Slashdot Mirror
Worlds Slowest NT Server
Thom writes "NerdPerfect is holding a contest to find the NT server with the slowest reboot time. The best (worst?) time so far is 49 minutes, 13 seconds. Go check it out."
← Back to Stories (view on slashdot.org)
← Back to Stories (view on slashdot.org)
Yeah, it's a "microkernel". :)
But yes, that did aid portability before they dropped support for everything. Now it probably just slows them down.
VMS had some cool features. Unfortunately, NT didn't use any of them, so all we have left is some legacy weird architecture. How stupid is that?
---
pb Reply rather than vaguely moderate me.
pb Reply or e-mail; don't vaguely moderate.
A VMS cluster can take all morning to boot. But once it boots it will stay up until your UPS's fail for the next time.
What's the point of having a short boot time? Does Linux crash so often that it matters that it boots in only a few seconds? Or is this yet another meaningless benchmark that the Linuxer's are bringing out?
Capable?...Yes. Easily capable?...I don't know if I'd go that far. There have been days when I've had to reboot my NT machine at work half a dozen times or more. Oh, the joy!
-- Linux...find out what you've been missing while rebooting Windows NT!
"Longest time for a Linux newbie to figure out how to change the screen resolution in Gnome" I
guarantee it'll be weeks....
So we are talking about an linux newbie with a MSCE right?
I have to return some videotapes...
or like watching NT boot....
The angle of the Dangle is equaly proportional to the heat of the beat. ---Beavis
I win.... i gave up and reinstalled NT.
it doesn't make decent-sized partitions
WTF is that supposed to mean?
You DO realize that NTFS partitions can be up to 2^64 bytes in size, right? You knew that, right?
You also knew that NT has to have a boot partition within the first 1024 sectors, right?
Yep, you knew all that. Keep talking like *you're* the expert on NT.
Actually, go to his site and click around for a while.. You'll find a funny article about how M$ wanted to charge him for tech support on bugs listed in their tech base.
I called them once to confirm that a bug existed in Internet Information Server (the key word is *confirm*, it was plainly obvious there was a problem, and I already had my own work-around). Because they gave me some stupid-ass workaround (that was unacceptable), they claimed to have solved my problem and wanted to charge my company. I argued about it until the phone lackey gave up by leaving the decision to his superiors. As far as I know we never got a bill.
The most infuriating thing was that the very next day, the bug, and their lame work-around, showed up in the knowledge base, when it was not there before. Apparently, they don't admit to bugs until they can charge someone for "discovering" them.
For the curious, here's the bug.
fsck can take a wile with just 4 1.7 gig drives. But since it never crashes I dont see that much. fsck is nothing compaired to what it takes that box (URL at top of post) to recompile. No 4.8BogoMips isnt THAT slow :)
I have to return some videotapes...
BOOT/DOS partitions are limited to 7.8gb. NT Setup is limited to creating 4gb partitions because it's running in real mode and limited to 1024 sectors when it creates them.
After NT is installed, you can create NTFS partitions with Disk Administrator up to 16 exabytes in size.
That's right, 16 exabytes
-witz
Newspapers report that an NT server was recently discovered in a cave on the bottom of the ocean near the comores islands. This particular type of server had previously been believed to have been written off over 65 million years ago, but here it was, and still booting.
If anyone want to win, all one needs to do is to setup a start up script that caught into infinte loops.
Our exchange server has taken well over an hour to shut down before. And, of course, since it's an NT box with a lot of traffic hitting it, it's gotta be rebooted every couple of months.
Now that they're moving everything to Active Directory services, just imagine, you can have your website's registered users, your email users AND your in-house people all mucking up the same "technology" at the same time. And the API is almost useable! Lucky us, I tell ya.
At the last place I worked our Exchange servers had like 300 GB of disk space each. When it came time to reboot (about once per week they would crash/blue screen) it would take about 4-5 hours to check the disks. Most of the time admins would just do a preemptive reboot every Saturday. Sometimes the machines wouldn't shut down though (causing disk corruption and the subsequent 4-5 hour wait).
What I see - and I see it very often - is that a NT machine will start up pretty fast, slow down over a the course of a few hours of hard work and finally all but stop. At that point, the system may not have technically bluescreened, but it sure isn't any fun to work on. The only cure that I know of is a reboot.
I never see this problem on Unix systems - I think it has to do with the way Windows applications interact with the OS, with all the shared gunk that clogs up everything.
D
----
Its not your site deal with it
The angle of the Dangle is equaly proportional to the heat of the beat. ---Beavis
Perhaps RH should start starting sendmail in the background, then? I've never seen sendmail take any more than 2-3 seconds to start. And I'm not a sendmail fan either (go Exim! ;-) ).
/* Steinar */
(This comment is of course GPLed.)
Oh, most certainly. SBS has so many servers running on it that it takes forever for them to shut down in order. I had a similar experience with a client where SBS took nearly 2 hours to shutdown on its own.
The "official" fix recommended by Microsoft books is to use a batch file that shuts down the servers in a particular order to speed things up. It does help, it cut the shutdown time on my client's machine from 2 hours to about 20 minutes.
Here's a writeup on the batch file and how it works:
How To Do a Quick SBS Shutdown
I got the point of this immediately - they're not trying to make NT look bad, or make Linux look good.
It's like a sysadmin bitching contest - whoever has the best story wins the prize.
Not when it's NTFS, it doesn't...done on the fly, not on boot.
Of course, that is what is supposed to happen but my experience has been quite the opposite.
We have a rather flakey NT file server at my company that will slowly begin corrupting new saved files. Of the many proceedures tried to fix this problem, the only consistently successful one is to run chkdsk c: /f which will wait for the next reboot and then run a full check of the hard drive before starting up.
This command is more benificial (and worthwhile) than the regualar file checking performed by NT during standard operation.
Ever notice how the real power to Windows NT is accessed though the command line like the file checking command above?
Neither was the Mindcraft test. Fight fire with fire.
That would have to be a big fsck, or a really big drive. My 2x4 GB RAID-0 cluster (long live Linux, taking RAID to the home users!) takes perhaps two or three minutes to fsck. Should be even better (hopefully), when I get them on my new Promise ATA66 controller.
Of course, in a business environment, you could have a 8x30GB RAID-5 cluster...
/* Steinar */
(This comment is of course GPLed.)
If the samba implementation is better, I see no problem here...
A few folks spend five minutes putting up a web page and telling people to email in their results. That's the point. Oh, and to spread FUD about NT. For some reason there is a substantial population out there that thinks that geeks invent lies about how bad NT is for no reason whatsoever.
Ours are Netware 4.11, but the central computing lot are a mix between 4.11 and 5; not sure which ones they had which were taking ages to boot.
--
My Novell servers 4.10 or 4.11 (36Gig of total disk space) NEVER took more than 5 minutes to boot even after a crash. Lovely machines with little RAM. Now we have to run some NT monsters, but boot times are not that much worse. Of course, these are nothing but file and print machines with no other services. HP-UX handles everything else :)
The 6x86 has the RDTSC (read timestamp counter) function -- you just have to enable it by toggling some CPU flags (I think). I've seen a Windows program to do this. It wasn't entirely standards-compliant (!), though -- didn't always generate unique timestamps (over a period of 30 years or something...).
Check the `amendment' to the FAQs -- no emulators (OK, VMware isn't an emulator... or so they claim) allowed now.
/* Steinar */
(This comment is of course GPLed.)
...something an NT server will be good at! ;)
"Total destruction the only solution" - Bob Marley
He said to start timing when you select "Yes" -- I'm assuming that's when you select "OK" or whatever from the shutdown dialog, not when you select the OS after the BIOS/Hardware initialization. So that includes BIOS and hardware, a non-OS-specific slowdown. I start my HP Netserver LC with 2 slow-to-detect SCSI controllers, it takes a while, no matter which OS I'm running.
Of course, the point is apparently poking fun at something people have to do all the time, not specifically MS-bashing. In that context, I suppose it makes sense for that pain to mean something. :)
phil
Read the KB articles. The system partition (acc. to MS, this is actually the BOOT partition, the one where the bootloader files are located, not necessarily %systemroot%) is limited to 7.8gb.
This sounds like a cool contest, but how are the times being checked? Or is this like the Darwin awards, where if you can come up with a reasonable answer, you're good to go?
Not that I don't believe that it could take Windows NT 49 minutes to reboot, I'd just like to see it.
by Mike Buddha -- Someday the mountain might get him, but the law never will.
The 99.9% uptime figure is likely for a failover cluster. That would leave only 8 hours of downtime a year and while NT isn't as horrible as most of /. makes it out to be, it isn't that good either. The only way I can imagine having less than 8 hours of downtime a year without redundant servers would be if you needed no service packs or hotfixes and you did not have to install or upgrade any software on it.
As for the people who have noted slow boot times on high-end machines, I have too. I seriously think that NT takes significantly longer to boot on a large machine though I have only my personal experience to back that one up and that's not worth much, eh? At my last job I had the 'pleasure' of administering a whole mess of NT boxen and the PII/233 with basic SCSI booted SIGNIFICANTLY faster than the quad Xeon with hardware RAID-5. Of course, once they were running the Xeon was a wee bit faster.
