Microsoft .NET Patch May Make PCs Go "Haywire"
yuna49 writes "Various people are reporting that the MS07-040 patch for .NET released on Tuesday can cause a variety of seemingly unrelated problems. According to the SANS Internet Storm Center 'the reports we got so far seem not to lead to any specific thing that happens in many cases, just various things going haywire.' Some commentators on The Register's report of this story indicate that the patch failed to install at all, while others report things like the mouse suddenly failing to work or long periods of hard drive thrashing. In some cases a hard reboot seems to fix the problem, but other reports suggest that a reinstallation of the .NET framework itself is required. The problems may be related to the MSCORSVW.EXE process which recompiles all the .NET assemblies when the patch is downloaded. While the recompilations are supposed to run as a background task, in some instances the recompilation will drive the processor to 100% usage."
That's why my box was running so slow yesterday morning. Drove me and the other IT guys a bit nuts trying to figure it out. Eventually it got better on its own.
Best Slashdot Co
is ALIVE!
And this is why I sit on patches for at least a couple of weeks.
"Declined"
Don't Tread on Me
Okay, I noticed my laptop thrashing away like crazy last night just before I went to bed. One of the offending processes was MSCORSVW.EXE. Since I was tired, I just shut it down and figured I would look into it later. This saves me some research!
It didn't seem to cause a problem on any of my other PCs, though.
...then I'm glad to see others having problems. It tried to install twice, but kept coming up as a pending patch. On the third try, I figured it must be fucked on MS end, and disabled the install and told update to ignore it from now on. *shrug*
Is it just my observation, or are there way too many stupid people in the world?
"If Assumption the mother of all fuck ups, then surely it is also the father of all Microsoft engineering."
When this 100% cpu utilization was happening I called up Process Explorer http://www.microsoft.com/technet/sysinternals/util ities/ProcessExplorer.mspx
in some instances the recompilation will drive the processor to 100% usage
No, kidding ? You mean the background task don't deliberately leave CPU cycles for the sake of increasing idle time ? Amazing.
This kind of summary don't push me hard to RTFA.
I installed this on my Windows 2000 box yesterday and I haven't seen any problems so far. *shrug* Maybe it's just a Win XP thing.
I haven't seen any of the issues mentioned, but after I installed the update my PC failed to wake up after being put in standby mode. Fans and drives powered up, but no signs of intelligent life. This happened the first two times I put it in standby after installing the update and rebooting. Since then I've put it in standby 3-4 times without any problems.
I don't know if it's related or not, but with everything else on the machine working fine, I was suspecting the update before it magically started working again.
others report things like the mouse suddenly failing to work or long periods of hard drive thrashing. In some cases a hard reboot seems to fix the problem, but other reports suggest that a reinstallation of the .NET framework itself is required
;)
Wait; so, random failures, hard drive thrashing, rebooting and/or reinstalling works? Isn't that the normal user experience in Windows anyhow?
DUPE!
Returned Peace Corps IT Volunteer
OK then, so everyone goof up every once in a while, I can't really blame them for that, but when is there a patch for the patch then?
Beware: In C++, your friends can see your privates!
For a minute there, it sounded like Microsoft had moved to Gentoo for their package management... ;)
Disclaimer: I use and like Gentoo, for all its misgivings, so no flames please!
Moderation Total: -1 Troll, +3 Goat
After installing patches on Wednesday I started having a peculiar problem when running vmware 5.4. The host box network connectivity would be lost. It would take a reboot of the machine to reestablish the network connection. I also had one unexpected system reboot when running the arp command while troubleshooting the problem. The problem appears to have gone away once I uninstalled the latest patches. No way to know which one was causing the problem. This was on a Windows XPSP2 box.
While the recompilations are supposed to run as a background task, in some instances the recompilation will drive the processor to 100% usage.
Sounds like Microsoft are Secretly Monopolizing the CPU Without Being Root
They'd better not have nicked my code or they're in trouble. It's GPL 3 I'll have you know...
Summation 2
on a[n] [un]related note, the win2003 server at the company I work for had its partition table corrupted by wednesday morning. Last time the server was alive was midnight tuesday. They can't figure out what happened, other than "some updates were installed".
I rebuilt the partitions with some magical software. Everything seems to be okay at this point. Anyone know if this is related? The only unusual thing that happened to this otherwise "reliable" server were the updates.
A background task that's taking 100% cpu is perfectly fine, so long as it is a background task and is running on a below normal priority.
