RAID 0 Crashing Games

M

mInk3nine

Weaksauce
Joined
Apr 2, 2002
Messages
66
I just setup a RAID 0 array using two WD 80GB drives with 8MB cache. I'm using 64K clusters and all is fine until I run UT2K4. It crashed on both attempts to play it. First time on the ingame server browser and the second time it happened about 5 minutes into playing. Would you recommend decreasing the cluster size to 16K? I'm new to RAID so i'm not exactly sure how to fix this problem.

I just recieved my 160GB WD 8MB Cache as backup, and I installed both UT2K3 and UT2K4Demo on it and they run fine.

What could be causing these games to crash on the RAID array?
 
you mean stripe size?

Stripe
NTFS Cluster use \ FAT32 Cluster use \ Slack


while the RAID array may be the main change that has occured since the problem developed
it doesnt necessarilly follow its the cause, Id troubleshoot the crash like any other, starting with the Event Viewer, testing the memory, cleaning the registry, reloading drivers (vid chipset ect)
 
I have a SiI680 RAID0 (PCI card) with 2 WD 120GB SE drives (WD1200JB) and 64KB stripe size. The UT2004 demo plays fine off it and I like how load times are very fast. :D

Your RAID might be flaky. Be sure to back up anything important before you continue.
 
Er...stripe size...yes. Sorry, i'm new to this ;). Everything is backed up on my 160GB drive. I bought it so I wouldn't feel like i'm walking on egg shells.

I've used Sandra's Burn-in Wizard. I ran Prime95 last night and left it running for about 9 hours. Ran Diskspeed32 numerous times. No crashes. Everything seems fine. it's weird because the first night I loaded up UT2k4 it kept crashing. The next day I attempted it and I played for about a half hour before I got tired of playing and exited.

I guess I'll try reloading the chipset drivers...
 
This almost seems like a power issue to me. Prime and Sandra only tax your mobo/RAM/HDs, but not your vid card. Try running the system with all three drives connected to power (no IDE cables), but only the 160gig actually in use and see if the system still crashes during UT2K4.
 
Originally posted by BigBadBiologist
This almost seems like a power issue to me. Prime and Sandra only tax your mobo/RAM/HDs, but not your vid card. Try running the system with all three drives connected to power (no IDE cables), but only the 160gig actually in use and see if the system still crashes during UT2K4.
Click to expand...

It can't be...I just bought a 500W Powersupply...I wouldn't be able to do what you suggest anyway since i'm booting windows off the RAID array...
 
That doesn't mean it puts out 500W, and, more importantly, a sufficient quantity of amps on the right lines.
 
Well, it is a Modular PSU afterall...I'll try hooking up my 350W Enermax PSU and see if anything changes....

But honestly, do you think its a PSU issue? I can do everything, i mean everything except play a Game sufficiently off the RAID array...

EDIT: Keep in mind it was crashing the game before I added the 160GB drive to the mix
 
Ok...i'm running into more problems now....

This is really odd, i'm beginning to think that I got what I paid for in this 500W PSU ($99). I loaded up MBM5 and have been checking the voltages since my last post. And Snugglebear, you're totally right, i'm sorry I doubted you. The Voltage on the 12V rail is very low compared to my Enermax 350W. Everything else seems just about dead on, but the readouts for +12.00 and
-12.00 Fluxuate between +11.55V to +11.77 and -11.67 to -11.84.

Now here is when it starts getting weird. I got fed up and reinstalled UT2K4 back onto my 160GB (D:\Backup) and started playing. It locked up. I couldn't believe it. So out of spite I ripped out my 350W out of my old case and hooked it up. It booted, I went into the BIOS and monitered the Voltages. Alot better results. It has always been such a good PSU, but I digress. I got out of the BIOS, the computer went through the POST like it usually does, loaded the RAID array....Then here's where the problem started.

I got an error msg saying that a certain file has been deleted, it instructed me to try booting off the WinXP CD and hitting 'R' to repair bla bla bla. I tried that, no dice. The setupdd.sys file couldn't be loaded? So i'm thinking something is horribly wrong, maybe I friend something? No. I plugged the 500W back up and it booted normally, and I wouldn't be typing this now if it hadn't.

So all I have to say is, WTF?! Last time I installed a new PSU i don't recall running into so many problems. So I guess the whole problem with my crashing issue just happens to be this sh*t power supply. I think i'm going to return it and try the Antec Modular PSU. Anyone have any suggestions or comments? I'd like to know what some of you are thinking...and Why my rig wouldn't boot when trying to swap PSU's...
 
