Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[marko]

I've got either a ventilation problem, a hard drive problem, or a motherboard problem. Either way, it's no fun. I'm going with ventilation to start with....

One of my drives keeps going AWOL, and when it does, the whole system goes mental, then crashes, and when I finally get it to boot again, and chkdisk has run, half a dozen apps have corrupted settings files that need restored from backups.



 

[JimH]

Get a good backup.  My money is on the drive.

[rjm]

SpinRite http://www.grc.com/sr/spinrite.htm works as advertised.

My mother-in-law's laptop drive quit a few days ago and SpinRite repaired it in a couple hours. I still changed out the drive to be safe but SpinRite successfully recovered all the data so it was easy for me to clone it to a new drive.

[marko]

Get a good backup.  My money is on the drive.

You might be right...

Copying files to another drive on the LAN, I'm only getting 4Mb/s transfer speeds from the dodgy drive.
Copying from any other drive to the same LAN drive yields just shy of 10Mb/s.

So, fair to say I've either got a drive with one foot in the grave, or a dodgy SATA controller on the motherboard.
Does anyone happen to know if the six SATA ports on the MoBo are independant of each other, or do they all have the same controller?

If one controller takes care of the six SATA ports, fair to assume the HD's on its last legs?

--------
If the drive is knackered, that makes that the second Seagate 500 Gb drive to give up on me this year, and they were both bought in May 2007.
The first one got RMA'd under warrenty due to S.M.A.R.T. errors.

You know you're getting old when you start saying things like "They don't make things like they used to..."
Our very first PC, bought over eight years ago came with a whopping 15Gb hard drive. It was a Quantum Fireball. That PC was broken up for parts, and that 15Gb drive is still working just fine in a PC I built for my daughter about three years ago. The nVidia Riva TNT2 with 32Mb memory still works just fine too, though that's currently wrapped in an anti-static bag in my 'box of bits' that the wife is so keen to throw out with the rubbish!!!

I've got no cash for a new hard drive, boohoo.
Hopefully I can get this resolved with Seagate without putting my fighting head on (if it is indeed a faulty HD).
Why do things always start breaking down at Christmas time? 

-marko

[KingSparta]

If the drive is knackered, that makes that the second Seagate 500 Gb drive to give up on me this year, and they were both bought in May 2007

They May Be Left Over Maxtor Drives.

If I Am Not Mistaken Most Seagate Drives Come With A 5 Year Guarantee (I Bought 3 Segate 500 Gig Drives In The Past Year, And That's What Is On My Box).

[marko]

tomorrow, I will be checking to see if the 150Mb/s SATA jumper has been left on that slow drive...

[KingSparta]

I don't think that would create an Error.

[marko]

No, but it would explain the pitiful 4Mb/s transfer rate I'm seeing from the drive, along with the random "the device has out of date firmware" mesaages I've also been finding in the system logs.

The jumper was in place. I removed it and now the drive is operating at full speed again.

The drive in question is, would you believe it, the replacement drive I received from Seagate under an RMA back in April. What a palava that was, let me tell you. I had to obtain a copy of the signed P.O.D. from the courier and pass that on to Seagate before they would accept that they had actually received my faulty drive!!

I've bought a new drive. A cheap as chips 'no brand' affair in a plain brown box. Turns out the drive is a Samsung HD753LJ which has extremely favourable reviews out there. Only time will tell, but initially, I'm happy with that. The drive is virtually silent, which is always a nice bonus.

Hopefully, things will settle down now...

[marko]

So, the 'smart' money is now resting on the motherboard being faulty...

Bought new hard drive, swapped it out in place of suspect drive and within 1 day, it went AWOL too and brought the system down... sheesh!!

Unsure if it was related or not, but the new also began throwing S.M.A.R.T. errors on the next boot.

I went through yet another low level format and took the drive back. I hooked the replacement drive into a different SATA port on the board and so far, so good...

[KingSparta]

one time i air dusted the Firewire 800 controller board, and the drive stopped going AWOL.

I then bought a new Firewire 800 controller and have not had a problem yet.

[marko]

a quick update, just in case anyone was wondering....
8 days uptime and a clean system log....

