people to get into the game late and do well. I applaud them for that. Linux now has an opportunity to come into those same offices and show its stuff, probably by way of a couple of borrowed RedHat CD's and a spare Wintel machine. What will it do? Again, I say to you, borroring Dr. Clark's phrase: Boys, Be ambitious. (http://www.hokudai.ac.jp/bureau/socho/message-e.
Jim O'Connell
Tokyo
-- My Weblog.
...but a form of this is an existing, shipping feature: An NT DNS server can refer to WINS for hostname resolution.
Adding the functionality to the NT DHCP server is a rather minimal, but helpful step: it just means that DHCP clients that aren't WINS clients can participate.
The real question is, who in their right mind would use NT as a DNS server in a production environment?
Unfortunately, there are places that deny access if you're not registered in DNS. This was the main reason I pushed to get the office I controlled at the time, in the DNS.
[John]
Shit better not happen!
People may say that this is bullshit all the way, or they may just love anything that comes in a shrinkwrap saying Microsoft on it. Personally I don't think that anybody who has their sound mind intact will implement this until it is standardized and proved working on all platforms. I can't really see where this would be beneficiary at all. So Unix admins, be calm. And remember that Microsoft is scared. It must do these things. It will backfire upon them eventually though. Right now, I don't believe that anybody will implement anything until it is proven.
Wealth is the product of man's capacity to think. -Ayn Rand
The article said it resolved by machine name too. Is this an smb thing? if so then it's a real embrace/extend/shutouteveryonebutmicrocrap. Those stupid Machine names ala SMB are not the way to resolve this, but Microsoft has been known to do stupider stuff. They're trying to make everyone that doesn't know what they're doing switch to windows so they don't have to RTFM.
I'm not dumb here. Having put together Windows NT networks, I know that WINS resolves NetBIOS names to IP addresses. If you bothered to read the rest of the post, and the reply thereafter, you'd see that the MS DNS server can query the WINS server, so that DNS and WINS names of a machine are identical.
Branded UNIX possibly, but Linux has much more security issues than NT. Simply pout, Linux is _less_ secure than NT.
maybe i'm completely off, but cmu uses dhcp, yet i have to give them my mac address first. that way when i ask for an ip cmu knows which ip to give me.
this has been in place since i have been connecting to cmu with an ethernet card. (4 years?)
is this what they are talking about?
It is your opinion and you are in minority. I like Windows interface, never used SGI but for example I found Mac to be very unfriendly.
When I moved from Amiga to Win 3.11 it was sort of let down but Windows 95 has very good interface (save for couple problems like unresizable file dialogs.)
Every Unix I used has shity GUI ( KDE is nice but due to X slowness is half the speed of Windows 95 GUI)
That's bullshit. I just did install of NT server 4 and aplied SP5 and works perfectly. You are spreading FUD my friend...
MetaIP UAM is doing the dynamic name to IP for a year or so, and this product is available for Solaris. LINUX and winnt.... No need to wait for DDNS from M$...
---------------------------------- Joni Eyal yoni@tty.co.il
and you truely are an "Anonymous Coward".. create an account you lazy freak.
How we know is more important than what we know.
Its a shame thats all you could comprehend
Hypothetically, anything hypothetical is possible.
This is the guy who complained in his last "article" that Linus doesn't answer his phonecalls anymore. (surprise!)
I guess Scott has no idea about DDNS whatsoever,
PLEASE Slashdot, stop posting his FUD
Being the eternal optimist I have to say something nice again. Maybe this thing is actually a GOOD thing provided: a) Microsoft sticks to the RFC's (i haven't heard anything that says they don't - except maybe their track record) b) Linux/*BSD/... gets this functionality too - i'm sure they're already working on it. As to whether thit is useful: I definately think so. It would be nice to have a fixed DN even though we're ip-hopping. Most of the people without a T1 line (including cable/adsl/isdn) get their ip by DHCP (if i'm not mistaken DHCP was also an extension by MS of BootP, no? not sure - but everyone uses it) It irks me a bit sometimes that ppl never give MS the benefit of the doubt. It's true that they may have a crappy policy and/or OS, but it's not by bashing MS that linux becomes a better OS. (cfr linux-advocacy howto) I'm just a user that wants to get some work done. obs
I suppose I should just read the RFCs and see what they say...
