What does this error message mean?

[NetRyder]

Could happen to any program, even Firefox.

Of course not! Haven't you heard? Firefox was created by demi-Gods. It transcends the realm of all errors...

...supposedly.

 

[David_L6]

OK, I'll run the memory test thing....

I was concerned that there may be a spyware program running because sometimes I.E. will slow down or stop (won't open a short cut to a site). When that happens it never fails that I'll find that iexplorer.exe is hogging all the resources when I check in Task Manager. I have all kinds of anti spyware / anti virus stuff that I run once in a while. PestPatrol, Spyhunter, Spybot Search and Destroy, Ad-Aware SE, McAfee, Trend Micro, ...... just about anything you can think of!

I was wondering if maybe I have a RAM stick going (gone?) bad due to the "can't read memory" message.

Here's what I have:

Dell 8300

P4 2.8GHz (Northwood) 800fsb

1.5GB RAM - 2 x 256 that came in the computer plus 2 x 512 that I added. The memory that I added I bought from Crucial.

Leadtek A380 video card (nVidia 5950 Ultra)

Audigy 2 ZS sound card

Thanks for the help and suggestions.

David

 

[xtweaker]

Shakes you rock!

Edit:

If you want to read more on that 'The memory cannot be "read"' Error... check out this link...

http://search.microsoft.com/search/...na=82&qu=(The+memory+could+not+be+"read")&s=1

Find me where it says in any of these articles, that the ram is at fault.

Just some more proof it's not a hardware issue

 

[American Zombie]

If you think IE is messed up you could try IEFix.

 

[xtweaker]

If you think IE is messed up you could try IEFix.

Nice find man! I knew about the IE.INF fix, but that's even better! I'll add that to my arsenal

 

[LordOfLA]

Okay lets get one thing straight. If an applciation or as in this case an extension of Ie was trying to access memort space allocated to another process withour read permission you would get an Access violation not an application error.

Application Errors indicate that the application code itself is corrupted.

THe only way you will corrupt application code is to have a transistor in a ram chip stick on the wrong value just long enough to be read.

It is unlikely, in fact I'd go so far as to say highly improbably that spyware is going to cause this as it needs IE to work in order to perform its task - its not going to sabotage itself.

It coul dbe a virus but a virus will do much more than alter a few values in ram causing your IE to crash.

Therefore it can only be a transitor stuck in a ram chip. It doesnt even have to be a permanent failure. those things are switching sof ast so oftern you would expect one switch to stick at some point. 9 times out of 10 windows can account for this and correct the mismatch, however it is sometimes missed and will cause the error mentioned in the first post.

You may continue your desperate struggles to prove otherwise however I would encourage you not to waste the bandwidth and persue more productive activities.

If I was not certain my stated description of the problem was correct I would not have posted. I quick memtest will confirm if the ram is in a state of permanent failure however I would guess it was just a case of a random sticking of one or two bits in ram. These are not uncommon and not anything to worry about.

If you would now like to set about proving that mechanical switches do not stick go right ahead.

 

[FishBoy]

I know you're only half joking here, but Linux crashes too you know... it's not all 100% fullproof like you guys like to say it is!

Just put as many applications as you put on Linux as you put on a Windows PC and we'll see how rock solid stable it will be... If Linux was as stable as you guys say it is, there wouldn't be as many package updates all the time

yea i know but i still prefer it better

 

[xtweaker]

Okay lets get one thing straight. If an applciation or as in this case an extension of Ie was trying to access memort space allocated to another process withour read permission you would get an Access violation not an application error.

Application Errors indicate that the application code itself is corrupted.

Funny... but Microsoft doesn't seem to think so in this case...

"Access Violation" or "The Memory Could Not Be Read" Error Message When You Quit a WinForm Program

It's not IE, but still goes against your sayings... Your logic doesn't make sense at all.

If you would now like to set about proving that mechanical switches do not stick go right ahead.

Since when is RAM mechanical?? I didn't know RAM had moving parts? Unless I didn't get it right, a Transistor Switch opens/closes by voltage saturation... Not by a mechanical intervention? How can this get stuck?

I'm not saying it can't ever happen, but I'd like to have some more information about that, and not from your own mouth... Could you give me some links to KB articles or online manuals that explain this behavior? I'm seriously curious about that.