--------[ UpTime ]------------------------------------------------------------------------------------------------------

    Current Session:
      Last Shutdown Time                                10/12/2008 07:13:36
      Last Boot Time                                    10/12/2008 07:14:48
      Current Time                                      18/12/2008 07:34:07
      UpTime                                            692388 sec (8 days, 0 hours, 19 min, 48 sec)

    UpTime Statistics:
      First Boot Time                                   10/12/2008 07:14:48
      First Shutdown Time                               10/12/2008 07:13:36
      Total UpTime                                      692388 sec (8 days, 0 hours, 19 min, 48 sec)
      Total DownTime                                    72 sec (0 days, 0 hours, 1 min, 12 sec)
      Longest UpTime                                    692388 sec (8 days, 0 hours, 19 min, 48 sec)
      Longest DownTime                                  72 sec (0 days, 0 hours, 1 min, 12 sec)
      Total Reboots                                     1
      System Availability                               99.99%

I'll live with that. Hopefully, the failed SATA port is an isolated affair, if the rest start to fall over, I'll have to look for new motherboard, and I really don't want to be doing that!

-marko.

[JimH]

Well, I lost the money, but I'm glad you found the problem.  Bad SATA port.

Bob at JRiver is one of the best problem solvers I've ever known.  His approach to problems like this is just what you've done -- start swapping stuff and see what happens.  It's amazing that something as smart as a computer can be so dumb sometimes.

[glynor]

Hmmm... That probably doesn't bode well for the board.  Do you know if the SATA port in question was in a different "block" on the board (physically) from the one that worked?

The SATA interfaces should be hung off of the chipset (southbridge probably, but it depends on the specific chipset you have), unless you have an additional auxiliary SATA chip on the board.  Some (many more "high-end" boards) include the 4-6 SATA ports from the chipset, and then throw in a couple of extras supported by an additional chip on board.  If only that aux chip went bad, then you could be okay.  However, if your chipset is going, then you could start seeing other I/O issues as well (USB would be something in particular to watch).

Could also be the traces to that particular port, I suppose (or the actual port connector itself), but this seems a LOT less likely.  Did this SATA port seem to "go" bad, or was it probably always bad and it just took a long while to figure it out?

If the board happens to be still under warranty, I'd really try to send it back.  If not, I'd seriously consider swapping it out.  It could be a forbear of further issues down the road a bit...

[marko]

Hi glynor,
My board's an eVGA 122-CK-NF68-A1
At 18 months old, it's an absolute pensioner in motherboard years!!

It has 6 SATA ports.
4 of them (3 - 6) are grouped near the center of the board, and the other two (1 - 2) are hanging off the side opposite the backplate.

It's ports one and two that appear to have gone bad. I'm not really sure if they run off their own chip or not. The manual is here.

-marko.

[glynor]

They do not.  All 6 of your SATA ports are hung off of the 680i SLI MCP (the Southbridge, which is identical to the one on Nvidia's older 590i chipset).

There are well known problems with the SATA support on the 680i chipset (one of the reasons I've avoided Nvidia chipsets for Intel with the current generation).  Nvidia has released a number of hotfixes.  Here is the latest info from Nvidia on these issues: http://www.nvidia.com/object/680i_hotfix.html

This update is related to SATA disk drives and system instability. System instability is observed in the following ways (not a complete list):

    * Random application shutdown
    * Corrupted boot drive
    * "Blue screen of death"
    * Corrupt data

This update will improve system stability and prevent future stability issues related to SATA disk drives on those systems requiring this hotfix.

Here is a post from Nvidia about the issue on their own user forums: http://forums.nvidia.com/index.php?showtopic=23254
Here's more info on the problem from HardOCP: http://enthusiast.hardocp.com/article.html?art=MTI0NCwxLCxoZW50aHVzaWFzdA==

I did see on the EVGA Forums that there is an even newer BIOS release for that board that includes as one of it's major highlights:

Fixed SATA / MAC / USB compatibility issues

http://www.evga.com/forums/tm.asp?m=546893

So if you aren't running the latest BIOS, it is certainly worth updating.  This vP33 BIOS appears to be from September 2008.  However, I do see that there is a HUGE number of other problem reports on the EVGA forum and the PCPers forums regarding that board (and Nvidia 680i boards in general) for a wide variety of issues.  It appears that current BIOS upgrades aren't always guaranteed to fix the issues, and that many people have had to RMA boards.  Good news is that both Nvidia and EVGA are well aware of the issues, and they appear to have been fixed with current revisions of the boards.