I have a machine sitting right next to me that has an incompattible IDE card. I once installed NT server on it in fact, for a client of mine (as a backup server they could switch on if the other one went down). But NT loaded a driver for the IDE card (the wrong driver) and it had problems booting (I would get an IDE interrupt! message over and over). Must have taken at least 3 or 4 hours to boot (when it did), as for rebooting, it would freeze during shutdown, so I guess my time would be from the time I started the shutdown untillteh power went out long enough for the ups to run out of juce. I'm not sure I could put up with having to reinstall NT on that damned thing though!
the slowest i have personally seen was about 15 min.. and that to me is WAy too slow.. i dunno what it is.. but 49 min to reboot would make me have to hurt someone.. bad... :)
----
That's right. 8M of ram, 300M HD. I wouldn't even wait for the log in screen. Just turned it on before going to bed, login in the morning. :D
And yes, NT really did have a 3.1 version. It was fab.
That's what Coda (or AFS, if you want) is for. NFS has some fatal flaws (like waiting infinitely), but it was the only working thing available at the time it was written.
/* Steinar */
(This comment is of course GPLed.)
It wouldn't be that hard to add up the time & rig it for slowness. Put the maximum amount of slowest speed ram possible, max out your pci slots with scsi cards & max out the scsi devices with really slow drives (they spin up one by one to minimize power strain), mess up your drivers and the network configuration cuz it takes extra time to error out, and there ya go. 49 minutes is quite unbelievable tho, unless it's a 486 with all the stuff I just talked about.
It's hardly a credit to the ball bearing manufacturing community either. I bet the Gulf of Mexico fishing community is feeling pretty uncredited as well. And the US Civil War re-enactment community is left right out in the cold. I don't even want to mention how alienated this contest must make the disco community feel.
What does any of this have to do with Linux? The contestants are all NT users.
... exhibitionist masochist network administrators. Is this news, or daytime television?
Oh, and my linux box (P60,40MB RAM) boots in the same time as my NT box (P200, 80MB RAM)
Computers, Linux, NetBSD,
I have a copy of NT 3.51, running on a 12.2mhz OC'd V-series x86 clone, with 4M of memory and dual 400M IBM full-height MFM/RLL drives.
At thirty+ minutes to restart with no services/processes running, I think it could win. Anyone ideas on how to make it slower? (I'm not going to engage in anything like the Exchange kill timout bug. Thats cheap.) Can't strip out any of the HW, I'm already below the minimum req's..
.sig: Now legally binding!
I don't think you can say that SMB is totally proprietary. Quoting from the Samba meta-FAQ:
SMB is a filesharing protocol that has had several maintainers and contributors over the years including Xerox, 3Com and most recently Microsoft. Names for this protocol include LAN Manager and Microsoft Networking. Parts of the specification has been made public at several versions including in an X/Open document, as listed at . No specification releases were made between 1992 and 1996, and during that period Microsoft became the SMB implementor with the largest market share. Microsoft developed the specification further for its products but for various reasons connected with developer's workload rather than market strategy did not make the changes public. This culminated with the "Windows NT 0.12" version released with NT 3.5 in 1995 which had significant improvements and bugs. Because Microsoft client systems are so popular, it is fair to say that what Microsoft with Windows affects all suppliers of SMB server products.
/* Steinar */
(This comment is of course GPLed.)
I'm sure I could load NT on my 486 currently running Linux. It'll take at least 5 minutes to boot right out of the box. If I add a few service packs and essential services, it should get up to at least 20... What else?
æeee!
It's an interesting, and almost plausible theory, but Microsoft themselves recommends AGAINST this.
;-), the Microsoft recommendation was to never have more than either 2 or 4 thousand users/domain. (and this was the official microsoft documentation). If you ever check out the network structure of a large company (someplace like Merrill Lynch where Private Client Services has something like 60,000 nodes on the network) you'll note that it is NOT a single domain.
When I got my MCSE (know thy enemy
I can't see ANY reason why you wouldn't be better off to have a PDC and BDC at each satellite and then set up trusted domains.
If you have a single domain then that means that the central administrator would have WAY too much access. 'the admin walked away from his desk and forgot to log out of the god account, so I ran a program which changed 10% of all the passwords, and randomly deleted 2 files off of every machine'. This would be a Bad Thing(tm).
I think it was NT on Alpha at SP3 (or was it SP2???) that had a known problem that would take upwards of 30 minutes to shut down. I can't remember if it was related to a service or what but it used to annoy the hell out of me before we upgraded to SP4! just 15 minutes downtime I would tell the users...
(And now, I'll see massive holes ripped in this postulate. Bring it on!)
"So, what do you want to hack for, Bickle?"
"I can't sleep nights."
"They got porno theaters for that."
"Whatever happened to fair use?"
-- Duff-Man
Is it just me or is the very idea of this seem too stupid to bother with?
Oxryly
linux kernel compile times!!!!
and yes i know that there are different journal filesystems available, but which ones are being used by a distribution ? I was talking about using a stock install with ext2.
I have been a linux user for a long time myself, and my first machine with linux was a college lab computer that we made to dual boot! A 486 with 8 Megs of RAM! but i took longer than 45 seconds to boot!!!!
Either way, chill out, i am sure a stock install off linux with a HUGE file system will die too!
Non-Deterministic Finite Automata
...something an NT server will be good at! ;) I would have to say that an NT server taking 49 minutes and 13 seconds to boot would make a really good paper weight.. add some paint to the case and you have a work of art....
----
Sorry, forgot to respond to your first comment.Q 114/8/41.asp
Yes, I meant cylinders, my bad. It's still 1024. You may get it to work going beyond that, but MS isn't going to support that configuration and it can definitely lead to issues.
Read: http://support.microsoft.com/support/kb/articles/
Whereas the Master Boot Record is generally operating system independent, the Boot Sector of the active partition is dependent on both the operating system and the file system. In the case of Windows NT and Windows NT Advanced Server, the Boot Sector is responsible for locating the executable file, NTLDR, which continues the boot process. The only disk services available to the Boot Sector code at this stage of system boot up are provided by the BIOS INT 13 interface. The Boot Sector code must be able to find NTLDR and file system data structures such as the root directory, the File Allocation Table (FAT) in the case of an MS-DOS FAT volume or the Master File Table in the case of an NTFS volume. These must be present within the area of the disk addressable by the 24-bit side, cylinder, sector structure used by the BIOS INT 13 interface and the partition table. This limits the size of the system partition to 7.8 gigabytes regardless of which file system is used.
-witz
Sure, leaving out any form of security can only help your boot times.
BIOS support for large IDE boot partitions is a pretty new thing in the PC world, and NT 4.0 is getting pretty long in the tooth. When NT4 shipped, even SCSI BIOSes generally only supported a 2GB boot partition.
(I'd be happier if they just had shipped an NT4.1 rather than pretending that Windows 2000 was just about read for the last 2 years. That would have at least resolved the stupid install issues like large IDE disks.)
--
Business. Numbers. Money. People. Computer World.
Get some new RAM.
Preferably ECC.
The only way that would really be true is if someone was running a really small and slow machine, although, technically, the slowest NT machine possible would not be the slowest Linux one...I've seen Linux run on a 5 meg 386-33, which isn't even possible for NT, so you have to conclude that no matter how small a machine you managed to shoehorn NT into, if you put Linux on that box, it would still run faster then that brave little 386 that it took 60 seconds to pop up the prompt after you logged in.
If corporations are people, aren't stockholders guilty of slavery?
Look, at all the people proving your point for you! How wonderful! You are "childish" because you disagree. You should "grow up" because you said something that could be be remotely construed as pro-MS. Serves you right, MS jerk. Next time maybe you'll think twice before you try to be un-biased.
12 seconds, after the BIOS screen - with FTP, telnet, mail, NAT and a WEB server running. This is on a Celery 300A, 192Mb RAM, 6.2Gb hard disk.
Two years ago I was the Admin of a NT box with Exchange 5.5 on it. It had about 50 mailboxes on it, but all the mail was kept on the server (In one large file, no less). It would routinely take about 30-40 minutes to reboot, depending on what exchange was up to at the current moment.
Unfortunately, we had a power failure overnight and the server went down. Hard. Very Hard. Exchange was NOT happy about having open files all over the place. Exchange would take forever to try and bring up all its messaging services, and usually fail in thr process of initializing the internet gateway. Before we blew the whole mess away, it was taking almost 2 hours to reboot.
Surprisingly, the box was a PPro 200 with 224MB memory (128+64+32)
"I may disagree with what you have to say, but I will defend to the death your right to say it"
"I disapprove of what you say, but I will defend to the death your right to say it."
- Evelyn Beatrice Hall
"The slowest machine to compile the Linux kernel"
;-) /usr is NFS mounted at boot (the other machine is in .au)
Hardware:
- 386sx16
- 2MB RAM, 1GB swap
- MFM disks, UMSDOS fs
-
Services:
- X
- Apache
- Oracle
- Quake (all 3 versions)
Clients:
- X server
- Enlightenment (with nice bloated theme)
- Netscape
- StarOffice
- Quake 3
Background apps:
- MP3 encoder
- povray
- gcc (also compiling X11R6, JDK and Mozilla)
Did I forget anything?
Agreed. The only thing that Exchange and it's long shutdown prove is that the NT Services Control Manager is retarded.
This is an old problem on NT, and it would be nice if it was solved.
--
Business. Numbers. Money. People. Computer World.
Just install stuff.
:)
SQL server, Exchange, Transaction server, etc.
It'll take days to boot.
Hmm. With Windows 2000, you need 2 reboots, from putting the CD into the server, to having IIS installed and serving up web pages. Once installed, zero reboots. None whatsoever. Hah.
