PDA

View Full Version : Computer Hangs After 49.7 Days



sulekha
September 13th, 2010, 03:39 PM
Hi all ,

See this article :- http://support.microsoft.com/kb/216641

It says that win 95/98 machines wont work for more than 49.7 days continously. my question is , following the lines of above article , does linux distros (ubuntu in particular) have any such issues ?

endotherm
September 13th, 2010, 03:45 PM
I certianly would hope not. many linux servers can run for months and years without reboot, if correctly administered.

CharlesA
September 13th, 2010, 03:47 PM
I certianly would hope not. many linux servers can run for months and years without reboot, if correctly administered.
This.

I am curious why you'd worry about something from an OS that is like 15+ years old and obviously obsolete.

endotherm
September 13th, 2010, 04:58 PM
btw, thanks for the lolz. that support doc gave me the best laugh I'm gonna get today. makes me think of what happened to my 386 (WFW311) at midnight 1/1/2000. the date jumped back to 1986, and the keyboard got all wonky.

perspectoff
September 13th, 2010, 04:59 PM
Hi all ,

See this article :- http://support.microsoft.com/kb/216641

It says that win 95/98 machines wont work for more than 49.7 days continously. my question is , following the lines of above article , does linux distros (ubuntu in particular) have any such issues ?

I have had a Linux DNS server (for our small hospital) running for 7 years without needing a reboot. It has rebooted automatically after power outages a few times, though, but without needing intervention.

After a Windows 2003 terminal server crashed after only 3 years, requiring an expensive replacement (Microsoft demanded an upgrade to Server 2008 minimum -- very expensive), we duplicated the DNS server, just in case some hardware would fail.

(I'm actually impressed the DNS server hardware has never failed after 7 years 24/7.)

Elfy
September 13th, 2010, 05:01 PM
moved to the cafe

samalex
September 13th, 2010, 05:13 PM
Yeah, Linux definitely isn't plagued by this problem... At my former employer we had a DEC Alpha running Red Hat Linux 7.2 as our FTP and proxy server, and until we replaced the UPS and had to bring down all servers it had about 4 years of uptime. And even after the scheduled reboot it ran for about another 2 years before it was replaced.

I've had impressive uptimes on my home systems as well, but generally due to power outages I'm good to get a few months if that between reboots.

Sam

Xianath
September 14th, 2010, 07:20 AM
I've always found it ironic that it took them several years to find that particular bug :D

handy
September 14th, 2010, 07:29 AM
Reboot on day 49!

scottuss
September 14th, 2010, 07:32 AM
Even modern versions of Windows don't suffer from this (they just suffer from everything else!)

Don't worry about it, you'll have years of uptime with Linux

s0rc3r3r
September 14th, 2010, 07:42 AM
Even modern versions of Windows don't suffer from this (they just suffer from everything else!)

Don't worry about it, you'll have years of uptime with Linux

aye aye!!:popcorn:

kamaboko
September 14th, 2010, 08:16 AM
Ridiculous. I've had Win95 machines set up to deliver services that went for up to five years w/o a reboot. They were setup, put in a server room closet and generally forgotten about.

V for Vincent
September 14th, 2010, 10:01 AM
Ridiculous. I've had Win95 machines set up to deliver services that went for up to five years w/o a reboot. They were setup, put in a server room closet and generally forgotten about.

It's on microsoft.com and it's been confirmed, so it must affect some versions of windows. Not saying that's worth slagging MS, though. After all, 9x is ancient.

m4tic
September 14th, 2010, 10:29 AM
I once went for 73 days on a OpenSuse desktop installation. I shut it down once i saw my electricity for the two months its been on.

Khakilang
September 14th, 2010, 10:31 AM
I don't remember my computer hang at all using Window 95/98. Hardware failure yes but that does not involve Windows. At present my Linux didn't hang either. But I am curious to know whether it happen on the latest Window OS.

roddie
September 14th, 2010, 10:33 AM
I once went for 73 days on a OpenSuse desktop installation. I shut it down once i saw my electricity for the two months its been on.
:lolflag:

Grenage
September 14th, 2010, 10:34 AM
I am curious to know whether it happen on the latest Window OS.

Certainly not. :)

julio_cortez
September 14th, 2010, 10:41 AM
It says that win 95/98 machines wont work for more than 49.7 days continously.Well, they still are rock solid compared to the one that came after: ME barely worked for more than 49 MINUTES in a row!!
..sorry, couldn't resist.. :p :p :p

NMFTM
September 14th, 2010, 11:07 AM
I'm going to be installing ME on my main computer later today. Because it seems like an easier solution than getting them to work on XP or 7.

julio_cortez
September 14th, 2010, 11:30 AM
I'm going to be installing ME on my main computer later todayWell, I wish you good luck, brave fellow poster :D

grobar87
September 14th, 2010, 12:16 PM
Well, they still are rock solid compared to the one that came after: ME barely worked for more than 49 MINUTES in a row!!
..sorry, couldn't resist.. :p :p :p

hahahaha :p:p:p

Johnsie
September 14th, 2010, 02:20 PM
Modern versions of Windows are on par with, if not better than Linux when it comes to uptime. Obviously it depends on your hardware compatibility/setup as well. If linux is running pretty hot on a machine (which it does quite often) then you might need it serviced earlier. Sometimes Linux locks up on some hardware combinations or freezes because a program is using too much ram. I ran a grep command on a server ther other day and it locked up. I've also seen the shoutcast server cause several ubuntu machines to crash. The same can happen with Windows.

blueturtl
September 14th, 2010, 02:33 PM
Windows 95 also had a bug that caused it to BSOD pretty much immediately on boot if your CPU was faster than 350/400 MHz. The Microsoft patch page spesifically says there is no fault on part of the CPU, but that is a timing loop issue in Windows itself.