Thanks.

 

[Dublex]

RAM isn't mechanical as far as I know.. I will dig out the A+ hardware extract that I have about RAM. (I've just done / passed the A+ cert as well so its stuck in my memory atm). From what I remember RAM is all controlled by eletcricty flow, voltages and transistors. Its possible a transistor may not hold a charge but I can't see how it would get "stuck" either.

 

[X-Istence]

That is something I didn't know. I figured Windows would be more demanding if anything, UNIX being so efficient and all. I'll find some stuff to read on it. Very interesting.

Windows also has error recovery as everything gets written to the pagefile as well as memory. Whereas Unix prefers memory over writing stuff to the pagefile, AND assumes it can use the entire memory chip. It does error recovery differently, but will in the end start panicing when memory is corrupt.

Since when is RAM mechanical?? I didn't know RAM had moving parts? Unless I didn't get it right, a Transistor Switch opens/closes by voltage saturation... Not by a mechanical intervention? How can this get stuck?

I'm not saying it can't ever happen, but I'd like to have some more information about that, and not from your own mouth... Could you give me some links to KB articles or online manuals that explain this behavior? I'm seriously curious about that.

Thanks.

It does not matter if ram is mechanical or not, what matters is the fact that as memory ages (Or even quite fast after it is new to your doorstep, cause of the speed up process they have, instead of quality control), the conductors and tiny transistors can begin to wear out and lose their voltage easily, and or not take voltage anymore, or always have voltage. This is when it will either always send a 1 or 0 back to the CPU, even though the value should be the opposite. Now you might think it is not possible, but it really is. The reason for this is that shorts can occur, transistors can burn out, and cause electricity to always flow.

Not that it matters in this case, as it is spyware or adware that is probably causing the problem. If you wish, run Memtest to check your memory though. Download Knoppix, and run memtest on the cmd line.

RAM isn't mechanical as far as I know.. I will dig out the A+ hardware extract that I have about RAM. (I've just done / passed the A+ cert as well so its stuck in my memory atm). From what I remember RAM is all controlled by eletcricty flow, voltages and transistors. Its possible a transistor may not hold a charge but I can't see how it would get "stuck" either.

Ill answer this here, rather than above. Transistors can over time deteriorate and not live up to the full speed that they have to be changed. This causes them to either old electricity, or not accept electricity for a small period of time, which can be just enough to throw off the CPU, and cause an error. This is what LordOfLa means with stuck.

 

[xtweaker]

Thanks X-Istence, one rep for you That explains it better. It's sort of what I ended up understanding from the reading I did since my post, but the way Lord said it wasn't as clear.

I'm not doubting that there can be invalid reads/writes in the RAM caused by stuck bits or bits where he value is flickering due to a bad transistor that can't hold its charge properly anymore...

But I highly doubt that it's the problem David is having with his, since I really think it would give errors for other programs, not just IE!

David told me he did do a spyware sweep and didn't find any, but he hasn't repaired IE yet using IEFix... and he had trouble with the memtest, so I sent him more detailed instructions on how to make the bootdisk. We'll see how that turns out.

Thanks for the clarification

 

[LordOfLA]

Thanks X-Istence, one rep for you That explains it better. It's sort of what I ended up understanding from the reading I did since my post, but the way Lord said it wasn't as clear.

So why didnt you say so instead of telling me I'm wrong when clearly I am not? tut tut

I'm not doubting that there can be invalid reads/writes in the RAM caused by stuck bits or bits where he value is flickering due to a bad transistor that can't hold its charge properly anymore...

But I highly doubt that it's the problem David is having with his, since I really think it would give errors for other programs, not just IE!

Well you explained why this wouldnt occur yourself. Application memory is protected. A transistor being a mechanical switch (I've yet to see a non mechanical transistor a contact still has to move) can stick just long enough to give the wrong value at the time it is read. At the time IE had this error, IE was the only application with access to that transistor.

Being as I suspect its a one-off fault its unlikely to affect other applications. If it was a more serious sign of impending module failure he would have reported several applications crashing and general instability of the computer.

David told me he did do a spyware sweep and didn't find any, but he hasn't repaired IE yet using IEFix... and he had trouble with the memtest, so I sent him more detailed instructions on how to make the bootdisk. We'll see how that turns out.