The slowest reboot for any OS would probably be the machines where you had to dip-switch in the BIOS. Then you had loaded the diskette boot driver, and from the diskette, you could load a program to boot from HD...
A really experienced person, working 100% all the time, could do that in 30 minutes or so. But then, the last machine of that type in Norway came around 1990 (I think), and it _still_ runs!
/* Steinar */
(This comment is of course GPLed.)
It may support huge partitions, but trying to get it to install on a > 8 gig partition (IDE disk), or IDE disk that 12 gig regardless of how it is partitioned is a royal pain in the ass as I discovered trying to help a friend. Out of the box NT doesn't have large IDE disk support so it can't install on one. As far as we could figure out Microsoft's solution was to extract the large disk support from the appropriate service pack and insert in into the .cab file on the boot floppy you use for installing.
It took ages to get that install working!
Start up.
Real mode portion of installation (if you boot to the CD).
Reboot.
Finish installation, set up networking, installed options, set password, etc.
Reboot.
Log in.
That's right, you're at SP1 (NT CDs now ship as SP1), but you're logged in after 2 boots.
Install SP5.
Reboot.
Now you're at SP5, and logged in, after 3 reboots.
Not 5.
Keep going, FUD-boy.
Not to mention that you can install IE5, MDAC 2.1 and assorted software/hardware drivers in the same boot. So with an additional reboot, you could be at:
NT 4.0, SP5, IE5, MDAC 2.1, and some 3rd party mailserver.
-witz
linux could _easily_ reach boot times that windows NT could never dream of, because you have the source.
Just open up the kernel source, and somewhere in the code for the opening crawl insert code that simply runs an empty while() loop until GetDateTime() is equal to January 7, 2004.
Hell.. long as you're at it, you might as well make it a for(;;);. Infinite boot time! woohoo!
Irritable, left-wing and possibly humorous bumper stickers and t-shirts
No, not for my NT server to reboot. :) :) oh, and I have a Mac Classic in the bedroom as a cheap notepad/xterm for my shell account, and it boots in twelve seconds (not much to boot). ;) "So what? They're only customers."
That's how long it took for my old powermac to quit Netscape, shut down PPP, reboot, start up PPP, launch Netscape, dial in, load Slashdot, load this page and return to where I was reading
For some reason I find this whole discussion insanely amusing
Steve Jobs once threw a fit at Apple developers, getting them to make the Macintosh boot faster. "Twenty seconds??? There are X million customers! If you waste a single second you're wasting time equal to X many entire lives! Are you ready to waste that many lives to your laziness???" (not a quote, just paraphrase). Maybe we ought to send him to Redmond to yell at some people. I guess the Microsofties don't mind the toll of lives they waste
something definitely wrong there. the closest config I've had to that one was PII 300, 128 ram running exchange 5.5 AND proxy 2. reboot time was in the 5-10 min. range... just had a thought, how does size of message store affect reboot time?
Ah, the Linux zealots are spreading FUD again, I see. Production NT servers SELDOM NEED REBOOTING as well. Or for that matter, NT desktops. My machines are usually up for months at a time (never had a crash with NT 4.0), they only go down for power problems or hardware changes.
My company used to do all our network stuff on a single NT server, and I mean ALL out network stuff. It was a case of upper management wanting to 'stick with proven software' but being too cheap to shell out for decent hardware to run it on. Needless to say, this machine was SLOW and took FOREVER to boot.
It may actually beat the current 49 minute record. Unfortunately, I don't know exactly how long it used to take to start, because I never had the patience to wait around while it booted. I do know that one time I powered it up just as "The Simpsons" came on TV. When "The Simpsons" was over, the machine still wasn't finished booting. However, by the time "Home Improvement" had finished, the machine was up and running. Therefore, it had a startup time between 30 minutes and an hour; probably closer to the full hour.
We're running everything off linux now, and the old server has been shoved in the closet. I'm halfway tempted to dig it out again just to see how long it takes to start.
Beer recipe: free! #Source
Cold pints: $2 #Product
NT usually deserves the flames..as anyone who has used it knows.
"The fastest boot sequence of any reasonably general purpose OS I've encountered was on Univac 494/Omega that I worked on back in the early 70s (when it was already nearing the end of its lifetime): given suitable peripherals it could boot and have the online workload running again in under 10 seconds - because it had been designed for a target market which could not tolerate extended outages. " Hmmm, try BeOS. It boots up in 10-15 seconds for most people.
hmm.. cntrl - alt - + now go kick yourself.
The Alphaserver running NT4 at terraserver.microsoft.com takes 24 hours to boot - it takes that long to init the drive arrays. I had to boot it once....
Irish
until it crashes.
Oops..thats 25GB disk space, not 25MB
My system always performs a fsck on startup on all the partitions, since I'd like to know my server is a happy bunny before starting anything else
Donte Alistair Anderson Roberts - hi son!
Karma: Chameleon
The kernel leakage is extremely low (if at all), so this claim is bogus. If your NT machine keeps slowing down, close your apps and re-open them (_NO_ OS can stop an app from being shittily written). At worst kill explorer.exe and reopen it.
I'd wager a pretty good amount that >99% of NT reboots are completely and absolutely unnecessary and are the result of ignorance on the part of the user, not at all anything to do with the OS. As Linux is hitting the mainstream these "horror stories" are starting to appear that are obviously the creation of a fool and his ways rather than any factual measured metrics.
This is a rather silly contest. Wait let me replace all my memory with a 16 MB strip of RAM and then severly underclock my CPU. What is the point of this contest again?
I did this once to see exactly how far down I could clock the CPU and still have Windows NT work. I only tested to 60 MHz but it took approximatly an hour maybe less. Login took aprox. 25 minutes to authenticate.
"Help me Obi-/.-Kenobi,your my only hope!" -$
Maybe I should change MY name to Anonymous Coward, because it seems like all the people who come onto slashdot and say that their NT servers have these unbelievable uptimes are Anonymous Cowards.
This whole AC thing reminds me of the spell-check police on Usenet: It is so ridiculous it is laughable. How does some bogo nick make you somehow more culpable? I don't both making an account simply because it's not worth the bother, and secondly I'm not trying to be a "Slashdot hero". I'm just another guy blathering away in the noise and a nick or not doesn't change that.
However on to your claims:
So I'm changing my name to A.C., and maybe, just maybe, I can make it through the day without a BSOD on my desktop machine, just like all these other Anonymous Cowards.
I do very heavy development in several different tools, SQL Server work, IIS work, I browse the web quite heavily, I run Office 2000 apps all day long, etc. etc. etc. I run everything through the network and my PC has a load of services. For the past six months or so I was running NT 4 SP5 (without a single BSOD or any sort of crash necessitating a reboot). For the past month or so I've been running 2000 RC2. I NEVER get BSODs. NEVER. This continual FUD that Linux wankers are trying to spew is quite hilarious, and the only one's having the daily reboots are the idiots. Check what's BSODing and FIX it. Got a crappy video card driver? Well get rid of your Trident 9680 and get a real video card (the Matrox G220 is an extremely basic good workstation card and Matrox' WHQL drivers are absolutely stable).
Heh. I had this fileserver running NT 4.0 SP3 one time, with five 9GB drives running as a volume set (makes them look like a single volume, but it isn't RAID). So when we ran out of space, I added a sixth drive to the set and rebooted. It shut down happily and went into NT's happy happy blue screen of fsck^H^H^H^Hchkdsk. About ten minutes later, chkdsk had incremented to 1%. Of the first stage.
All in all, that particular reboot took over eighteen hours. Beat that.
Causation can cause correlation
Nice Script. However, I doubt that using a batch file is a Microsoft approved method.
Microsoft spokesperson Dan Small once said "Yeah, but the scripting is almost a failing of Unix, not a virtue".
I use an NT system at home. I have dual Celeron 366s running at 550mhz. Using U2W SCSI (Quantum Atlas 10,000rpm), NT4 w/ SP5 takes about 20 seconds from OS Loader to Desktop, if I can type really fast. Once I login, the Desktop is loaded before the startup sound gets done playing. Can't say that about my BSD and Linux machines.
Actually, you can do this from the GUI as well. But, yes, a lot of NT can be done from a command line if you know what's being run behind all the buttons.
Either this guy really sucks as an admin or he some how managed to shove 20 different SCSI adapters into this thing.
NT has to be the worst system in the world when it comes to the speed of intializing SCSI adapters at boot. But I really think this guy just "super sux" as an admin. Or maybe he is sysadmin for ECHELON and they ported it to the super computer OS, the unstoppable Windows NT!
HEHE. "Unstopable Windows NT" I loved that commercial, it's funny, you don't see it on any more though.
"Help me Obi-/.-Kenobi,your my only hope!" -$
I don't think that VMWare would run on a 386, it requires the RCTSD (sp?) instruction. I don't believe that was introuduced until the Pentium, or maybe the later 486 (DX-2/4 5x86). VMWare won't run on my Cyrix 6x86 P150+ because it lacks this instruction, one of the few pieces of software that I have had a problem with.
-- Remember: Wherever you go, there you are!
Yes I have had this problem too, even longer than 49 minutes to reboot. You're right about the services getting deadlocked. Also part of the problem is users logged on to exchange. Exchange's client kill function isn't so hot, the easiest way to crash a win9x box is shutdown a exchange server while user are connected. Outlook crashes and takes down windows with it.
But if you want to cut down on reboot time do a net stop on the exchange services and then reboot your box. Guarantied to speed reboot
The world isn't run by weapons anymore, or energy, or money. It's run by little ones and zeroes, little bits of data.
