PC Problem...

[itadakimasu]

some wireless devices might not support WPA2

http://www.dslreports.com/forum/remark,12691890?

oh, and i found tkip to be - Temporal Key Integrity Protocol. This is the encryption used w\ WPA, WPA 2 uses both TKIP ( to be backwards compatible ) and uses AES.

[Buffalobiian]

Yeah, all my devices support WPA2. Just that there's no option for WPA-AES.

I haven't done the hotfix yet, and I don't really feel like having WGA bugging me, though it's a genuine copy.

I'm just left wondering what encryption method it's using since the settings don't match, but everything still works.

[itadakimasu]

I think if windows says AES and the router also says AES, its using AES...

and if not its using TKIP.

Do you have some secret data you don't want to risk anybody intercepting?

This has me wondering why I even have my wireless router enabled since we have both pcs and ps3 directly connected to the router...

I was studying some wireless stuff last night and today because I failed a test this weekend mostly because I hadn't payed close enough attention to the wireless stuff. Luckily there is a free retake so my $150 isn't wasted yet.

[Board of Command]

I thought TKIP is just a key protocol, not data encryption. AES is data encryption, not a key protocol. They're not mutually exclusive and hardly related to each other.

http://en.wikipedia.org/wiki/Tempora...grity_Protocol
http://en.wikipedia.org/wiki/Advance...ption_Standard

[itadakimasu]

"TKIP comprises the same encryption engine and RC4 algorithm defined for WEP. However, the key used for encryption in TKIP is 128 bits long."

TKIP was made as a quick fix to weak WEP security. WPA2 is a finalized standard and it uses TKIP to be backwards compatible w\ devices that maybe can't do AES.

[Board of Command]

I see, thanks.

[Buffalobiian]

Do you have some secret data you don't want to risk anybody intercepting?

Does online banking count?

Thanks for the info guys. I'm still not sure what it's using, but I'm fairly confident it's AES, and windows just recognises it as WPA rather than WPA2.

Netgear's still as clueless as ever.

[Xelbair]

NTFS stopped being risky pre-1999.... have you been living in a cave?

ehh - with clonezilla NTFS support its still in testing phase - i know that others work well with this but i'm suggesting open-source software.

NTFS has been the standard Windows filesystem for years

yeah i know - it does not waste bytes like fat32 in each sector.

So my router is configured with WPA-TKIP OR WPA2-AES. It'll try WPA2 first, then back down to WPA-TKIP if that fails.

Windows says WPA-AES!!

So....which encryption is it using anyway

Its using AES i think, you set "WPA-TKIP or WPA-AES" so it uses better encryption if possible. I got netgear router too - it only failed with my cousin integreted Nvidia card.

[Buffalobiian]

I'm using an E6600 (2.4GHz Dual core) with the stock heatsink & fan. I know at low enough temps/voltages, the intel fans start and stop - that's normal. But has it stopped and failed to start for any of you?

On idle, C1E downclocks the multiplier so it runs at 1.6GHz. At ambient of around 28C, it ithe core idles at 39C (coretemp), 37C (easyTune), and stock fan runs at ~888rpm.

The normal starting/stopping of the fan happens when temps get low enough on the Default EasyTune Settings (20C - spin 31%, 61C - spin 100%)

The situation where the fan fails to start has only happened once before, about a year ago. I noticed some sluggish performance, and check the CPU temp...85C!! I rebooted the computer, and temps were back to normal with the fan spinning. (Note: I didn't actually witness the fan stopping when it was at that temp).

I thought the fan may be stuffing up when it's trying to start up again, so I jacked SmartFan settings up to (20C - spin 49%, 61C - spin 100%) so that it'll keep spinning at all times - no chance of stopping.

A year later, in summer, I changed the settings back to default to lower some noise. It's not that bad, but it does help. 3 days later, I came back to find web pages loading slow, EasyTuneCentre encountering an error, and the temp display on CoreTemp showing

Core 0: 85C (?)
Core 1: 85C (?)

It showed exactly that, question marks and all.

Rebooting yielded stable temperatures again, which made me think back to the same even 1 year ago.