Which again proves my statement that it wasnt spyware/virus though I was quote ready to hold my hands up and stand corrected if that was the case.

I doubt IEFix will do anything its more of a placebo imo than anything really usefull.

 

[xtweaker]

LordOfA, thing is David said it happened more than once, not just that ONE time. It's highly unlikely that IE would try to read the exact same memory cell with a wrong bit more than once after several reboots don't you think? I didn't say you were all right just yet, just that it is "possible" that memory goes bad like you said, but it would give more symptoms than the same error coming from the same program even after multiple reboots. Do you understand my logic here? RAM gets flashed when a PC is shutdown, and as far as I know, chances that Windows will use the exact same memory cell for the exact same piece of code twice in a row after the RAM has been flashed, are very slim if not impossible... That's why I said, by the symptoms David described, that it is not likely to be a hardware issue at all, or he'd see errors of the same kind occuring with many other programs at random intervals, not everytime a specific program starts.

Anyhow, we'll all know for sure when the memtest has been ran by David, and I will stand corrected as well if it finds problems with his RAM.

EDIT:

Here's a quote from David's PM to clarify the symptoms he's experiencing:

I couldn't figure out how to use the memory tester thing. I'll try again later when I have time to read through the site.

That error message has only showed up a few times and usually several days apart.

I ran all my virus and spyware programs, but they didn't find anything.

If I am able to figure out for sure what the problem is I'll be sure to let you know.

David

 

[Mastershakes]

Lord: Please back up your assumptions with actual links. This little flame war is getting out of hand, and you've done nothing to support your opinions on this issue. Makes it easy as pie to refute them with my right hand man, Mr. Google.

Please stop referring to 'mechanical' there is nothing mechanical about 'electronic' RAM

Time to learn: http://computer.howstuffworks.com/ram1.htm

Still gonna use mechanical as a way of describing this functionality? Fine, but go to dictionary.com and read the definition, you may be out of context.

Lord, you are still way off. For the last time, it's not hardware.
It's software. Sorry Tweak got you all excited. Please read carefully.

I'll reserve most of my reply for after the memtest has been run. I've throughly researched this issue, I've even resolved it on client machines by reinstalling IE6 and Sun Java. I don't use iefix, but I use a functionality it provides, reinstalling via the ie.inf trick. Saves a time wasting re-image. No mem test or replacing RAM. I still havn't found any evidence that Windows can correct a memory mismatch 90% of the time, but my technet searches are slowly being refined. If I find something to back up this assumption, I'll post it.

Thought it might have something to do with Java - http://support.microsoft.com/?kbid=269446 -- but I think MS addressed this issue in IE6

Cool page of IE errors - http://www.colba.net/~hlebo49/erriexpl.htm


IEFIX a placebo. hahhahahahahahahhaha. In my opinion, that's the broadest and most baseless assumption you have made yet. It's the furthest thing from it. Perhaps take a gander at it's functionality. I'll answer any queries you have concerning it. It is simple to use, you shouldn't need any further instruction. Latest version they took out the automatic system file check (not sure why) but it can be run manually. Start, run -- type 'sfc /scannow' (no quotes) and have your xp CD in, or point it at your I386 directory

In short, it's a valuable tool in any Windows arsenal.

 

[translucenthoughts]

I have seen software cause this type of error. Application software. I have in past days, caused these type errors when coding (before releasing for general use of course!). In older programming languages, it can be caused by referencing incorrectly defined variables among other things...

But it does not mean this is definitely the problem. There could be an area on the hard drive (used as virtual memory) that is unreadable which a program (legitimate or spyware) is trying to access (unlikely at reboot, but not impossible if poor programming) or one of the other potential causes already mentioned...

I don't think there is only one right "possible" answer to explore - even though there is probably only one cause in this case.

I would go ahead and run memtest, can't hurt. I would also run anti-spy ware. If problem is not fixed, next I'd disable startup programs where possible - maybe even one at a time, to see if one stops the problem. Then go from there.

My thoughts are... break down the things you might try, start with the easiest and least expensive option (since you don't know which is the most likely) and go from there....