I frequently make processes that run at 100% CPU run as a background task.
thank God the internet isn't a human right.
yeah no kidding somethings broken, my task manager screensaver used to display usage from 0-100% (Win2k AS)
until those updates and now according to the TM iam using
-152% cpu
So apart from the negative cpu process usage numbers everything else seems fine, oh apart from the installer process needing 450mb! free space to install/compile an 8mb patch
a complete 1.1 & 2.0 net install now needs 1+gb space to run that 300kb widget , bloatware at its finest (soon be catching java on how big a library can get)
You sure it's not related to patch KB935807? On three seperate computers running Vista, I could not get this patch to install. It would try to install, then after a reboot or two, it would report back its status as failing. After doing a quick google search, I soon found out that I'm not alone!
And yes, I've tried downloading the patch file and installing manually. No go.
Life is not for the lazy.
Just wait for the Mono enhancement.
my computer asploded
Yeah, this one seems to of got me as well, or at least I think so. Patches installed by themselves, and Windows did it's reboot thing overnight. When I logged in the next day, one of my USB hard drives wasn't recognized. Now whenever it's plugged in, just AC, not USB, the activity LED goes crazy. Anyone heard anything like this, or any way to rectify.
For me, that particular patch installation failed. Then the windows update service informed me updates were available, including that patch. Let it try again. Failure. To stop the update service from informing me that this broken patch was available for me to try to install, I had to tell it to ignore that particular patch.
Woo, QC.
Come on... We all know the routine here on M$ boxes... Reboot it a couple times until we realize it is shot... Stick in the repair cd so that it can finish the job of killing it... Then wax the whole thing and reinstall... Explain to the user that all their data is gone and when they get that "deer caught in headlights" look, tell them they should have backed up to their data... Hm... Missing anything here??
Just a typical day in windows land...
"While the recompilations are supposed to run as a background task, in some instances the recompilation will drive the processor to 100% usage."
Um, so? If the processor isn't doing anything else, why shouldn't a background recompile use up 100% processor time? Don't tell me Windows gives time to the "idle" process when there are other processes, even background ones, that could run?!?
Why doesn't the gene pool have a life guard?
How you during the installation had to choose a kernel and were given advice on which to choose. It went something like this.
Gentoo-sources is the recommended choice.
Linux.org base kernel is for those who like to live on the edge. (This is gentoo, we REALLY mean the edge)
linux BETA, for those who like to live dangerous.
I al ways thought one was missing.
MS Windows, COME AND GET ME!
If running Beta linux code is risky then running MS windows surely is like playing russian roulette with a pistol. (For those who don't know, you play russian roulette with a revolver. Google the difference)
Oh sure, this may be flamebait but ask yourselve this. What exactly are you paying MS for? Were is all that support and testing that free software just can't provide?
To all the people that have to say something about how this is just typical Windows for you, what exactly are you doing to your poor machines? I work on XP machines all day, and I find that if they're kept up to date with patches, antivirus, and antispyware then this kind of thing is very rare. 95% of my "Windows" problems are caused by 3rd parties like Adobe doing stupid things with their updaters and such.
Quit clicking on every shiny ad you see - you won't get an Xbox for catching a monkey.
Friday 13th, anyone?
How anyone would install an MS patch without first performing some exocism and have a Voodoo priest sacrifice a chicken is beyond me anyway. I have been doing this for years now and so far, no incompatibilities.
Ok, using Ubuntu and Gentoo might have something to do with it, too, but I'm fairly confident of my chicken patching technique.
We used to have a Bill of Rights. Now, with the rights gone, all we have left is the bill.
You know things are getting bad when even supposedly technical types start to use this kind of language. In a few years we can expect to see serious techie-to-techie channels postings saying things like "The CPU went kerblooie" and "The disk became discombobulated" and "Don't apply this patch if you're not a real computer genius..."
"How to Do Nothing," kids activities, back in print!
This is what worked for me. After I installed the patch a few of our .NET 2.0 sites broke with a generic "object reference not set to an instance of an object" error, so I called microsoft tech support. After sending the requested log files I decided to uninstall and reinstall the patch. It worked! It's worth a try...
Interesting and mature response. Sorry, have no points today...
Looks like there's lots of experts from Microsoft on these boards everyday, posting 'mitigating factors' in response to even the slightest criticism.... okay, I've got a real important qn. for you:
Let's say I've got a 100 PCs running XP... every month, on average I think there's about 20MB or so of patches downloaded and installed. That would mean 240MB per year, or 1GB over 4 years. Now suppose a system goes broke, or a new one has just come in. Is there a simple way to install XP ALONG WITH these patches, in one go?
Maybe a separate Ask Slashdot qn. is in order, but I thought I'd rather ask now, seeing as we have another broken patch.
If you keep throwing chairs, one day you'll break windows....
my kids have been having trouble the last few days with wireless connections dying and their mice stopping.
my Debian Laptop is the only computer in the house that hasn't had trouble.