Originally posted by mInk3nine
Modular PSU afterall...I'll try hooking up my 350W Enermax PSU and see if anything changes...
Click to expand...

its no longer total wattage that is the concern and often its not even the amperage per rail, but the stability of the PSU that is often the issue these days...

Winbond Launches New Bus Termination Regulator April 4th 2003

"Winbond Electronics Corporation, a leading supplier of semiconductor solutions, today launched the W83310S, a new DDR SDRAM bus termination regulator. The solution, new to Winbond's ACPI product family, is aimed at desktop PC and embedded system applications with DDR SDRAM requirements.
Computer systems architectures continue to evolve and are becoming more complex; CPU and memory speeds continue to increase ever more rapidly with every technology turn. More and more high current/low voltage power sources are required for PC systems. This is particularly true for high-speed components such as CPU, memory, and system chipsets. The performance of these components is highly dependent upon stable power. Therefore, motherboard designers require accurate, stable, low-ripple and robust power solutions for these components.

Many system designs use discrete components to implement bus termination functions. This approach creates several problems including poorer quality load regulation; higher voltage-ripple, increased usage of board space and inconsistent designs when different discrete components are used.
Click to expand...


Transient Response: As shown in the diagram here, a switching power supply uses a closed feedback loop to allow measurements of the output of the supply to control the way the supply is operating. This is analogous to how a thermometer and thermostat work together to control the temperature of a house. As mentioned in the description of load regulation above, the output voltage of a signal varies as the load on it varies. In particular, when the load is drastically changed--either increased or decreased a great deal, suddenly--the voltage level may shift drastically. Such a sudden change is called a transient. If one of the voltages is under heavy load from several demanding components and suddenly all but one stops drawing current, the voltage to the remaining current may temporarily surge. This is called a voltage overshoot.

Transient response measures how quickly and effectively the power supply can adjust to these sudden changes. Here's an actual transient response specification that we can work together to decode: "+5V,+12V outputs return to within 5% in less than 1ms for 20% load change." What this means is the following: "for either the +5 V or +12 V outputs, if the output is at a certain level (call it V1) and the current load on that signal either increases or decreases by up to 20%, the voltage on that output will return to a value within 5% of V1 within 1 millisecond". Obviously, faster responses closer to the original voltage are best."

a few other terms that bear on the discussion

Load Regulation: Sometimes called voltage load regulation. This specification refers to the ability of the power supply to control the output voltage level as the load on the power supply increases or decreases. The voltage of a DC power source tends to decrease as its load increases, and vice-versa. Better power supplies do a better job of smoothing out these variations. Load regulation is usually expressed as a "+/-" percentage value for each of the voltages the power supply delivers. 3% to 5% are typical; 1% is quite good. (The -5 V and -12 V signals usually are no better than +/- 5% even on very good units; there's no point bothering getting them better than that since they are low-current and mostly unused anyway.)

Line Regulation: The complement of load regulation, this parameter describes the ability of the power supply to control its output levels as the level of the AC input voltage varies from its minimum acceptable level to its maximum acceptable level. Again, a value for each output level is usually specified as a "+/-" percentage. +/- 1% to 2% is typical.

Ripple: Also sometimes called "AC Ripple" or "Periodic and Random Deviation (PARD)" or simply "Noise". The power supply of course produces DC outputs from AC input. However, the output isn't "pure" DC. There will be some AC components in each signal, some of which are conveyed through from the input signal, and some of which are picked up from the components in the power supply. Typically these values are very small, and most power supplies will keep them within the specification for the power supply form factor. Ripple values are usually given in terms of millivolts, peak-to-peak (mVp-p). "Peak-to peak" refers to measuring the AC voltage from its negative maximum to its positive maximum (see here for an illustration of what this means.) Lower numbers are better


[H]ardcore PSU Info


so, it could still be a power issue, eliminating it as a suspect would be nice
 
The long & short is that the amperage level per each output should be listed on the PSU itself, and you should give yourself a good 10-20% buffer over your equipment requirements, more if you want to add stuff later. A year ago I ran into this when adding another disk to my fileserver, in which I was using a 450W Antec supply. Turns out it has less amps on the 5V lines as compared to my 350W Enermax. Swapping the two returned the machine to normal functionality. The rest... well, if you have a hacked BP6 and goofy VRs on the board, you begin to see that even little $2 parts aren't made equally.
 