I would think that our original poster would like help as opposed to reading pages of debate!! It's great for learning (and sometimes chuckles) but its sure hard to weed through for the valuable input after awhile!

 

[LordOfLA]

Lord: Please back up your assumptions with actual links. This little flame war is getting out of hand, and you've done nothing to support your opinions on this issue. Makes it easy as pie to refute them with my right hand man, Mr. Google.

What makes you think I'm assuming anything? Everything I've stated here is common knowledge to anyone worth their salt in IT. As a windows developer I know what issues cause what errors with applications and largely how to fix them. As a PC and datacentre technician I know how the majority of harware in pc's work. Having studied electronics at school I know full well how the base components (transisitors, resisistors, capacitors) all work.

Please stop referring to 'mechanical' there is nothing mechanical about 'electronic' RAM

Time to learn: http://computer.howstuffworks.com/ram1.htm

Still gonna use mechanical as a way of describing this functionality? Fine, but go to dictionary.com and read the definition, you may be out of context.

Your own link there backs up what I have stated. A transistor is required (in fact 2 by that link) to switch on or off current to a capacitor in a grid. A capacitor is a mechanical switch that switches power on or off depening on which of 3 pins gets current. If that switch "sticks" on or off for a fraction of a nanosecond too long for whatever reason the capacitor being read will have the wrong value.

When RAM fails, it is the transisitors which have ceased to switch properly.

Lord, you are still way off. For the last time, it's not hardware.
It's software. Sorry Tweak got you all excited. Please read carefully.

I'm still of the firm belief it is hardware. Mechanical hardware fails. Computers are fully of microscopic mechanical switches. it only need one to not switch to throw out a whole system.

I don't use iefix, but I use a functionality it provides, reinstalling via the ie.inf trick. Saves a time wasting re-image. No mem test or replacing RAM.

If you have to reinstall IE it tells you there is someting more at fault with the machine. The fact you negelct a RAM test is worrying as the need to reinstall IE would indicate to me that a file has been corrupted. and that only happens if one of 3 compontents is failing/has failed: cpu, ram, IDE controller.

I wouldn't trust any of my personal PCs nor the servers at this datacentre to you to save my life. You portray an image of a not-so-good engineer.

I still havn't found any evidence that Windows can correct a memory mismatch 90% of the time, but my technet searches are slowly being refined. If I find something to back up this assumption, I'll post it.

Windows has always had some kind of software ECC for as long as I can remember. Windows has to tolerate hardware in a poorer state of repair then any other OS therefore it has to account for and where possible repair dataloss caused by a glitch in data storage devices or indeed transit pathways. Therefore allowing you to use a partly failed ram module or hard disk on a windows pc far more sucessfully than on a Linux or UNIX server for instance.

 

[LordOfLA]

LordOfA, thing is David said it happened more than once, not just that ONE time. It's highly unlikely that IE would try to read the exact same memory cell with a wrong bit more than once after several reboots don't you think?

Not necessarily. It would depend on whether that setion of code requires to be at the same address in ram all the time or whether it will permit being moved around at each load.

If it inists on the same place (non-Position Indepenant Code) then its highly likely and almost certain to end up using the same cells and transistors as the last time. If this is indeed the case and the same error with exact same values occurs repeatedly over a number of days I would investigae RAM failure further as a potentially failing module.

If however the values are changing in each error popup I would lean more towards a failure elsewhere in the system, say cross talk on the ide cable or even failure of th ide controller onboard.

...but it would give more symptoms than the same error coming from the same program even after multiple reboots. Do you understand my logic here? RAM gets flashed when a PC is shutdown, and as far as I know, chances that Windows will use the exact same memory cell for the exact same piece of code twice in a row after the RAM has been flashed, are very slim if not impossible...

See my response above

 

[Mastershakes]

Alright, here we go. I apologize in advance for the long post.

David, we need the results of that mem test pronto.... get them posted when you can.

When I've been personally attacked, I prefer to duly defend my position rather than go silent.