...to do this short of manually "building" an installed running machine with all the patches, and then making a disk image (e.g. with Ghost or similar disk partition image utility). However, creating any such custom installation media of the XP plus all service packs and patches merged together in one install is technically a violation of the EULA unless you are an OEM licensed by MS to create install medias.
I spent all day Wednesday trying to figure out what the bloody hell was going wrong with our server running the helpdesk software (SQL, web based, that sort). Stupid me, I had loaded the patches deemed 'critical' by MS. I unloaded that patch and all was fine. No one believed me when I said it was the culprit. I AM VINDICATED!
That being said, at least the damned thing uninstalled easily.
The patches were applied on 3 of the PCs I use and on 1 Virtual Machine (all running XP PRO) running a combination of .NET framework 1.0, 1.1, and 2.0. I noticed it took a while for .NET to be updated (especially on the VM) but that was it. I did not notice 100% CPU utilization even on my single core, non hyperthreaded office machine which I use for Visual Studio 2005 development. Overall, no problems with the patches here.
If you follow the update KB article, you'll find MS has already found issues with the update.
.NET 2.0, reinstall .NET 2.0 and try to update again. Sounds kind of cyclical to me.
See:
http://support.microsoft.com/kb/928365/
Which leads to:
http://support.microsoft.com/kb/923100/
and
http://support.microsoft.com/kb/934711/
and
http://support.microsoft.com/kb/923101/
and
http://support.microsoft.com/kb/934793/
and
http://support.microsoft.com/kb/931846/
923100 says if you get hosed doing the update, uninstall
machinator omnis sine licentia
David Notarios blog entry
At the end of the day Tuesday, I did the MS autoupdate thing, and when I came in on Wednesday, one of my .NET custom controls was busted in a really odd way.
Specifically, I have this line of code in a custom control.
myPanel.visible = true
A pretty standard line. Worked fine on Tuesday. Now, I set a breakpoint right on that line, and right after I run that line, the visible property of the panel is still false.
Since I hadn't done anything to the control, I was sure that the update had somehow broken it, but since that had never happened to me before, I just kept plugging away. I probably spent 20 hours working on this over the last couple days, but I couldn't find the error. I'm glad to hear that I might not be completely crazy ... it may actually be a problem with the update.
*relief*
burrocrisy
and that would be what? Ruling by jackasses? Never has a slashdot misspelling been more apropos
Since my work runs Microsoft's WSUS server, all the "critical" security patches have been configured to automatically get applied overnight on our XP workstations. (Yeah, I know... this *may* cause me some headaches if a patch breaks things, which it has in the past. But I also figure that in most cases, a patch for the patch is released pretty quickly when MS figures out they screwed one up ... so the issues have been temporary ones so far.)
.NET update patch was reported to break this particular type of thing? Crypkey licensing is used by quite a few software packages.... not just Pronest.
Anyway, we have a specialized program in use on a few PCs here called Pronest 8. (It's a package that calculates the most efficient layout of odd shaped parts for torch cutting from a flat piece of steel. It also has some parts inventory capabilities.) It uses a shared network license on a Windows server, and the "crypkey" service to check the license info.
I know I had this working fine on one of our workstations a few weeks ago, but now, they went to use it today and just receive an "error -24, driver busy" message from the crypkey service - and can't launch the app!
Does anyone know if the
This just in the news: microsoft screws up pcs. WOW!
On some computers updates always fail and I have no clue why. I have never even had the urge to go messing with .net assemblies although I admit I might have killed mscoree a time or two.
.NET crap floating around in versioned folders in the windows.
It uses rediculous amounts of disk space. Tuesdays security updates stole an additional 200+MB. This is after deleting the temp folders in the windows directory. There are some 10 copies of the same
I don't install .NET in the first place, thus (hopefully) not requiring any patches to said .NET. My machine thus remains much, much faster. The instant you install .NET, your machine will be slower. The more .NET applications you install, your machine will be slower. That's my own experience, anyway, on multiple machines. So, yay me! :)
Comment removed based on user account deletion
MS has thousands of employees working on hundreds of evil and greedy pursuits, but isn't it interesting that not even one of them seems to be addressing this serious problem?
Pages on their websites mostly seem to say "THERE'S NOTHING TO SEE HERE, YOUR PROBLEMS ARE ALL IMAGINARY, MOVE ALONG...". If it was MY company, there would be a link to the solution top center on the home page of support.microsoft.com.
Or perhaps MS is intentionally trying to break XP, to get users to shell out for Vista crapware.
You can have my SIG when you pry it from my cold, dead hands.