Ice Czar: Thank you so much for taking the time to write that explanation. Now i don't feel so dumb :rolleyes: .

BTW, the [H]ardcore PSU Info link is dead.

Snugglebear: So are you saying i'm SOL by trying to revert back to my 350W PSU? I'm going to need to use something while I send the 500W back and switch it with a more reliable brand name.
 
Originally posted by mInk3nine
Ice Czar: Thank you so much for taking the time to write that explanation.
Click to expand...

LOL just cut and paste
all the term descriptions are from the PC Guide (link)

Your Antec idea is a pretty good one, if you go the True Power Line, see that [H]ardcore PSU info link (and the site its talking about)
I fixed it, inadvertanently linked to my local HDD
file:///E:/Documents%20and%20Settings/Administrator/Desktop/FAQs
/Diatribes/PSU/hardInfo/%5BH%5DardForum%20-%20%5BH%5Dardcore%20PSU%20info.htm][H]ardcore PSU Info :p

The True Power Line is a very stable PSU
 
haha, silly me. Yea i was wondering how you posted that so quick :eek: I just assumed you had god like typing skills.

I just read the review on the 480W Antec Modular TruePower at TwistedMods.com. They tested it with a multimeter and the results impress me. I think i'm going to go with that. I wanted to buy it originally but its pricey after FrozenCPU made it modular, but eff it, I need to play UT2K4! I'll do anything! lol

Thanks again.
 
All I'm suggesting is that you read up on your equipment specs, add together the required peak loads on the various devices, then make sure your PSU can support it.
 
dude...i'm so pissed. FrozenCPU won't take the powersupply back because it's been premodded, voiding it of it's warranty. So i'm out $114 bucks....guess i have to stick with my 350W for now....I'm going to go give myself a blackeye, maybe two.

I'll read through those articles now. Thanks again.
 
Sorry to hear it. Perhaps some day ATA drives will support delayed motor start commands. It would definitely allieviate most of the power problems around.
 
there is alot of "windage" both ways when it comes to any calculator or supply

calculators use the maximum voltage per rail if ALL the components where employed at full load at the same time, something that rarely if ever happens
your "normal" load is probably about half the total or so (varies with the components)

alternately, some PSU manufacturers lie :p
the rails tend to be over stated (or combined)

Dans Data
Unethical PSU Marketing 101.

Here's how to make overly optimistic power supply specifications. It's really simple.

First, power the thing up. You can make an ATX power supply that isn't connected to a motherboard turn on by grounding pin number 14 on the big motherboard power connector. It's easy to spot that pin, because it's the only one with a green wire going to it.

Use any handy bit of wire - like the paper clip in this picture - to connect pin 14 to any ground contact. The ground contacts are the ones with the black wires going to them. Presto, the PSU will turn on.

Now, break out your brick-sized power resistors and load the heck out of one of the output rails - the +5V rail, for instance. Measure the current as you increase the load, until the voltage sags unacceptably far below the rated voltage.

How do you tell what an unacceptable voltage sag is? Well, you could choose a nice conservative small permitted sag - say, 0.1 volts - so that your results are genuinely useful to your customers. Or you could just ignore the voltage and say that when a fuse (or some other component...) blows, that must have been the limit, right there.

OK. Now you've made a big fat amperage number for the +5V rail. If you blew up the PSU in the process, get another one, and repeat the process for +12V and +3.3V, and for the low current rails as well.

On no account, though, should you test more than one rail at a time. This is the key to the whole scam.

A big beefy PSU may be able to deliver 50 amps (say) on the 5V rail when nothing else is under load, and 25 amps (say) on the 12V rail when it's similarly all alone. But the 12V and 5V rails together may only be able to deliver, say, 350 watts between them, when they're both under load. Watts equals amps times volts.

In a real PC, all of the power rails will always be under load together.

But you're not testing what the PSU can really do - you're making pretty numbers for the sales brochure!

So test all of your rails alone, get an amperage figure for all of them, multiply that figure by the voltage of the rail it came from (the nominal voltage, not whatever the voltage had sagged to as the PSU pumped electrons through the dessert spoon you'd soldered to the circuit board), then take all of the resulting wattage figures and add 'em up. That's a wrap, folks. Ship it!
Click to expand...

:p :p :p
 
You must log in or register to reply here.
Back
Top