--
This is not my sandwich.
Okay, I'm both a Unix/Linux sysadmin and a Windows NT sysadmin, and I can tell you this. Win NT 4.0 already does this in an ad-hoc sort of way. The Win NT DHCP server can "bind" to the MS WINS server so that when leases renew/expire, the WINS server will be properly updated to reflect changes. And, you can also "bind" the MS DNS server to the WINS server. So, in essence, you have a workaround way of getting dynamic DNS.
So, this is nothing new really. In fact, most corporations that I know of (including the one I work at), already use Win NT-based DNS servers on the LAN so that it can be "linked" into DHCP. This is essentially a requirement already, unless you want non-NETBIOS (normal old socket-based) TCP/IP to not work at all for your clients.
Essentially, MS is already forcing you into using their DNS server on your LAN if you're using DHCP. The other option, is to use static IP addresses, have no WINS server, and use DNS for NETBIOS lookups. I suspect, however, that that is changing as an option in Win2k.
This doesn't mean that BIND is going to lose any market share. As far as I know, most corps use Win NT DNS internally and BIND externally on the Internet. That's what we do at my work. And it's what we did at the place I worked at before, and the one before that too. No one in their right mind would use the Windows NT DNS server on the Internet, although some people probably do.
If Microsoft uses Solaris on their main web site (http://www.microsoft.com), then how come I constantly get ASP/VBScript errors on the site?
Perhaps you are referring to Hotmail, which runs on Solaris servers and is supposed to take 4 years to migrate to NT?
It runs on dedicated hardware and it doesn't come cheap, unfortunately...
-------
Warning: Slashdot may contain traces of nuts.
I have a laptop (usually running Linux, occasionally running Win98) that gets its IP address via DHCP (from a Linux server at home, an NT server at work).
At home, because it's almost always the only DHCP client, my laptop always gets 192.168.0.10 (the beginning of my assigned DHCP range), so I can pretend it has a fixed IP address for local DNS purposes. At work, it gets a different IP address almost every day. WINS can resolve its name anyway; DNS can't because we don't have DDNS yet. MS supporting DDNS is good; my Solaris and Linux machines (which have clients for DNS but not WINS) would be able to look up my laptop by name, just like my Windows box (which has clients for both DNS and WINS).
Yes, MS might screw up DDNS, through malice or incompetence, and provide something only 99% compatible with the RFC. Recall the pump DHCP client included with Red Hat 6.0, which worked great with most Unix DHCP servers but not with NT's. But note that it was quickly patched to work with NT. Open-sourced clients can quickly deal with a bit of incompatibility, whether malicious or accidental.
The fact that MS supports a new open standard like DDNS before your favorite OS does is a reason to start working on an open DDNS client, not an excuse to bash MS. DDNS is good. NT becoming more standards-compliant is good. If at some point in the future MS starts changing their DDNS server around to deliberately cause problems with other people's clients, *then* bash MS, and suggest to your local sysadmin that he run DHCP and DNS from a cheap Linux/*BSD/whatever box instead of an NT server to maintain maximum compatibility with existing clients. But bashing MS in advance just for announcing the intent to support a good, new, open standard is counterproductive. Would you really prefer WINS?
-- David Ripton
I could be totally wrong about this but couldn't this be a security risk? Just allowing computers on yout network to set their own name? I know my
logs have a few entries already where people on the campus where I work have tried to enter stuff into my DNS server. If somebody wanted too they could just make up a name that is close to that of another one
Example:
a server has bubba.admin.whatever.int
a devious person sets his name to bubba.worker.whatever.int and somebody telnets/ftps/etc... to bubba and gets bubba.worker instead of bubba.admin because of a search path issue..
I could be overly paranoid which is very possible, so somebody set me straight on this...
>> Get with it, Microsoft. You'll never win by going this route.
That's just the problem. They *will* win by going that route (or at least they can, it depends on competitor's response). The problem is that they cut so many corners, and throw their weight around the market so much that they can muscle out the competition, and have a "newer, better" [read: embraced, extended] product.