Whatever.
ye gads, get an Honda or Toyota, much more reliable and better trimmings than *any* domestic car.
Installing Exchange Server always adds about 15 minutes to the restart time, but I don't think you're exactly right about why. According to Microsoft (there's a technet article on it), the system pauses for a period after stopping each service to make sure each has time to complete. You can change the delay time in the Registry to make things shutdown faster, or you can, as you note, manually shut down services first.
Can you wait till NT goes into SuperSlowMode, and then try rebooting? I notice that all my NT servers that I forget to reboot weekly will eventually hit "SSM" and crunch along, easily taking 10 minutes to bring up the task manager. So once in SSM, you can start a reset and wait for literally hours before NT can complete a basic shutdown. I ususally just pull the power (it's actually somewhat satisfying).
Note that this is on simple hardware (no loopy hardware conflicts), 20 Gig filesystems, and fairly modern systems (P II/300's).
sectors??? Even if you are refering to the 1024 CYLINDER limit, you're still wrong. I have NT 4.0 installed on an 18G SCSI drive with the entire NT partition beyond the 16G point. The machine works just fine.
So far, I've had no problems with any OS and those archaic cylinder limits -- that limit is only valid for the 3D mapping in the PC partition table; everything uses the linear address now.
This one wouldn't work for this contest; but, there was this machine at work once that took over-night to shutdown. It was saying "please wait while the system writes unsaved data to the disk" at the end of one day and still said it the next morning; a little later in the day it finished.
The t-shirt probably reads:
"I am the NT Admin whose machine booted the slowest in the big contest. Hire me, I am an expert"
It's to be worn at job interviews. And when applying for food stamps.
Screen resolution. Hell that's easy (Ctrl-Alt-+), now if someone could tell me how to change bit depth on the fly I'll send them a beer. I know how to do it at startup but I don't know how to do it without taking down X. Windows does it...so it must be easy right.
--
"L'IT c'est moi!"
Yeah, I was pretty pissed off when we bought Norton Tools for NT for a flaky machine in the lab, it kept doing Diskcheck every time we rebooted, so we wanted to use a different tool to do the disk checking while the server was "up".
Turns out, Norton CAN'T do a disk doctor on NT while NT is up, (like it can for 95/98 and Mac), all it does is reboot the machine and do a diskcheck.
IMNSHO, that's retarded.
At least it defrags - a file system that supposedly never gets fragmentation, haw haw!
I wish I had a nickel for every time someone said "Information wants to be free".
These are my friends, See how they glisten. See this one shine, how he smiles in the light.
Here's the hardware:
486-33
12 MB RAM*
The biggest, *SLOWEST* hard drive mobo can accept
Painfully slow network link
Any other stuff to be used
Install NT + service packs (they're at 6 now (!)) on another machine (unless you want to suffer, and have 20 more MB of RAM to spare). Install usual apps. Set NT to do network login (domain server) over painfully slow network link. Ensure several startup files happen over network link. Make a large swapfile (it'll be needed!). Fragment hard drive as much as possible, and make sure swapfile is likewise extremely fragmented.
Boot the thing up, go to bed, then do your reboot trial (might have to bring several books to read while you're at it!).
Remember Wierd Al? "You're using a 286, don't make me laugh/Your Windows boots up in what, a day and a half?..."
* Yes, I remember someone booting NT up in VMWare stating it needed 32 Mb to install, but can be brought down to oh, 12 megs for which it will boot up [extremely slowly]...
Just bring up your Linux system, then yank the power cord while it's doing a disk write.
Do that two or three times.
Get out the stopwatch and have some fun.
hm. Maybe I should change MY name to Anonymous Coward, because it seems like all the people who come onto slashdot and say that their NT servers have these unbelievable uptimes are Anonymous Cowards.
So I'm changing my name to A.C., and maybe, just maybe, I can make it through the day without a BSOD on my desktop machine, just like all these other Anonymous Cowards.
I wish I had a nickel for every time someone said "Information wants to be free".
These are my friends, See how they glisten. See this one shine, how he smiles in the light.
Don't forget that NFS was once a Sun proprietary network protocol and is certainly not optimal as a distributed file system. I'm perfectly willing to believe that Samba is as good or better than NFS or at least that the Linux implementation of Samba is as good or better than the Linux implementation of NFS.
--
"L'IT c'est moi!"
Hey, didn't you read the Linux Journal "Help" column a few months ago. Somebody wrote in with an NFS question regarding Linux.
The column maintainer answered by strongly discouraging the questioner from running NFS on Linux. NFS is one of the Achilles heels of Linux, and from what I've been able to gather up, it's not anything that's being worked on.
Get this- the column maintainer recommended the user instead install Samba on both systems and use SMB to communicate to filesystems between the machines.
Isn't that ludicrous? A writer in Linux Journal willing to badmouth NFS and talk up and promote a Microsoft Proprietary (and now reverse-engineered) network protocol.
Nah, that's nonsense - A friend at boeing has a Linux box connected to an 800 GB symmetrix drive - and no, it doesn't "die", whatever that's supposed to mean...
Oh sorry, you will just have to leave the 10 tons of butter out here in the sun. Computer problems you know..." (true story)
Do tell.
--
"L'IT c'est moi!"
I worked on a network server that would take an hour to boot because it had multiple network cards that all used DHCP which never worked on the first try.
So what would happen is each card would attempt DHCP, fail 5 minutes later. Then when the machine got into the login screen it would start loading services... including web services. It would then try DHCP again, and fail 5 minutes later with an error popup.
The sysadmin had rigged a macro to do DHCP the right way. Why they don't use non DHCP, I will never know. Or why they won't replace their Windows DHCP server, I will never know that either.
Windows is bad. Don't use it.
- Hugh Buchanan
- Userfriendly.com
Is that correct? To duplicate that kind of performance on my Linux box, I would have to turn of the L1 and L2 caches. BDKR
And what difference does it make if you post as "Anonymous Coward" or post with some bullshit fake name anyway? Either way it's anonymous.
My 486 DX-100 with 8 MB of RAM took over 24 hours to do a make zImage on a vanilla 2.2.12. Of course it was running an MUD bot (niced) at 80% CPU at the time.
--
"L'IT c'est moi!"
What could be done to speed up Linux boot ups ?
Write scripts in Perl instead of bash, perhaps ?
I don't think this would help much. The most time spent booting up is not the scripts themselves but the processes they start. Many of the processes have to configure hardware, open ports, and do DNS lookups. The biggest time savings would probably come from parallelizing startup of independent daemons. This would have the advantage of increasing speed, but the disadvantage of increasing complexity and adding potential race conditions.
--
"L'IT c'est moi!"
nuff said
Assuming a typical reboot takes = 99.99%
Our IT guy figued out after about a week of complaints (when I was in Chicago), that our local BDC had froze, and every person with NT would take like 5 minutes to boot because we were synching to Orlando.
But then again, after this IT guy quit, and we had NOBODY watching the NT machines for two weeks, we had not one single network failure. Before he left, about twice a week, something would get fucked up. He'd reboot boxes at lunch time, just because. There was another time when he jiggled a SCSI cable, just to check it out, and the RAID array went down, and took 4 hours to bring back on line. I pity the people he's working for now.
I wish I had a nickel for every time someone said "Information wants to be free".
These are my friends, See how they glisten. See this one shine, how he smiles in the light.
Assuming a typical reboot takes 10 minutes:
uptime = 100% x (1-(10/7*24*60))
= 99.99%
Well, Rob KNOWS I'm not some Steve Barkhto.
I wish I had a nickel for every time someone said "Information wants to be free".
These are my friends, See how they glisten. See this one shine, how he smiles in the light.
Where does the Linux community come into it?
I thought the page was a Linux advocacy page, but you're right, I may have jumped the gun.
Correct, almost. You cannot install the NT system onto a new disk in a large partition. You can, however:
- plug your new 20G disk into an existing NT system
- format the disk as one single 20G NTFS partition
- plug the formatted 20G disk back into your system
- install NT into the 20G NTFS disk on your system
I used to own an Impala with that piece of crap engine. It was in the shop more than my neighbor's MG. And for a car of that weight, it was about HALF the power it needed. The person who decided that this piece of soggy diarreah should be put in the Impala ought to be shot. Maybe in a Cavalier.
It's like the original Celeron chip, the one with NO L2 cache at all.
I wish I had a nickel for every time someone said "Information wants to be free".
These are my friends, See how they glisten. See this one shine, how he smiles in the light.
These contests are kinda going a bit far if you ask me. What do you really see from looking at reboot times ? Heck its NT!
lets talk kernel compile times !!! flags used ? ?
I am sorry, but reboot times for NT is just something that tells you nothing! Hell Linux would be BAD as hell on this one tooo... imagine having a VERY VERY large filesystem! well that is one place where linux is not too strong!!!
Non-Deterministic Finite Automata
Mmmm, c:\ ..... Yea no dos in there at all :)
/' means this linux system must contain some bits and pieces of the original pdp-7 unix...
and the fact that I can do 'cd
apparently you havent learned the look and feel concept.
incidently, command.com is not DOS, its just a shell, same as bash is not linux.
NT sucks total shit over a WAN, even with a BDC at each site. If you must run NT in a large organization bite the bullet and run a multiple-master model and learn how to maintain the trusts among the domains that need to share resources.
And why would you want to install it on a partition larger than 8 gig? Most admins will install the system on a fairly small partition and make a huge second partition to store all their data, much safer in case of a system failure, which can happen with any OS.