It begins with the evidence I've been slowly compiling on this type of issue. Then it deals with LordofLa's fruitless character attacks. This is all for you man. Enjoy.

http://forums.pcworld.co.nz/archive/index.php/t-40026.html

Same errors - was fixed by full reinstall - corrupted system files suspected.
Instead of reformatting - could have tried iefix, or the MS recommended ie.inf reinstall

http://www.seriousliving.net/new-2499262-741.html

Looks like they talk a bit about generating these application errors by bad coding - Mr. Developer.

http://www.codeguru.com/forum/showthread.php?p=873187

Again, coding seems to cause these errors. They couldn't solve it, or if they did, they didn't update the post.

http://www.lanwar.com/forumsh/forum...te&REPLY_ID=115271&TOPIC_ID=13391&FORUM_ID=40

Same error - ran two mem tests - nothing wrong. hmmmm

http://www.experts-exchange.com/Web/Browser_Issues/Q_20591182.html

You may need to be a member to see the response - but in it they discuss - guess what - problem with coding of the pages.

http://www.experts-exchange.com/Web/Browser_Issues/Q_21103256.html

In this one, they take my logic tack. iefix, windows update, all the good stuff.
Eventually the HIJACK THIS logs are analysed, and it appears there are several addons to IE that may be causing the problem.

http://ask-leo.com/iexploreexe_has_generated_an_error_now_what.html

A lot of errors can be generated via ActiveX protocols, IE addons....

http://www.easywindows.com/wmessages/2339.html

Points out it may be a virus. (and it does alot more than alter a few values )

http://www.sptimes.com/2005/03/14/Technology/Error_message_could_b.shtml

More in favour of my assessment getting the picture?

http://www.geekstogo.com/forum/yoursearchws_infection-t14633.html

Slightly different error - virus suspected. Nothing mentioned about bad RAM

http://forums.windrivers.com/showthread.php?t=69923

Same errors. Gee, there seems to be alot of bad RAM out there. Where is quality control ? (bleeding sarcasm)

http://reviews.cnet.com/5208-6620-0.html?forumID=14&threadID=49949&messageID=594648

Check out dem replies. I like their logic.

http://forums.wugnet.com/-iexplore.exe-Application-error-ftopict282591.html

Again, good clean logic.

The following are more of the same. Software is suspected in every one, sometimes resolved through patching the code, or fixing IE. It is the code that is suspect, and adware, spyware, and malware frequently have code that will mess up IE causing 'application errors'- until you fix it, or reinstall Windows.

http://www.annoyances.org/exec/forum/winxp/t1057940873
http://www.milonic.com/mfa/2003-August/002216.html
http://www.issociate.de/board/goto/...wing_up_leaving_preview_on_FP.html#msg_703723
http://www.mcse.ms/message1215062.html
http://www.mytechsupport.ca/support/topic.asp?TOPIC_ID=8307

http://support.microsoft.com/kb/262490/EN-US/ woot. Exact error. Seems to be OS related. hmmm

http://oca.microsoft.com/en/windiag.asp - Microsoft's very own memory diagnostic tool.

----------------------------------------------------------------------------------

Too many people have reported this error - it's unlikely that many people have failing RAM.
My PC at home was put together initially with value RAM. She bluescreened plenty of times, mostly in games, and then on a more regular pace. I replaced it with Kingston, and she's never done it again. I got BSODs when my RAM was messing up. Not a pretty little application error popup. Future BSODs were eventually traced to my onboard LAN - which I disabled until ASUS fixed it with a BIOS revision. I've left it disabled anyway. NIC card are dirt cheap, and I got PCI slots galore.

The fact that the same application errors out time and time again is indicitive of a software issue. If it was all sorts of programs, I'd start leaning your way.

It will be nice to see what the memory tests tell us. It's truely amazing that you have stuck to your guns on this thoughout despite the barrage of contradictory information.

In this case, if in fact there is no spyware, adware, malware on his system .... cannot remember if we've seen a hijack this log yet... fine. Once removed, some hijacks have managed to damage IE, Winsocks, and any amount of system files. Great thing is, with System Restore in XP - eventually even your restore points are infected with the corrupted file / reg entries.

I attack problems in a logical fashion. Hardware is the last thing you check. It would be extremely rare for this only to occur at startup if it was an imminent RAM failure. Even more rare that it's always iexplore.exe. Even MS concedes this is an OS problem, others in the above links surmise that it's coding. One got it solved by reinstalling the OS completely.