The uptime cap seems like another example of badly underengineered or just plain botched design and/or programming.

This level of reliability was a joke as much back in 1995 as it is today.
Just something for you all to keep in mind when you say it's ok cause it was such a long time ago.

A certain nerd who started his own operating system and only targeted it for his own 386 initially had higher design standards than Microsoft had when they had already set to take over the world.

kelvin spratt
September 14th, 2010, 02:54 PM
How can windows run for months years without a reboot seeing as it forces you to reboot monthly when it updates with security patches come on boys and girls all this trying to get one over the other. with Linux you have to reboot for Kernel patches so what are you guys saying you never update, Well thats a bad practice. somebody ran a Elive Live cd for i think over a year maybe 2 continual and kept it updated was featured on the old site.

endotherm
September 14th, 2010, 03:00 PM
How can windows run for months years without a reboot seeing as it forces you to reboot monthly when it updates with security patches come on boys and girls all this trying to get one over the other. with Linux you have to reboot for Kernel patches so what are you guys saying you never update, Well thats a bad practice.
agreed, but at the same time it's not quite apples-to-apples either. kernel upgrades can be done without reboot via ksplice if you are running a high uptime server, and due to the lack of a central registry that is partially autogened at boot, most other updates can be done with no reboot (at most a service restart).
additionally, kernel upgrades are often put off for planned downtime. kind of like installing a service pack. you don't just let windows update install it whenever it wants to.

scottuss
September 14th, 2010, 03:01 PM
How can windows run for months years without a reboot seeing as it forces you to reboot monthly when it updates with security patches come on boys and girls all this trying to get one over the other. with Linux you have to reboot for Kernel patches so what are you guys saying you never update, Well thats a bad practice. somebody ran a Elive Live cd for i think over a year maybe 2 continual and kept it updated was featured on the old site.

I would presume they don't include planned downtime, as any and all systems have this at some point.

Grenage
September 14th, 2010, 03:05 PM
I would presume they don't include planned downtime, as any and all systems have this at some point.

Bingo, planned downtime is not generally detracted from total uptime.

CharlesA
September 14th, 2010, 03:07 PM
agreed, but at the same time it's not quite apples-to-apples either. kernel upgrades can be done without reboot via ksplice if you are running a high uptime server, and due to the lack of a central registry that is partially autogened at boot, most other updates can be done with no reboot (at most a service restart).
additionally, kernel upgrades are often put off for planned downtime. kind of like installing a service pack. you don't just let windows update install it whenever it wants to.

+1 to ksplice. I've done security updates and had to reboot after they were installed. If you are running a machine that requires a high amount of uptime, it would be worth it to look into ksplice.

But yah. Planned downtime is a fact of life when servers are concerned. Hardware needs maintenance even if the software doesn't need it all the time. All software needs maintenance as well, not just Windows.

NCLI
September 14th, 2010, 05:34 PM
I have had a Linux DNS server (for our small hospital) running for 7 years without needing a reboot. It has rebooted automatically after power outages a few times, though, but without needing intervention.

After a Windows 2003 terminal server crashed after only 3 years, requiring an expensive replacement (Microsoft demanded an upgrade to Server 2008 minimum -- very expensive), we duplicated the DNS server, just in case some hardware would fail.

(I'm actually impressed the DNS server hardware has never failed after 7 years 24/7.)

How did you make the server turn itself on after outages like that?

V for Vincent
September 14th, 2010, 05:47 PM
This level of reliability was a joke as much back in 1995 as it is today.

Windows is stable nowadays. I don't think I've experienced any lockups so far and I've been using 7 (quite casually, I'll admit) for months.

Maybe you've had less luck on your hardware, but I can't help moralizing here: false claims about windows (or OSX for that matter) do the open source community more harm than good.

Rasa1111
September 14th, 2010, 06:06 PM
false claims about windows (or OSX for that matter) do the open source community more harm than good.

but false claims about Linux by MS people are cool and do allot of good, right. There are MS ppl who talk a whole lot more crap about linux distros than any linux user does about windows. lol

It's just karma man,
Universe gotta stay balanced somehow...

TiBaal89
September 14th, 2010, 06:10 PM
hmmm...

http://i192.photobucket.com/albums/z121/tibaal89/Misc/top_uptime.png

blueturtl
September 14th, 2010, 06:36 PM
Windows is stable nowadays. I don't think I've experienced any lockups so far and I've been using 7 (quite casually, I'll admit) for months.

Maybe you've had less luck on your hardware, but I can't help moralizing here: false claims about windows (or OSX for that matter) do the open source community more harm than good.

Neither of the mentioned issues in my post are false claims.
http://support.microsoft.com/kb/216641

http://news.cnet.com/Windows-95-patch-for-K6-2-chips-now-free/2100-1001_3-218430.html

I'm glad that you've found Windows 7 to be more reliable (it would be terrible if it wasn't!).

tatose
September 20th, 2010, 07:54 PM
I had an eeebox connected to 2 usb JBODs and serving my lans. This machine keeps dying every few days and fails to reboot due to ACPI issues, probably a bug after I had installed win7 home.

Out of rage and frustration of getting ACPI to work correctly, I VMed the whole win7home under vbox (non-ose).

I thought the whole thing will die sooner or later... since the processor is atom and stuff and its not meant for virtualisation.

Never had I knew it went for 4mths of uptime at least, only manually rebooted twice last due 'critical security issues',

just because ubuntu is running on bare-metal instead of windows.