FUD+Flames on NT = good and +1 score
FUD+Flames on Linux = flamebait.
These rules should be published on the top page so everybody understands them.
You know the only people I ever see who say "anyone can admin an NT box" are Linux advocates spreading fud about NT and MS. NT is EASIER to admin than Linux/Unix but still not brainless.
Of course on a SysV init system setting up the order which daemons are started and killed is trivial. On NT you have a blackbox kiosk type OS that expects to be able to do everything for you and becomes very confused when stuff doens't go according to plan.
-- Remember: Wherever you go, there you are!
How about vmware machines? I might install NT into a vmware test box on my old 386! It'll take ages to come up without even applying a service pack! ;)
I've personally found NT to be more unstable than Linux on the same hardware (and I've seen funny things on Linux too) - so I think your claim on hardware being defective is illogical
how much would you wager? how would you design the settings where we could put this to the test? (no Mindcraft III, please)
Great contest idea, though a T-Shirt for a prize seems like a pretty half-assed prize imho :) Here's how I would handle the prizes if were running things:
Me: Congratulations, you have won the contest with an amazingly pathetic reboot time of 4 Hours, 57 Min! (or something to that extent)
Contestant: WOOHOO... What did I win? What did I win?
Me: Oh we have a special prize for you... this brand new, top of the line server... running on your favorite OS... WINDOWS NT!!!!!!!!!!!
Contestant: NOOOOOOOOOOOOOO!!!!!!! AGGGHHHHHHHH!!!!!!!!!! (goes insane and flys into a fit of violent spasms before falling to the ground, dead)
Now wouldn't that be funny and ironic?
BS. I could start a class action suit for the amount of time that is wasted at my company waiting for NT reboots. Change ANY IP parameter? Reboot! Change the FONTS(??) on your display? Reboot! Install some new software that shouldn't have anything to do with the operating system? Reboot! etc, etc... It is the BIGGEST fault of Windows (IMO). Sure, in a totally static environment, you may not need to do this. But static environements are boring, and we need to adapt our network and setups to incorporate new hardware, or shift resources, etc. Linux (and our other Unix boxes) do NOT generally require nearly as many reboots (perhaps one when all the configuration has been completed)
So yes, in my experience, NT requires at least an order of magnitude more reboots than Linux/Unix systems for similar amounts of configuration, installation, administration, etc.
Your point MAY be valid for production servers, but NOT for desktops.
"It's overkill, of course. But you can never have too much overkill." - Anonymous Slashdot Coward
I just got back from IBM's AS/400 Fall Technical Conference... one of the offerings was session on IPL. For a large system, they hope to bring the time down to 15 minutes (for a normal IPL). While that sounds all well and good, our ERP will then take several hours to 'catch back up' - running reports, processing unfinished requests, etc.
In any case, stability and up-time are much bigger issues than the time required to restart the system.
This story should've been in the Humor category, not the Microsoft one. ;)
I'm here to refute your claim that all NT users who claim remarkable uptimes are cowards. Note that my nick is real. The NT server that I manage for our development group stays up for MONTHS at a time!!! The only reason it goes down is 1) Long power outages beyond the battery life of my UPS (certainly not an NT problem); and 2) the crappy 3COM NIC with the damn bent bracket causes the PCI card to not seat properly!!! (again not an NT problem). Other than that, I never reboot it. The biggest problem I have found with most OS's requiring frequent reboots is the fact that the sysadmins won't leave well enough alone. They always have to tinker. The exceptions to this are Win95/98 (memory fragmentation problems in the crappy memory management require frequent reboots), the Mac OS, Solaris, and my own Linux box that I constantly tinker with (therefore I am to blame not the OS).
Take two linux computers, put 10 filesystems on each. Now NFS mount the 10 fs from one machine on the other, and vice versa. How do you shut them both down, without logging on as root and unmounting all the filesystems manually first, because the power has failed perhaps?
In case you don't know, if you have some NFS filesystems mounted, and try to reboot when the NFS server is down, it will take about 5 minutes per filesystem before umount -a times out.
Do you know what happens when win 98 gets a mem leak error? You can't do jack with it, because the memory is GONE..........It's there, but it isn't there in win. I've had to reload win98 when this happens and it can happen on any win9X/NT40 system out there. The point of windoze is to be easy enough for my grandma to use. (I gave her a box with win on it, and she still doesn't get it) The software needs to be replaced. What they did is take old Scraps of code and patched them together in a nice GUI pattern. And Coward talks about linux FUD? Has he ever heard about DR dos or OS2? That's just plain hippocritical.
Don't call my crazy, that's what they called me back in the home!
Just get a slow 486 Stick in ten 36BG scsi discs... Boot the machine Do something that access all partitions on these drives and press 'RESET' Now, it's time to wait! /AC
...'nuff said...
One of my Network Engineer friends told me this story.
A large insurance company with offices in Chicago had outsourced their IT dept. One day the said Engineer got a call from the dallas branch wondering why it was taking so long for the workstations to boot (2 hours!). He was able to trace the problem down to the fact that for some insane (and unknown to him) reason the dallas PCs were authenticating against the Chicago servers.
War is necrophilia.
Okay, explain to me how this makes Windows NT a bad OS ? (Apart from the obvious). Of course if you jack up and load the fucker with lots of apps, it'll take forever to reboot, the same with linux and all other os's.
/etc/hosts name.
Hell, do you even realize how long it takes to fsck and parallelize drives upward of 20gig on a linux box ???. And if the name server is down, sendmail takes forever until it times out to take the
Like I said, stupid, retarded, and it only serves as to satisfy all those rabid NT-haters out there.
One could easily have the login authenticated accross an artifically slow link creating very long login times. I think they should take this out of the contest.
Then again, this isn't scientific, and is good clean FUD, so sure, whatever!
-AP
The last company I worked out had a dual 400 PII with 512 megs of ram... it took sooooo long to reboot, that we usually just powered the damn thing off--screw drive intergrity! I remember waiting over 3 hours once!!! hahahaha... our clients would call for hours bitching us out! I'm going to shoot and email out to the sysadmin and have him time it. I doubt he will be willing to sit in front of it to get an accurate time though! hahahaha... I think I'm goin to bust a nut from laughing! peace!
NT always boots faster than Unix, especially when the Unix box has to fsck the disks. NTFS is great.
Not so, my man. This is only measuring one thing: a contest with slow NT reboot times.
:)
If you want a contest for slow server boot times, so be it. Put up a webpage and ask for submissions.
My linux box is running quite a few services, and it isn't a monster like some of the hardware for some of those NT boxes, and it boots up quite nicely, thank you.
fsck can take forever. Fortunately it doesn't always run. However, NT checking partitions takes a while too. Sendmail doesn't take *that* long to timeout IMO, but I could be wrong here.
A 50 minute boot time sounds like Exchange is a bad *app* to me, though. I wonder what it does.
What do you think creates rabid NT-haters? NT! It's DOS, VMS, and Windows, all rolled into one! How monolithic and anti-UNIX can you get? It doesn't tell you anything, it's slow, it doesn't make decent-sized partitions, it has weird driver support, it doesn't support a lot of windows extensions from the past 4 years, it's insecure, it's bloated, and it's generally annoying. Therefore, the occasional anti-NT page is entertainment.
If you can write a good, sensible anti-UNIX or anti-whatever page, and get it posted, go right ahead. Start with X, it annoys me sometimes. Even if it does have all the features Windows Terminal Server wishes it has.
---
pb Reply rather than vaguely moderate me.
pb Reply or e-mail; don't vaguely moderate.
Heh, heh....
I mean, in normal operations you only have to reboot maybe once a year or less. Oh, sorry, this was NT. In that case I guess you really need to tune those reboottimes in case you have to install some critical OS-patched like... Internet explorer for example. :-)
There is a known problem with Microsoft Exchange 5.0 (and 5.5) that in certain circumstances can cause an NT machine to take a LONG time to reboot (approx. 1-2 hours). I believe that a machine with Exchange will be the winner because of this problem (which I spent days with MS trying to track down this problem).
This problem is apparently due to a service deadlock (i.e. the service does not respond to the kill command). There is a registry entry (what a shock!) that dictates how long NT will wait for a service to shutdown. Of course, this is a PER service command. If you have a deadlocked service subsystem (as can occur in certain circumstances with Exchange), it will take approximately * to bring the system down. I believe the default is 5 minutes. Changing it to ten or twenty seconds does a lot of good. =)
Looking at the website, the leader with 49 min has Exchange running on it. Maybe we should notify him of this registry entry... Nah...
Later,
Justin
Mu. P.S. The address you see is real. =)
I think that's the slowest processor that will run NT, and I just happen to have one sitting at work. I'll just pop that into the Compaq 486 w/24Megs of RAM, and we'll see how long it takes to INSTALL, let alone go through reboots... :-)
Thank God it's Friday, and I can actually slack off and do this all day!
...Its no problem to make a server running any OS rebooting really slow if you try.