Basically, do Intel fans do this?

[darkshadow]

Hmm I don't think so, my fan is also always at ~2400 rpm's at my 3ghz. The cores idle at:
core 0: 45C
core 1: 44C
core 2: 38C
core 3: 40C
And their MIN/MAX are:
core 0: 39C - 59C
core 1: 38C - 54C
core 2: 33C - 50C
core 3: 36C - 52C

Maybe your sink is clogged with massive amounts of dust? or your voltages have become irregular?
After I set my voltages higher for some stability, my idle temp jumped to 80-85C for some reason which is the same as shutdown temp, even after putting it back, temps would be easily 15C higher then normal.
So I shutdown my pc for the night, and now its fine again.

[Buffalobiian]

Dust is definitely not a problem, since I cleaned it out when I chucked everything into a new P182 case about 2 weeks ago.

For some reason, the overheating thing happened again this morning. I left my computer on so it could finish with some encoding, which would finish about 1 hr after I went to bed. I wok at 6am to find:

1) encoding as finished
2) procexp had a "Application error: memory could not be written" error
3) Core Temp was reading 85C again
4) Computer was sluggish and kind of unresponsive.

Core Temp logs show:


Note: I truncated the data with ... to highlight the important data.

Interestingly, there's a gap between the logging between 5:05 and 5:11. The event viewer shows that starting @5:05, this happened:

5:05 -> System restore filter error regarding file ALSysIO.sys (a temporary coretemp sys file)
5:09 -> Application errors, memory could not be written for procexp and EasyTuneCentre (3 entries total)
5:11 -> The ALSysIO service failed to start due to the following error: The specified driver is invalid. (This same error logged every second from 5:11 to 6:21, when I closed CoreTemp)

CoreTemp calculates Intel temperatures using the following formula:

Temp = Tjmax (85C for Conroe) - Delta (Variable read off the DTS)

If I assume ALSysIO is critical for reading Delta, a ZERO for that value would explain why the displayed temp was always 85C. It may also explain why the detected clock speed was 2400MHz (program used default BIOS settings while ignoring dynamic downclocked multiplier (6x-9x)).

Upon exiting and restarting CoreTemp (without rebooting the PC), temperatures were back to idleing 39C.

System event log also displays "The ALSysIO service was successfully sent a start control." at that time.

-----------------------------------------------------
From what I see, it looks like a CoreTemp read error, rather than a real overheating issue. I've upgraded CoreTemp from v0.99.3 to 0.99.4 after that, so you can consider it a reinstall.

Reasons I'm not 100% sure it's purely a software driver stuff-up are because:

1) Overheating's happened before, fan's stopped before.
2) The gaps in the logging (5:05-5:11, and 6:21-6:23) doesn't exactly rule out hardware problems, since it's ample time to crank up the heat if it was really a fan failure, and 2mins is also adequate to cool a CPU down from 85C.
3) It doesn't (or if it does, I don't know how it) explain why memory errors occurred for procexp and EasyTuneCentre as well.
4) Sluggish performance was noticed for ~20-30sec after I turned on the monitor.

So yeah, I hopefully my CPU didn't get to do any roasting, but it's possible.

And that system restore error wiped out all my restore points

[Deadfire]

It could be any number of alot of things.

A failed/failing CPU cooler fan is a definite possibility. Another possibility is that the cooler has loosened some how and is no longer making good contact with the processor. Also, you might want to make sure that all BIOS settings are at their safe default values.

Check new drivers and BIOS updates.

Because the encoding uses CPU processing it could be that the app is FUBAR in how it handles your CPU/motherboard/memory.

Again it's a odd thing, hard to judge exactly what it is.

[Board of Command]

Load it up to 85 degrees and touch the heatsink after a minute or so. If it feels too hot to the touch, then you're really overheating. You should try to rule this out first.

Also, run Memtest 86+ to make sure your memory isn't dying.

Oh, and just for the heck of it :P


I also have a screenshot of a Phenom II socket-AM3 version, but I'm not allowed to post it

[Buffalobiian]

@DF:BIOS and drivers are all good. I've used the encoding program for a good year or two, and it's worked well. The encoding would have stopped at 11pm at night anyway, so I shouldn't have been doing anything (shown by the CPU load). I get what you mean, how it might stuff up memory. I'll have to see if it does it again.