MS has been so successful because they're practicing business, whereas most others are just trying to improve on current technologies. Only MS "improves" their own technologies because they cut so many corners in creating it that nobody else (at least nobody else knowledgable enough to be improving on current technologies) would want to work with what MS has created.
With DDNS, the hostname is bound to a device and the IP changes. With static DNS, the IP is bound to a device and the name changes.
ayottesoftware.com
Not quite ... the WINS lookup feature on NT DNS
will, as you pointed out, return WINS client values in response to dns queries to that NT
DNS server. However, it does not actually add
the WINS entry to the domain zone like DDNS does.
This doesn't really become too much of an issue
until you start setting up other DNS servers like
BIND to become secondaries to that NT DNS server.
Because the WINS entries aren't actually in the
zone, the secondary BIND servers CANNOT retrieve
the WINS values or lookups. Further, they cannot
retrieve the info from the WINS server directly
since the WINS resource type is not RFC compliant
and only works on NT DNS.
So, in this scenario, the non-NT DNS servers
are incapable of returning dns lookup queries
for any WINS client.
MS implemented DDNS in win2k to actually inject
netBIOS names into the dns zone like DDNS does
in ISC DHCP 3.0 and the nsupdate tool in BIND 8.x.
Even if UNIX couldn't do dynamic DNS (which it can), why should this cause any UNIX admins to get all stressed about NT admins taking over the DNS?
I mean, isn't this sort of thing what subdomains are for? "Here, NT doods, you're authoritative for windows.corp.foo.com." (Edit subdomain name to taste, of course.)
No stress, no power struggles, no problems.
The DHCP server communicates with the DDNS, advising it of necessary updates. You set up your DDNS server so that it only accepts DNS change requests from the DHCP servers in your network.
How is DDNS related to DHCP? I would think that the DHCP server implements DDNS... and I thought that clients and servers for DHCP were already available (for Linux). But I am not wise in the ways of Bind. Am I missing something?
i don't think a client has anything to do with this, in the standard terminology. Instead, the DHCP server just lets the (D)DNS server know that the machine with MAC number xx-xx-xx-xx-xx-xx was just given ip address zzz.zzz.zzz.zzz. The DNS server then updates the DNS record corresponding to machine xx-xx-xx-xx-xx-xx. All the client needs is a plain, olf-fashioned dhcp client. Of course, i could be off on this. micros~1 probably did it differently anyway, to break compatibility. Don't flame, just correct.
i am sure, however, that the bind version 8 name server does allow dynamic DNS updates. bind 8 has been out for Linux for quite a while. Also, several dhcp servers have the ability to automatically use this feature of bind 8. So you can't exactly say that "Linux/*BSD/whatever" doesn't have this ability, since it is already available and has been proven (which micros~1's hasn't yet).
A lot of the M$ bashing is because of two points.
- The article claims Linux has no support for dynamic DNS (as do you). This is incorrect, see above.
- M$ has a history of taking something that works and changing it enough that their client/server only works with their server/client. Especially since they're claiming this is something new and something that won't work with Linux, there's reason to believe they will do so again.
Admins are also worried that there will be problems when micros~1 tries to tie WINS computer names into the dns database. Especially since many don't even know how to change their computer name, so you'll end up with 20 computers named "workstation"...-----
--
perl -e'$_=shift;die eval' '"$^X $0\047\$_=shift;die eval\047 \047$_\047"' at -e line 1.
HeUnique prints an excerpt from a separate work to provide context, and provides a reference to the original source.
Short and to the point, and follows the standards for good journalism.
Look, DDNS has had basic support in BIND 8.x for
a couple of years now (with increasing quality in
each update). BIND administrators could do DDNS
injections with the nsupdate tool provided with
the BIND distribution, they could roll their own
scripts with the DNS module for perl, or they
could incorporate DDNS into ISC DHCP with Irina Goble's scripts or with the new beta code in ISC
DHCP 3.0. Cisco's IP Registrar and MetaIP are
commercial products that perform similar functions.