Now, the contest ought to be for the slowest rebooting time on a machine set up by some certified MS whatever, in a place where the downtime really hurts. Like in a cold storage. "Oh sorry, you will just have to leave the 10 tons of butter out here in the sun. Computer problems you know..." (true story)
All opinions are my own - until criticized
Aside from the obvious humor of the whole contest, I do have one quick question about the actual rules of the contest, and the mini-FAQ didn't answer me. Is the object of the contest just to time a standard reboot, or to manipulate the server machine in such a way that it causes NT to take forever to boot? If it is the latter I would tend to find that somewhat unfair to NT. I'm not a fan of MS products, but I think NT performance should not be judged based on someone who sets it up on a 486 with minimal ram and loads the box with huge apps either. Since there really isn't a minimum requirement I bet we'll have people unearthing their old boxen from the closet and firing them up as lil NT servers. Nothing wrong with using it in such a competition, but I hope NerdPerfect notes the system specs on the "winning" box. I don't think they should go as far to require that people follow the standard system requirements listed for NT and run a regular fulltime server, but at least any tweaking with the boxen to help the contest should be detailed. Win NT may be inferior to a Linux box, but I'm sure you could give a *nix box crappy performance if you put heavy effort into making it as inefficient as possible. Maybe NerdPerfect could sponser a similar contest for Linux servers or whatnot, and challenge people to downgrade performance to the max. I do think that Win NT sucks, but it would be a bit obsessive to characterize its performance based on such a contest.
masochistic.
Just another reason I'll stick to Solaris, thankyouverymuch. UFS logging being a wonderful thing, especially when you have a 60GB RAID-5 volume.
--
--
One thing I don't understand is how these times are authenticated. There's no provision for it on the submission form. You just fill in your name and machine, and enter a time. You can just make it up.
-- "Sponges grow in the ocean. I wonder how much deeper the ocean would be if that didn't happen."
49 minutes, 13 seconds
Submitted by Sylvain Gregoire on Thursday, October 28, 1999 - 6:06 PM
Windows NT Server 4.0 SP5
256MB of RAM
1 x PII 400
Main applications: MS Exchange 5.5 (150 users)
This is a load of bollocks, and can't possibly be true. If his server takes that long to load, then the guy has more serious problems (like with his admin skill).
Man, some people will find any way to dig up stupid people from their sleep, why can't we leave them rest in peace?
By the way, "parallelizing" doesn't take any time at all. In fact, when fsck "parallelizes", it checks more than one drive at once. If you've got four drives, it'll check all four at once instead of one at a time.
- A.P.
--
"One World, one Web, one Program" - Microsoft promotional ad
"Remember when the U.S. had a drug problem, and then we declared a War On Drugs, and now you can't buy drugs anymore?"
To begin with, *lart* to slashdot editors for not classifying this under the right category - should've been "it's funny, laugh".
first - note this is more like a survey, where anyone tells their own story. how much science does that hold, i don't know. likewise i could tell my story of how i bluescreened my linux (yes, i did - a blue ascii screen displayed with 3 black columns, with OS lockup) and it would have no scientific value whatsoever.
furthermore - when your target (what counts as better) is making things worse, you just need to take slower computers and load more programs. hardly a challange. i could make my NT box try to spawn 1,000,000,000 processes in my login and watch it crumble. what i'm saying is, that there's not much challange in such a thing.
all in all this makes a funny interlude, but hardly something of any scientific value.
speaking of btw, linux will sometimes not reboot on itself at all. is anyone else familiar with the screen 'Incosistancy check: run fsck MANUALLY. give root password for maintanence or press Ctrl-D to reboot'? i've seen this one quite a few times.
is that all of Micro$oft's problems started when someone passed around an internal memo starting an obfuscated code contest with the stipulation that you had to use production code, unfortunately they didn't mention the date when this contest would end
I'm not a big fan of NT by any means, but this is just silly. As many have said, it would be easy to rig _any_ system so that it booted slowly.
What I'm curious about is if anyone has hard numbers comparing NT/Linux/BSDs/etc. boot times on systems that are running similar configurations. It's fun to bash Microsoft---hell, I enjoy it---but I'd prefer to see what the actual state of affairs is.
-zook
Last I heard, Microsoft claimed a 99.9% service availability on NT servers.
How does this add up? 99.9% and one reboot per day means that a reboot can only take 86.4 seconds...
Opinions expressed above are mine, and not my employees'.
49 minutes is pretty impressive. When I was an undergrad, our VAX 6110 running VMS 4.7 would typically take 30-40 minutes to boot. I guess that Cutler put more than a little bit of VMS into NT after all....
Well - this server is running Exchange 5.5 and here is the reason it can take that long:
There are various services running under Exchange and they depend on each other. This means that you have to shut them down in a certain order. NT doesn't know the order however, so it has to play a game trying to find out what the correct order is. When it shuts down a service it sends a terminate signal and then waits for the process to terminate or for the request to timeout. Eventually it will give up and ask another service to shutdown first and returns to the original service at a later date.
I've seen this done:
Manually shutdown Exchange services in the correct order.
Ask NT to reboot, 10 seconds later the machine already in the bootup sequence.
I hope this doesn't appear as defence of Microsoft products, because it isn't. NT process handling should be able to query processes what other processes are preventing the process to terminate or Exchange should have it's own termination sequence.
What's the point of this? To show that NT takes a long time to boot up? To make NT look bad?
It doesn't do either. The organisers come across as being incredibly immature, and the existence of such a lame contest is hardly a credit to the Linux community.
Silica is the `third worst' then.
...is a really lame excuse. Isn't one of the primary tentant that Windows is supposed to be easy to use? I have seen admins with credentials up their arm, screw something up and have to reinstall from scratch. Us "windows developers" pratically have to do it every day.
In theory, one could solve all your problems by knowing how every thing ticks, but most of time we work with a lack of information.
I'll tell you what - you release Windows in source code form and I will document it the way it should be - completely. No more obscure registry switches, no more half brain head services. I would love to know why SCM does the brain head things it does.
What do you think creates rabid NT-haters? NT! It's DOS, VMS, and Windows, all rolled into one! How monolithic and anti-UNIX can you get?
err, its not UNIX!? Now, supposedly the internals have a striking similarity to VMS, and it shared a major architect, but early on in both cases. Im not aware of any DOS bits in there. a point?
It doesnt tell you anything? How about all the crap I get in the event viewer app everytime something unexpected happens with my mouse? Doesnt make decent sized partitions? Sure 4 GB (IIRC) is not quite up to everyone's desires, but there's worse out there. BTW, anyone know if you can stripe 4GB NTFS filesystems to create a larger one? Ok, enough of this junk.
An anti-anything page can be amusing at times.
It's not monolithic - it's a microkernel with a HAL.
:)
Everything else you said stands
(Don't diss VMS - it's great. And has a real journalling file system, and proper clustering too)
--
Peter
the prize is a _t-shirt_!
When Banyan started supporting the Macintosh, the first release of the OS with Mac filesystem support was really flakey.
They added a separate FS layer after the fsck that checked additional filesystem structures. On a server with no macintosh files, this could take 2 hours on a 486 with approx 6GB of data. I came across a server hosting some macintosh services that took all day after an enforced reboot to come back up. This process could not be easily circumvented in those early days.
I can't recall anyone using Macintosh support on these servers for very long, most of the systems I supported were ripped out after 6 months.
If I set my Linux box next to one of my Win'98 systems, I know that my Linux system will take a lot longer to boot.
Some of this is due to some sysadmin problems that I can never find the time to fix, but it could also be down to the fact that my Linux box has:
* 8 partitions (25MB) to fsck,
* a dial-up Internet connection,
* DNS
* Apache
* SAMBA,
* DHCP
* sendmail, innd
* lots more.
Having a slow boot up time is normally due to either a bad setup , or due to the fact as system has a lot to do. Of course it is easier to find the former in Linux, as you can look at the startup scripting.
What could be done to speed up Linux boot ups ?
Write scripts in Perl instead of bash, perhaps ?
Donte Alistair Anderson Roberts - hi son!
Karma: Chameleon
When I was at DevDays 98, the Visual C++ guy commented that Version 6 could compile the NT Kernel in *just* 4 hours (compared to 8 hours with Version 5)
i can go along with that...
i only run linux and i i`m really not fond of MS - but whats the point of doing a "longest NT boot time" competition? If some people running NT experience an hour boot time - bad luck for them. If some people try to think of ways to slow down the whole stuff - looks to me as if they got way to much time to waste:) but it doesn't say *anything* about NT, does it? As quite a lot of former posts showed before linux can take hours as well, just think of the NFS mounting time-out...
It's all a quite stupid way of showing how to mess with NT, really nothing the OS community can be proud of.
wtf are some of you on? did you READ the article? I see people bitching about how long it would take linux to fsck big drives? the article says from the time you say "YES" as in a proper goddamn shutdown, so quit bitching about fsck'ing the drives, cuz it wont if you shut it down properly.
end rant
>"No Cheating, Production Servers Only" Jeez, this comment got scored up to "4" and it violates the rules of the article. Someone should write a paper on why stuff gets moderated here.
Don't be an idiot. Well, at least no more than you can help. Amoeba wasn't violating any rules by posting his observation on Slashdot. In fact, Amoeba has identified a genuine problem with the contest methodology, which I think is very constructive and - as the moderators seem to agree - insightful.
There was a bug in the Spooler service. It wasn't apparent until Exchange 5.x was installed on the system. The spooler would not shutdown because it was waiting for a connection that never came. Installing Exchange modifies the global shutdown WaitToKillService timeout to ~15 minutes. This was done to allow Exchange time to shutdown its databases without fear of corruption for being killed. If you install SP4 or greater, this is fixed. Long Reboots is no more funny than PCWeek not installing the security patches for Red Hat and calling it a fair security test. The only good thing that should be taken away from this is that all admins need to stay on top of the latest patches for their platform.