Load it up to 85 degrees and touch the heatsink after a minute or so.

0_o....you mean like, unplug the fan so it overheats? If not, I don't know how I'll crank it up to that, since it can't normally hit such a high temp. 100% load over a hours will probably get me something above 55C, but it won't hit 60C. I've fiddled with the EasyTune 5 fan settings. It still won't get that hot under normal settings.

So to get it there, I'll have to either unplug the fan, overclock it, or somehow get it to overheat again. If we count the last instance as a legit "overheating experience", it's only happened 3 times in the last 1.5 years. Both the first two times was when I left the EasyTune fan setting on default, which results in the fan starting/stopping at lower temps. Though that's completely normal, I've concluded that both times it's because the fan had problems starting.

Side info: I'm using a Gigabyte GA-965P-S3 with F14A BIOS. Google didn't say this board is commonly associated with fan controller failure, but it's entirely possible. There was one example of a guy who had a E6600 fan-stopping problem as well. (but not with all the coretemp/application stuff)

I'll encode another episode or two tonight and see how it goes.

[Board of Command]

Oh I thought you meant it overheats consistently. If you can't reproduce the problem, then ignore what I said.

But do run Memtest.

[Buffalobiian]

Thought I'd report: Memtest came up with no errors.

Interestingly, yesterday I was running CoreTemp 0.99.4 on my laptop (AMD Turion 1.66GHzx2), and for a minute or two, temperatures were shown as:

Core0: 0C (?) (!)
Core1: 0C (?) (!)

Since AMD uses a different formula (Core Temp = Value - 49), such a reading indicates yet another reading error.(Note: CoreTemp does not display temperatures below 0C, despite it being theoretically possible)

Core Temp resumed normal readings a minute or so afterwards, with (!) to indicate past invalid readings.

[Board of Command]

Temperature reading for AMD chips is a bit shaky. My Phenom II 920 is sitting at 11 degrees right now acccording to CoreTemp.

Take whatever you see with a grain of salt. It's only reliable for Intel chips.

[Buffalobiian]

My Dell Inspiron 1501 failed to boot this morning. I get through the first (and only) page of bios, and when it's meant to load the windows XP logo, it instead flashes grey/white for 0.5 seconds, and just keeps rebooting over and over in the same cycle.

I'm going to try my backup disk, but since the CD reader tends to pick up errors, I would prefer it as a last resort - unless I copy the disks to an external HDD and run the backup from there.

It's been working fine for a few months now. The only thing I can remember I did yesterday was uninstall the graphics drivers and upgrade to 9.2 for the ATI 1150? onboard card.

I've recently been switching to using hibernation and standby modes a lot more (in the past week), but that's it. I'll post some pics later when I get back home, since I'm at uni.

Currently running the PSA + Pre-boot Assessment/memtest - so far so good. Completed with no errors...

Interestingly though, during the normal boot-memory-test, the numbers climb up to ~6GB of ram, but then settles on 4GB - OK. I can't remember if it's always like this or not, but I'm pretty sure I've only got 4GB of ram.

---------------------
Weird. I went into bios -> selected "Load default settings", then changed the setting for "show boot-diagnostic screen" from "default - off" to "on", and it worked....?

Anyway, running chkdsk /r and defrag just for health's sake.

It was coming back from hibernation btw.

[itadakimasu]

can you hit f8 and boot into safe-mode?

If you just upgraded the video driver, that seems like the most likely cause.

See if you can boot into safe-mode and maybe roll back the video driver and see if it boots up normally after you do that

[Buffalobiian]

It boots normally now....I don't know why. I'll do another backup onto a network drive just in case. It was actually weird because since it's meant to be loading from hibernation mode, it actually failed before the "Resuming from hibernation" page, which looks like the bios layout as opposed to windows, so it shouldn't be a display problem neither.

Then again, as you said, the only thing that changed was the upgraded drivers. I'll see how things go and bump this if anything turns up.