In point of fact, the article got its history wrong. The examples cited above, along with
the relevant RFCs, were help up to Microsoft as
examples of why the WINS lookup feature in NT DNS
was so inadequate. The incorporation of DDNS updates into the actual zones via IXFR allows for
propagation of the DDNS injections to heterogenous
secondary DNS servers unlike the WINS lookup "feature" in NT DNS.
so, don't fret about MS getting the upperhand on
DDNS, in and of itself. BIND 8 (and the forthcoming BIND 9) has got MS beat. What this
article was trying to allude to is this:
Although the Win2k implementation of DDNS might
actually be RFC compliant, it is quite probable
that the Active Directory tie-ins to DDNS will only work with the Win2k DDNS server. This is
crucial: even if Active Directory sucks, it has
huge appeal for Corporate America. It will be
implemented far and wide. This is what the article
alludes to: you'll have Win2k Active Directory
admins trying to require the DNS hostmasters to
convert to Win2K DDNS to support their Active Directory application.
This is a crock! The first thing the Win2K workstation must do is check for an existing name same as it used to do with a NETBIOS claim-name call.
I can only guess that you have overlapping zones of authority in your namespace.
With a proper implementation of BIND (or Win2K DDNS) the conflict is detected, the registration is refused, and the client system will note the conflict in its errorlog.
So why can't you do that and attach some sort of responsibility to your postings.. personally I say, no register, no post. At least be accountable for your own opinions.
How we know is more important than what we know.
The kernel talks to hardware about services not features. That's user space software's responsibility.
C'mon get with the program. The radiator in your car is not part of the engine. Neither is [D,S]DNS part of the kernel.
The ship sank. Get over it. (This sig was cut out from another's shirt and painstakingly hand-posted)
DDNS has been a part of the OS/2 tcp/ip suite for quite some time.
Kernel = services.
User software = features.
Should we have sendmail in the kernel too? "Ok... now to reconfigure sendmail, recompile your kernel."
I don't think so.
And no I don't like the kernel fractals idea either.
The ship sank. Get over it. (This sig was cut out from another's shirt and painstakingly hand-posted)
With static DNS, the IP is bound to the device, but the name doesn't change. Since the device always has the same IP, the name can be tied to device or IP with the same effect. Because it's easier, the name is bound to the IP.
With dynamic IP assignment, the name is bound to the IP as above, because that's how the database was written. But since the IP is not bound to any particular device, both the device's name and IP change. If you manage to get the same IP assigned, you'll have the same name.
As a side note, in DDNS the name is still bound to the IP in the name server database. The difference is that this binding is changed when the device is assigned a new ip, so that the name is effectively 'bound' to the device.
-----
--
perl -e'$_=shift;die eval' '"$^X $0\047\$_=shift;die eval\047 \047$_\047"' at -e line 1.
At present, every time you change the network configuration, you REBOOT. Big [bleep] pain if you have a PC laptop.
If your machines are configured to dynamic addresses, then you don't need to reboot.
If you're going to use DHCP, then the Right Thing(tm) is to use DDNS for both forward and reverse mapping.
DHCP provides for automatic, dynamic assignment of IP addresses. As others have pointed out, you don't have to use that feature, and you can centrally assign static IPs based on ethernet hardware (MAC) addresses, but that is a feature provided by BOOTP, the ancestor of DHCP. If you're really using DHCP then you're dynamically assigning IPs.
If you're going to dynamically assign IPs, you should have a mechanism for tracking the name-to-IP mapping. DDNS is that mechanism. Using DHCP without DDNS is Broken and The Wrong Thing.
It is only logical that systems that use DHCP should also support DNS. The fact that many of us (including the distro packagers) are too lazy (self-incrimination, not flaim bait) to set up DDNS does *not* mean that it is a bad idea. The fact that MS is doing The Right Thing (in this rare case) also does not mean that it is a bad idea.
So all Unix sysadmins in mixed shops: if you want to continue to have responsibility for IP assignment and DNS, then hear and heed the warning... get DDNS working with DNS on your machines *now*. When Win2000 is deployed in your company (a year from now?) you'll be ready.
"in a coperate envoiornment, this isn't an issue"
I agree with you completely, because this strengthens my theory about MS's server strategy.
DDNS may not be a compelling solution for a global, public network, but it sounds as though it's a very nice option for a local net, and that's where Microsoft is concentrating their efforts.