Ever tried running the system info tools? You can get graphs of everything from vm swaps to cpu util in user vs protected mode to packet loss to ... well you get the picture. That app is the one thing I really miss about NT...
it doesn't make decent-sized partitions
Huh?? NTFS supports 16exabyte partitions. You have a device bigger than that??
it has weird driver support
Weird in that there are drivers shipped in the box with virtually any hardware device you can buy (except of course some video capture products and parallel port devices which I agree is a bit crap) and you don't have to write them yourself? :-)
it doesn't support a lot of windows extensions from the past 4 years
It doesn't support DirectX over version 3. Period. And Win2K will fix that...but hey only Linux people are allowed to use the phrase "that's in the next release" aren't they?
it's insecure, it's bloated, and it's generally annoying.
And I can't be bothered any more...
Seriously I'm not having a dig at the linux/solaris/be/*bsd/amigaOS/whateverOS guys & gals out there. Just don't fight FUD with FUD...it only lowers yourself to their level. And you really don't want that. NT's not too bad if you know you're way around it, and are aware of it's limitations. Different OS's are good for different things. Nothing is perfect....chill :-)
---- Den ene knappen er powerknapp, den andre er Bender voice knapp "Bite My Shiny Metal Ass"
You'd be surprised.. I have a system that fsck's 6 8.5 gig drives in parallel in under 60 seconds. (I think it is about that timeframe, it has been 65 days of uptime since it last booted.) Running a single fsck, on one drive (95% full) takes 35 seconds real time (as just measured)
:)
But lets put it this way, that system takes less time to boot up with an fsck than my personal machine which has one 90% full 5gb partitian formatted with the defaults among its drives.
Those 6 drives are intended for big datafiles, so I make the blocksize 4kb as compared to the default 1kb. This means 1/4 the number if indirect blocks. Also, since its big files, there aren't a lot of directories so it doesn't have to scan through them. Those directories that there are only require 1/4 the time to read because they are 1/4 the number of blocks. (I assume that sequential reads are relatively free because seek time should be the main cost.)
Then, I formatted it at one inode per 128kb, as compared to the default of one inode for every 4kb. This drops the number of inodes fsck has to read off the disk by a factor of 30 to only 69632 (compared to >2 million in the default format). (FSCK *has* to read off each inode, to check if the file is orphaned.) I think that this here is the main speed improvement. It also frees an extra 120mb of drivespace. (drivespace that an fsck would have to read in.)
Remember, since you can never create new inodes, so the default format always starts you with a large amount of them, usually an insane number of extra ones.
What I'm saying is to format your drives appropriately for their planned usage. FSCK's don't always have to be that painful.
If you need that huge quantity of extra inodes and also have to save the data, than you are screwed. If you don't need all the inodes, then formatting a filesystem without them is a lot more effecient. And if you don't need to save the data (say, a cache server), then it may be faster to just reformat it upon each boot.
If none of the above apply, then yes, with the current version of ext2, it is very painful to fsck a 20gb filesystem that has 3 million files on it.
Actually, go to his site and click around for a while.. You'll find a funny article about how M$ wanted to charge him for tech support on bugs listed in their tech base.
Lame, stupid, retarded, whatever, it's funny. Eat some fiber, or something. Go make your own web site, and post a lot of articles about why Linux sucks. Run a 'longest time to port linux to my backhoe' contest. Join the *BSnDwagon. Read about journalled filesystems for the linux. Etc.
--
Blue
i browse at -1 because they're funnier than you are.
My linux box takes about 10 minutes, spending most of its boot time mounting a big 20Gb ext2 filesystem.
The NT box has the same amount of HD space, but on more filesystems (about 10*2Gb FAT16). The hardware is almost the same, K6-3/400 256Mb for the NT, and K6-2/300 64Mb for linux.
A couple of companies ago, we needed to test some PCI hardware via emulation. The emulation of our board ran at about 1% of real speed (which is actually pretty good for that kind of thing) so we needed PCs in which the other components - CPUs, chipsets, etc. - also ran that slowly. We found a couple of companies that specialize in making the slowest PCs for just this purpose, and they compete very aggressively. "We're slower than they are! You should buy ours!" It was rather amusing.
;-)
Less amusing was the fact that they couldn't actually deliver the kinds of systems we wanted (450GX) at a low enough speed. It turns out that trying to turn _down_ the clock involves many of the same challenges as turning it up, and some of the components didn't take well to the low clock speeds. "We couldn't make it slow enough to satisfy the customer."
Yes, I know this would violate the rules of the contest. I also know that I'm a little off-topic, but hopefully this is also funny enough that I'll get to keep my karma points.
Slashdot - News for Herds. Stuff that Splatters.
These are very good tips. (In particular, I had missed the "1 inode per 4K bytes" bit in my brief encounter with the mke2fs manpage, but indeed that's the default.)
BTW, for the curious, the magic incantation that produces the above set of parameters is like so: mke2fs -b 4096 -i 131072 /dev/ whatever
I have a 17Gb hard-drive that I use for CD-ROM images for when I build music-compilation CDRs. These tips would work great on this HD, particularly for the odd case when I have to fsck. Note that a 4Kb cluster size should also speed up most other operations on large files, particularly those which require synchronous metadata updates, such as rm's.
(Note: I've only had one Linux kernel crash that I can remember since getting this drive, due to some strange interaction between burning a CD, playing a WAV file and insmod'ing the floppy driver from kerneld on a kernel that would have preferred kmod. The rest have been stupids like bumping the power switch, having a UPS go bad, etc.)
--Joe--
Program Intellivision!
Personally, I'd be more interested in how fast an NT system could boot. My NT Server (P200, 80 Mb RAM, 4,5 Gb FastSCSI) takes about a minute after the BIOS stuff, which is faster than when I had FreeBSD installed. But then again, I used a P133 and a slower SCSI drive back then.
- Peter
REM ha.. I win!
pause
One possible reason for such a competition is to actively seek out poor performing systems and use them to try to understand how best to fix the operating system. For example, in the Linux world, people now complain about long fsck times on huge filesystems, so the Linux community responds by developing journalled filesystems. (Note that in the early days, filesystems weren't huge enough to make fsck times an issue for most users. Nowadays, even typical home-user filesystems are large enough to cause a minor headache. This sort of model -- fix what the users say is broken now, even if it wasn't broken before -- is a reasonably good model.)
I'm not saying that Microsoft is behind this contest, but it would be an interesting (and good) if they were. This is an effective way to find out where a product is not doing as well as it should and can be improved, and it's something I think every software vendor should do.
Other industries do this. My dad used to work at the GM plant where they build the 3800 Series II V6 Engine. (This is the power plant that comes in the 3.8L Grand Prix, Bonneville, and some other cars. I personally own a supercharged Grand Prix GTP.) The 3800 apparently has a legendary track record as one of the most stable engines GM has produced in recent memory. Nonetheless, they purchased a used 3800 engine from a cab company that had somewhere around 300K miles on it, to do failure analysis on it so that they could make their engines even better. The disassembled engine is now on proud display at the plant as a testament to how good of an engine it was.
What would be interesting is if Microsoft did the same: "Show us your worst performing servers, and as a prize, we'll exchange your servers for new ones and use your old ones to do performance analysis so this problem doesn't happen again."
--Joe--
Program Intellivision!
I think I'll start a contest.. "Longest time for a Linux newbie to figure out how to change the screen resolution in Gnome" I guarantee it'll be weeks....
surely the world's slowest NT server will also be, by definition, the world's slowest Linux, FreeBSD, OS/2, Netware, whatever server?
Out production server regularly gets so f****ed up it won't reboot. Just try and shut it down - it never does. So, do I win?
Also, does it count if you're doing a chkdsk on 3 200GB Raids during boot-up?
First, make it work, then make it right, then make it fast, then, make it bloated!
Im not aware of any DOS bits in there. a point?
..... Yea no dos in there at all :)
Mmmm, c:\
--fatboy
Firstly, I never said linux sucks. I use linux a lot and do stuff like pack up 486 linux boxes with daemons and services up the ass. And as a testament to linux stability and efficency, it runs great.
/. have pretty much degraded into insane linux/open source-zealotry, an intense amd-zealotry.
I'm just saying a FUD campaign on NT is stupid and retarded. I won't complain if its on your homepage, but a so called "News For Nerds, Stuff that Matters" news website should not post it.
Also, stories on
Just cuz i like linux doesn't mean i have the right to cram it down other people throat's and tell htem that its good for them.
The time taken for *any* OS to boot is much less dependant on the OS than on the overall configuration of the machine - CPU platform, Hardware config, application config etc.
:)
/etc/vfstab entries to mount the array volumes then starup time came down to around the same as my 2 Dual 300Mhz CPU 512Mb Ram Compaq NT servers. One as an Application server, one running NT4 TS edition with Metaframe. Both would boot in under 4 mins including login. They just did it a lot more frequently than the Suns :)
A vanilla NT, Linux or other *nix system will all take about the same time to boot on equivalent hardware. OK one may be slightly faster than another but we're talking seconds here.
The slowdown comes when you start adding to the system startup sequence with additional hardware checks, file system checks and application startups.
For two years I ran a couple of Solaris boxes with 512MB ram dual 300Mhz Sparc CPU's and 550Gb of RAID 5 storage each. Start up from a clean shut down on these took 30Mins. Rebooting from a crash or other dirty shutdown could take over an hour. UFS file systems checks and/or FSCK'ing a crashed array takes time. You dont even want to think about how long it takes to resync a raid volume of that size
If I removed the
What I'm trying to say is you CAN ( and should ) blame NT for crashing, but you shouldn't blame it for slow boot times. Get your software config right, run it on a decent hardware platform and make allowances for what your asking it to do.