With regards to your personal attacks. Quite amusing. I work for one of the largest IT companies in the world. Our procedures and steps we take to resolve problems are tried and true. If our knowledge bases fail to find the answer, we resort to the web, and then 3rd tier (development) support. In production, the idea is to keep as many of our clients up and running as we can - it keeps them happy. So looking first at the source of the issue (in this case iexplore.exe) may lead to a quick resolution that gets them back up and running. On average, I fix IE about 10 times a week using a wide variety of tricks I've learned over the years. Since FF is not yet really an option at the Enterprise level of support for large corporations, I've got to be on my toes with regards to IE. Tools like Hijack This, and my extensive knowledge of the Windows registry help me immensely. I am eternally grateful to Spybot and Adaware as well.

Why do we go after the source of the issue? Well, a re-image takes them out of production for awhile now doesn't it. If I can sweep in using our remote tools, most IE issues are resolved in about 10 15 minutes. Some of the more intense spyware/adware malware issues take about 40, 50 minutes. They are usually quite content. Their browsing returns to normal, their in house web apps start working again, and there are no more popups. Since most of them have deadlines to adhere to, and their upper management that could care less that they've broken their PC - it helps that we can quickly resolve the issue - a touch less stress in their lives.

As far as looking into the reasons why these errors crop up, that's part of every major player's arsenal in the IT industry. Root-Cause analysis. I support production - I'll leave root-cause to them for now. Don't worry, it's done. When you deal with thousands of users, if all of them have the same issue, picture the impact it has on your helpdesk level, and 2nd tier support. We have service levels to adhere to, so those issues are resolved as quickly as possible to avoid fines that kick in on the contracts. We push out the fix as quick as humanly possible. The loss of productivity for our clients grows to an astronomical cost if we don't address it quickly and efficiently.

About me touching your datacenter. I'll pass man. I play with the big fish, not the little startups. BTW - you may want to take your digital-euphoria site back into your test environment for user acceptance testing. It failed my test. I really wanted to know more about your freeBSD Admin outsourcing services - but unfortunately none of your top menu links lead anywhere but index.php - I noticed the # sign following each, perhaps that part of the site is not finished yet - and you've placed bookmarks in your code. IMO it's better to release a website into production once it's finished - especially a corporate site. The title bar looks unprofessional as well. You may lose potential clients due to the disorganization the site is currently in. Again, IMO .... not mandatory... it's a private company.

That shoud about cover it. Now as far as me being a bad engineer. I treat about 80 or 90 issues a week. 2004 resolution rate was 86% on Windows related issues. Out of the 14% I missed, 9% was hardware related (and properly diagnosed, and subsequently fixed) The other 5% .... well then it gets tricky... mostly network issues, and some server outages - which is taken care of by another department. They fix it fast once they've received my diagnosis. So far this year, I'm at 82%

LordofLA --- Lord of Laughable Assumptions. Back your stuff up. I'm tired of your kind of IT. Explore your world. Avoid making any assumptions. Research the issue. If it's happening to you in Windows, it's likely it's happened to one of the other millions of proud Windows XP operators out there. It's also likely someone has fixed it.

While my colleagues have intoned that your scenario is possible - it's highly unlikely when you look at the symptoms. Fix IE, and potentially reinstall Windows is the correct course of action. If the error persists ---- well, run the mem test.

 

[LordOfLA]

Mastershakes, you've completely missed the point of all of my posts. So I must assume either I'm not typing in english or you are incredibly dense.

I have not once said "your ram failed/is failing" I said in my initial post it may have been a transistor stuck at the time IE was loading up on boot.

It was an "application error" and IE worked fine since.. Not a virus(malware) and highly unlikely to be spyware/adware as these depend on a stable IE to carry out their tasks.

So again. All perfectly logical explanation. I still think it was a stuck transistor. Not necessarily a sign of imminent failure but still worth a memtest.

 

[David_L6]

Still haven't run the memtest thing.... I don't have any floppy disks. Never needed one before. I'm working 12 hours night shifts right now and haven't had time to go anywhere to buy one. (20+ mile drive to nearest place to buy one.)


I have re-started the computer, run spyware and virus scans (found nothing), let computer go into power saver mode, turned it off and back on,.... haven't seen any error messages again (probably see one tomorrow morning when I turn it back on now that I've said that!).