It is important to remember that the Winxx platform is not the logical center of Microsoft's empire. MS Office is. MS Office is the "killer app" which makes most businesses buy Wintel boxes on the desktop, and Windows on the desktop is why those same businesses buy NT servers. The presence of MS Office for the Mac was a significant factor in Apple's resurgence in sales.
Microsoft is leveraging this advantage very effectively, integrating Office with IIS, and with DDNS they are now making it even easier for any salesperson to connect their Windows laptop to connect to any open ethernet port in the office and start working immediately.
That, all by itself, is a good thing. What is not a good thing is for MS to specifcially design their ActiveDirectory so that it requires DDNS. Novell's NDS doesn't require DDNS, and from what I've seen ActiveDirectory does less than Novell's solution. I'm sure that the programmers behind W2K are very good at their jobs, so I must assume that the decision to make W2K DDNS dependant was a conscious choice. If MS publishes a white paper stating the reasons for this, I will read it, (and the soon-to-follow slashdot commentary) and make my mind up then.
PC Week deserves criticism for not doing their homework on this (no surprise there). To state that Unix does not offer this service, when it does, is terrible journanlism.
But then, any "news" article about Windows 2000 which is followed by a link titled
"Check prices: Windows 2000" isn't actually journalism at all, it's an infomercial.
toing@deus:~ $ date
Sat Aug 28 07:40:03 PDT 1999
toing@deus:~ $ touch duh
toing@deus:~ $ ls -l duh
-rw-rw-r-- 1 toing toing 0 Aug 28 07:40 duh
care to explain?
...
Bitchslapped? Give Rob a bitchslap from bitchslapped.com.
I mean really -- most everyone uses some other form of comment sorting. Yet I still keep seeing some AC comments right smack in the middle (well, usually towards the bottom) of the comment list that still insist on being the first post.
Are these people just stupid or what?
As we all know this is exactly what MS described in the Halloween doc. As someone else said the DHCP/DDNS via WINS resolution has been around for a while and that created already enough damage. Let me explain. Since the majority of machines our there use Windows and they have name resolution working, the Linux users are seen has "trouble" by the IS people. It's just a fact. Even worse if only few people complain about lack of name resolution. It seems incredible but there are actually people working in a IS department that, when facing the problem to buy Networking solutions, buy Windows-based packages. Windows never played nicely with Unix and it never will. If youi find yourself in the situation where you are denied some network services I suggest to gather together with all the people that have the same problem and talk nicely to you IS people. They will probably buy better software than try to manage 50/100 IP addresses manually. After all they switched to dhcp exactly to avoid that kind of mess. I speak from experience, it takes some time and a lot of diplomacy but it can work.
My mom deals with a lot of students getting coops and internships at Drexel University. She has students who are becoming MCSE's who hate Microsoft and Windows, and love Linux. It's just like all the other bullshat you learn as a means to an end.
grep -ri 'should work'
We have static IP + DHCP...so its great!
not subject to tens of new kernel updates which severly effect the stability of the system
Absolute bullshit. Every Service Pack, every HotFix we had to apply, was breaking something.
I just did a plain istallation of NT. I consequently applied SP4, IE 4, VStudio 6.0,
SP2 for VStudio 6.0, then SP5, then IE5.0,
then Office 2000, then SP5 again. If I do it in
any other order - it dies and BS on reboot.
This system is a piece of crap.
You are full of shit, do not know what you are
talking about, and probably only use computer to write PP presentations. Clueless idiot. Go back
to your hole.
<^>_<(ô ô)>_<^>
Remember in 1998 when Microsoft invented the acronym OSS to stand for Open Source, but thanks to Microsoft we all started calling it OSS for Open Source Software. Now we're starting to go back to just Open Source. Now DynDNS has been around for several years but now that Microsoft has called it DDNS we can expect everyone to start throwing around the DDNS acronym. Where would we be without the innovation of Microsoft?
FireWire is dead anyways.
Compaq was the only PC manufacturer besides Sony that was pushing it, and Compaq is behind USB 2.0 which boasts 200 - 300mbps, which is far more than enough for most perhiphrals.
Really, if Apple wants their stuff to take off, get rid of the greedy bastard who thinks high licensing fees and countless lawsuits are needed to prove his stuff is best.
www.atacomm.com - The Leader in VoIP Product Distributi