A bad workman blames his tools. If you dont build it right, it wont run right.
My Win2k server box takes 45 seconds to restart. Specs: 1xPIII500, 192MB RAM, IIS 5, 180GB RAID, SQL Sever 7. Anyone can fluff an OS to take an age to boot up-- how about running a compo with the _fastest_ bootup times?
Of course the NerdPerfect contest is more ridiculous than a SlashDot poll.
How abot counting the number of reboots you need to do in order to install a working server with a certain set of functionality from clean HD?
Say, a mail server, an FTP server and a DHCP server? Any estimates for NT?
Remember that we are talking about the OS that needs to be rebooted about 5 times just to get a basic logon at all (NT 4.0 + Service Pack).
I've run it on a 486sx-25 with 16MB of RAM. It's butt-slow, but runs. :)
I have the same thing with crossmounts between a couple of machines. Situations where you need to reboot both simultaniously are rare though.
What you can do is be a good administrator and change your init scripts. eg make sure you have a scripts to forcibly unmount your nfs mounts before doing unmount -f.
jeez... you have complete control of how the machine boots/shutsdown cause it all done via scripts. so use that power..
I use Friend/Foe + mod-point modifiers as a karma/reputation system.
Unix box reboot times are irrelevant, really, because THEY VERY SELDOM HAPPEN. This contest is restricted to Production Servers. While some folks out there may like to reboot their desktop Unix machines on a daily basis and drum their fingers waiting while all the bootup messages scroll by, there is no excuse for constant reboots of a Production Server. Unless the server runs Microsoft NT, in which case frequent reboots are a necessity.
Run lsof, notice the problem, sync ; init 6
------ Curiosity killed the cat. {satisfaction brought it back | it didn't die ignorant | lack of it is killing mankind
Read the contest rules...it has to be a Real Machine doing Real Work, not some boat-anchor in the closet.
How exactly would you compare NT kernel compile times? I guess if you're a MS employee or one of those university researchers that get to see the NT source you could do it, but something like that is really out of the reach of most of us.
B through D model AS/400's would typically take 15-30 minutes normally even if nothing was wrong. Moreover because of the SLS architecture of the machine eg, the entire machine is a flat address space where every file, device and everything else is a memory address in an underlying database, an OS upgrade would take up to 17 hours.
...it's been going on for many years. The fastest boot sequence of any reasonably general purpose OS I've encountered was on Univac 494/Omega that I worked on back in the early 70s (when it was already nearing the end of its lifetime): given suitable peripherals it could boot and have the online workload running again in under 10 seconds - because it had been designed for a target market which could not tolerate extended outages. And because the machine had a maximum of 262Kwords of memory, and ran at only about 1.2Mips, there was a premium on careful programming that used resources efficiently. Even then, though, it was clear that the trend was towards the cost of implementation and maintainance of application software exceeding that of the raw processing power, so that it became increasingly cost-effective to use less expert implementors and to provide them with tools that traded implementation time against run-time performance. And that is a perfectly reasonable and defensible way of proceeding, provided that the various costs are taken fully into account.
:-(
However, there's a big difference between making a considered guestimate of the price of processing power over the likely lifecyle of a project and using these resources accordingly, and making the mistake of thinking that "cheap" resources are "free", and placing no constraints on their use (or, equivalently, blindly assuming that available systems will be fast enough to run the product by the time it is released). In the old days, if you did this you were liable to fail to meet contractual performance criteria, so there was a fairly direct feedback mechanism that kept product performance reasonably compatible with the capabilities of the platforms on which the products ran. More recently, with commodity pricing for software products, this feedback loop has been broken, and the costs of poor implementation practices are now born disproportionally by the customers - whether by being forced to replace systems more frequently, or with outages after major software failures becoming longer, or by finding that the newest products increasingly manifest inconsistent and unpredictable failure modes.
It's interesting, in this regard, that there have been rumours that MS may be considering leasing of its products rather than (or more likely in addition to) the customary "shrink-wrapped" sales. Major corporate users might just be interested in this approach, if it gave them some contractual leverage when the fitness for use of a particular product was less than satisfactory.
Just don't hold your breath
Putting an artificial step in, like making the server work over a slow link, would violate the spirit if not the FAQ of this contest. It's supposed to be for real conditions, not fake ones drummed up just for the contest.
My NT server has been booting for the past 500 years. I want that T-shirt.
Remember when www.windows2000test.com first went public? How long did it take to come back up after the router got "hit by lightning"? Wasn't it almost three days?
Now how do I attach that Monty Python foot to this post...?
--
QDMerge 0.4!
how to invest, a novice's guide
I don't know what this dude's doing, but my exchange server is a mere Pent. Pro 200 w/ 128MBs of RAM running NT SP5, Exchange Server 5.0 SP2 w/ 300 users, and the reboot time is roughly 5 minutes. Homey needs to learn how to configure an Exchange Server!!!
At my old job, one of my clients had a SBS that would take more than a half hour to shutdown. There was a patch released for the SBS though. It allows other services to begin shutting down, instead of waiting for Exchange to stop first. Still takes a while.
How does this make NT a bad OS? It points the finger at how NT manages resources. If you have a fast computer with lots of memory, NT probably won't have too many troubles rebooting (disregarding buggy software). But then again, no serious operating system is going to have troubles here.
On the otherhand, in an environment with less memory, NT is going to crawl. Your harddrive led will be solid and your box will sound like a coffee grinder. Some other operating systems happen to handle low memory situations a little better. This is probably due a lot to the fact that you don't have to boot into a gui.
The serious part of the linux community welcomes "benchmarks" like mindcraft because they expose areas of the os that need improvement. In the same way, NT hardcores should embrace the same kinds of criticisms because in the end it will improve the os. Who knows...maybe Win2K will rock.
Yes it should have had a 'funny' rating but..
On the other hand NT won't support that large of a boot partition (4g if I remember correctly) and getting it to recognize big IDE (>8gig) is a real pain in the ass during install. Once you get it installed though just make a big ass data partition. This backfires though. A lot of apps want to do crap in the \winnt directory and the boot drive. I've been screwed by running out of space on that primary partition.
I imagine that the winner of this contest will have a box that doesn't reboot. Its not hard to belive that there are things that can (and have) happened that keep a machine from shuting down. I have personaly seen shutdown times >45 mins.
BTW: I miss that perfmon tool too and the original poster is correct about lack of support for win98 things in NT. USB is the one that pisses me off the most.
Hmm.. At work there is a 4 cpu HP K580 system. It has 3 gig of RAM and fastboot enabled, which is supposed to speed up the process, however, it still takes about 1/2 hour to boot. I've also seen a horribly misconfigured tandem box (running SII software) that took 2 hours from power on till all the systems (partitions/zones whatever you call them) were up.
What is funny about this type of thing is that IE5 actually INCREASES the reboot time on all machines, becouse i automagically loads all of the IE DLL's into memory. Then they proceed to do ads stating how fast IE is.. Well sure, they preinit everything at startup, and your time is moved from when you click to when you turn on the machine. Not to mention the simply swap issue under low memory conditions..
-- I'm the root of all that's evil, but you can call me cookie..
I have heard of a machine that took six DAYS to boot. Running a netscape server (I think a news server product?). Lots of files (millions?) on a very big, very slow drive. took six days to finish running chkdsk when it restarted.
You must be kidding! I, as well as my cube mates(and countless other NT users), reboot at least once a week. I'm not doing any compiling locally either, just telnet'ing to solaris boxes and using word(and some occasional web surfing, like now.)
Technically the MS "requirements" are 486dx66, 24mb of RAM. So don't blame them if it's shitty =]
I can't say I like Impalas. Until the recent redesign, they just looked like Caprice's w/ wide tires. Big, slow, ugly.
Now the Grand Prix.... :-) Or if you want something a bit bigger, the new Bonneville....
--Joe--
Program Intellivision!
:) I think the replies this thread got cover most of my points, and it's too late now, but...
.DLL library clobbering problems, for Windows' sake. And only Linux people are allowed to use the phrase "check out the devel kernel with these patches if you really need it now." ;)
;)
1) You turn it on. What do you get? A blue screen. Any boot info? Not really... That's what I mean by "it doesn't tell you anything". It's about as helpful when it crashes, too.
2) In my experience, if I take a new machine and try to install NT, I can't make an NTFS partition take up the whole disk. Maybe it's just a limitation on IDE drives, but dude, it sucked. FAT16 is worse. 16 exabyte? Has anyone tried that? Jeez.
3) Driver support: try getting a modern video card to work, say. I had an ATI card that would run under SP3 but not SP4 or SP5 at all. Go figure. There are drivers in the box for virtually any hardware device on the NT hardware compatibility list. Past that, you're confusing it with Win98 Release 2. And you *can't* write them yourself, unless you have $$$....
4) It doesn't support FAT32. It doesn't have the same driver model. It breaks a lot of DOS apps. Sometimes Linux DOSEmu does a better job than the NT DOS VM. I hope NT5 fixes the
5) Ooo, didn't mean to waste your precious time.
6) I'm aware of its limitations. That's what I was outlining. And I felt the need to rant for a while....
---
pb Reply rather than vaguely moderate me.
pb Reply or e-mail; don't vaguely moderate.
Who cares how long it takes to boot if it's less than a day? The real test is how long it can stay up without having to be brought down.
Of course, I guess this is an issue with NT as with most software installs, you end up having to reboot anyway.