That's not a web browser (referring to IE), that's The Home Page. And several users a week will call in and say The home page is broke! or The home page is messed up! It's missing my links.
In most cases, they fugged up their IE tool bars.
I hate my users.
I drank what? -- Socrates
Never had my boxes go so slowly. Had to reinstall .net because some of our proprietary software requires it. Quite a headache. Lower TCO my aching ass.
Possibly this is all part of a drive to get people with no technical experience to buy new computers. If you apply patches, Microsoft has control over how fast your computer runs.
For example, Problems with an important Windows component, svchost.exe, can consume up to 100% of CPU time.
On one computer with which I am familiar, the RPC service takes 30%-70% of the CPU time.
I'm not saying Microsoft managers deliberately slow computers. I'm saying that maybe they are not particularly intense about fixing bugs that slow computers.
I'm not the only person who thinks that may be an issue. See this quote from the parent comment: "I've been thinking that MS would come up with something that would make XP less useful - some sort of bug or new type of unpatchable vulnerability to force Windows users to adopt Vista. Maybe this is the beginning of the end of XP."
For a lot of us, using Microsoft software has the feeling of partnering with an enemy.
The person who wrote the parent comment could fix the problem himself. Most people, maybe 99% of Windows XP users, could not. Most people who find that there computer is running very slow will buy another computer. The New York Times article Corrupted PC's Find New Home makes that point.
Has anyone else noticed that working in a IT environment means being perpetually involved in combobulation of the user base.
There is no right to feel safe thru security vaudeville at the expense of everyone's freedom, privacy and tax money.
Note that people have many, many different kinds of problems with Windows Update: Windows Update Discussion Group.
I'm guessing that millions of hours are lost every year because of the sloppy coding in Windows XP. Bill Gates is the Chief of Grief.
Please, don't say "PC" when you're referring to a Windows machine.
PCs can run other systems too, you insensitive clod.
factor 966971: 966971
I noticed severe cpu usage and when I investigated I found MSCORSVW.EXE was the culprit. After determining what it was I just said a few appropriate words, directed at Microsoft, altered the process priority, and went on. Eventually the stupid thing finished. The following day I found out that an acquaintance of mine, whose computer I help maintain, has upwards of ten messages whining about not enough virtual memory on the machine. After allowing the expansion of the virtual memory each time, it eventually went away. So it definitely wasn't a silent install. Though, I have to admit, the forced reboot afterwards annoyed me more. So much for fscking uptime for Windows. Though nothing catastrophic, it is yet another aggravation added to the MS pile.
OR just don't install .Net on your client PCs?
.Net are very remote unless you use it in house.
The odds that any of your client PCs need
My work mates keyboard decided to stop working yesterday, the keyboard would work in the BIOS, it would work to press F8 for safe mode. But the moment Windows booted, at the CTRL+ALT+DEL screen the keyboard wouldn't work! This was for both USB and PS2 keyboards.
So we decided to try a repair install, the keyboard worked through the initial phase (accepting agreement, choosing the disk etc.) when it got to entering the CD-Key the keyboard stopped working again. (By this point I was fighting the urge to install Ubuntu and be done with it)
We fixed it in the same way all Windows problems end up being fixed, we formatted and started again.
It is impossible to comment intelligently on this because no one with intelligence uses Windoze. Therefore Slashdot should ignore Windoze.
the military network environment I work with tends to very carefully evaluate these Microsoft patches before letting them lose on their systems.
Fixed that for you.
In general, it is safe and legal to kill your children. -- POSIX Programmer's Guide
The patch that patches all patches , a service pack2 is needed asap.
But I thought the new scheduler was merged for 2.6.22.
Well some bored slashdotter obviously found a way to let Microsoft know... I just came across this smackdown of this patch on a B0rg Blog.
What's really funny is that it actually sticks to the new Microsoft comments policy!
Right after installing this patch, windows crashed with an "unspecified" error, and then the next time I tried to run FSX, it was magically no longer activated. Not only that, but the Windows Installer is unable to repair, remove, or re-install FSX (or anything else).
So, now I have to reinstall windows, see if I can get someone on the phone in India in the middle of the night to re-activate windows, then do it all over again when I re-install FSX. Great.
Thanks a lot, assholes.
"We don't know exactly what this breaks but it breaks things!" Are you serious?
Now that you mention it, my Call of Duty 2 game went down the toilet around Tuesday. I game off a great weekend... kicking ass and taking names, but now, for the last 4 days I've been getting my ass kicked left and right.
MICROSOFT'S PATCH STOLE MY SKILLZ!!!
(though seriously, it does appear that while settings have stayed the same in the mouse control panels, my mouse sensitivity has changed. Coincidence? Or does Microsoft want me to game on Vista?)