Even though your Warranty has been expired, it might be worth contacting EVGA support.  They are known for having good support.  I'd just explain that you've been experiencing issues for a long time but that it took months to track the problem down to the SATA on the board, which is why you waited until after the warranty period.  I'd say you stand a fair chance of getting a new one if you act soon (you're only 6 months past the warranty date right now).  EVGA is well known for being pretty reasonable when it comes to customer service.

[JimH]

So if you aren't running the latest BIOS, it is certainly worth updating.

Or you could put a little duct tape over the bad ports.

[glynor]

The worry is that if his problem really is these reported issues, that it will not be limited to just those two ports.  These issues are systemwide, but sporadic and unpredictable.

[rjm]

hmm....

Might explain a problem I had a while ago. My e-Sata port is 1 of 2 ports hooked up to a Silicon Image controller. The other 4 ports are run by the Southbridge chip.

I noticed my e-Sata drives where running slow so I dropped using it altogether and switched back to USB 2.0.

[marko]

[lowest form of humour]
Cheers glynor... and a merry christmas to you too!
[/lowest form of humour]


Seriously though, thanks for doing the leg-work for me and thanks for the advice.

You know, before buying this board, I researched, and then I researched some more. I don't recall finding a single bad review for it.
I had had such a god-awful time of it with VIA chipsets up to that time, and was hearing nothing but wonderful things about nforce, that I had decided that nforce would be my next choice.
I believed I had researched and chosen well... perhaps I could have done better.

I wonder if ports 1 and 2 were at the heart of my Seagate RMA debacle... Seems quite probable...

I've read through all 8 pages of the eVGA forum thread, and I'm not convinced that a BIOS upgrade is going to help me.
I've got the iso image but think I'll take the "if it ain't broke" approach for now as things are peachy atm.
I understand the 'sporadic and unpredictable' nature of these issues, but, my experiences are nowhere near as catastrophic as those reported. No BSOD's, no drive corruption, just SMART errors, no USB issues and as far as I'm aware, no PCi issues either. I know that SMART errors are obviously not good, especially if the board is causing them, but, I don't count the second one as it appeared after only 19 hours of 'power on' time and the drive went straight back. This is what turned me on to the SATA port problem. Granted, it's only been 10 days or so, but, since ditching ports 1 and 2, things have been fine. I will keep monitoring the situation over the coming weeks...

I cannot see me needing SATA ports 1 and 2 anytime in the forseeable future, so, I think I'll run as I am for now, with those two covered with duct tape and the P28 BIOS that the board shipped with. Even with the P33 update, I still wouldn't have any greater confidence in the board, and I've read enough feedback that the update either didn't fix anything, or made things worse to put me off.

I'll give you a shout when I come to buy my next board and let you spend my hard earned for me as I'm obviously rubbish when it comes to research!!

-marko.

[glynor]

I read all the reviews too.  It was the most highly recommended one there was of that type.  Unfortunately, they always test those things for speed but not for reliability.  I've gotten burned with similar things too.

I find now, it is always best to wait after a brand new chipset comes out for at least 4 months or so to let the chipset BIOS and drivers mature, and to see what issues might have cropped up in real world use.

It is a lesson that is hard to follow, though.

[marko]

UPDATE:
Things got worse.
Just as I was beginning to relax, believing my system was back to stability and reliability, it starts losing drives again.
Errors galore, can't flush system data, delayed write failures, drives not responding, paths not found....

I ran the P33 BIOS update as provided by eVga, and if anything, that only served to make matters worse.

There's nothing worse than a system you have no confidence in (well, there probably is, but you know what I mean) so, a new motherboard was the order of the day. With help from glynor (thank you) I have replaced the motherboard with a Gigabyte Intel P45 model.


System rebuilt, windows installed from scratch, so far so good...

-marko

[glynor]

So often... Intel chipsets might not be the fastest, and they might not have the newest and fanciest features, but darn it if they don't "Just Work" and keep on "Just Working".

Glad you got her going again.  Random drive problems are terrible to try to diagnose.

[glynor]

I'm also partial to Gigabyte boards.  The blue color is soothing, I think. 

(Certainly much better than the DFI horrorshow.)

[hit_ny]

Not ASUS  ?

Marko, other than being stable, how does the new board 'feel', compared with the previous one ?