Help resurrecting my Sage server..

[panteragstk]

Originally Posted by RedR

You can still use the win7 OEM ISO for testing, after all we're just trying to get it to hit the drive. Don't need the key for that I believe.
Also just as panteragstk suggested (I believe), try downloading a current Ubuntu ISO and see if they dies as well. The fact it dies in the same place each time is suggesting something. As mentioned earlier, SATA issue/config?

Yep, that's what I was saying. ANY OS will do, I just want to see if something else will try to install. You need a spare harddrive so you can try to see if it will complete an install. If it does, then something else may be wrong.

[Skirge01]

Originally Posted by osx-addict

Nope ... didn't try reseating the CPU -- got no thermal paste.. When I did build it I was using the Intel provided fan + paste.. I guess I should get some Artic silver..?

One more trip to Fry's I guess.. Anything else I should get while there? Also -- do I need to fully remove the old thermal paste before putting new stuff on? If so with what sort of 'cleaner'?

Edit : I read elsewhere that I can clean the old paste with alcohol and cotton swabs or just leave it and apply a VERY thin layer of new compound.

The Arctic Silver is what I use, but any of the "popular" ones are fine. Any comparison review I ever read always noted zero difference between the name brands. Definitely clean everything off, first, though. Since you're troubleshooting, you want to make certain you undo anything which might be causing your problem. 90+% alcohol and cotton swaps are what I always use. As panteragstk said, ensure it's dry. Although, alcohol dries extremely fast, so I doubt you'll have any issues. Just use another cotton swap to get the excess off and it'll be dry in minutes. Make sure you don't have any drops on the mobo!

[osx-addict]

Sounds good! Thanks guys for following in my pain here! I will let you know how it goes.. Regarding the ISO W7 image -- I will probably give it a shot with booting but without any drive installed -- just to see how far it gets.. If it gets far enough to complain about no media to do an installation on then that would be promising.. But, if it dies somewhere before that stage then...

P.S. Do either of you know whether those OEM Win7 licenses get tied to a particular motherboard (e.g. ethernet MAC address) or just an install? The reason I ask is that I think it might be time to move from XP to W7 but I'd hate to install on the old motherboard but find I can't move the entire installation to the new one if the need were to arise.. Know what I mean?

Now, I think I did read that there was some sort of way in W7 (once it's installed) to deactivate it so it can be moved to new HW -- not sure if that's true or not.

[panteragstk]

Originally Posted by osx-addict

Now, I think I did read that there was some sort of way in W7 (once it's installed) to deactivate it so it can be moved to new HW -- not sure if that's true or not.

You can "generalize" the install so you can keep the installed programs and move it to a new machine. Works pretty well. You will have to re-activate windows for the new machine, so give yourself the 30 days to test before moving on to the new motherboard.

[david1234]

Originally Posted by osx-addict

P.S. Do either of you know whether those OEM Win7 licenses get tied to a particular motherboard (e.g. ethernet MAC address) or just an install? The reason I ask is that I think it might be time to move from XP to W7 but I'd hate to install on the old motherboard but find I can't move the entire installation to the new one if the need were to arise.. Know what I mean?

Now, I think I did read that there was some sort of way in W7 (once it's installed) to deactivate it so it can be moved to new HW -- not sure if that's true or not.

The motherboard is supposed to be the one thing you can't replace with the OEM license. I haven't tested that it actually will stop working if you put it in a system with a new board... yet.

[osx-addict]

Thanks guys! I went to Fry's at lunch and picked up a new OEM W7 home premium pack for $90.. Will try it out this evening..

[Skirge01]

As far as the activation goes, you'll probably be fine. MS has been extremely generous with allowing activations of OEM versions on new computers. I did this just last year when my client HTPC's mobo died. I installed the new mobo, reinstalled W7 (even switched to the 64 bit version), hit activation... denied. It walked me through the automated phone activation and I was reactivated in just a few minutes. Now, will it let me do that again in the future? I have no idea! LOL!

[osx-addict]

I tried that too with mine and it denied it via the phone system.. Lesson learned I guess -- perhaps I answered a question wrongly..

[BobPhoenix]

I use this to clean the thermal paste up.

[osx-addict]

Thanks Bob! I'll probably order some of that for next time -- since I'll be monkeying around with this tonight.. I picked up some thermal paste at the local Fry's at lunch and even ordered (but not picked up yet) a Lenovo machine they had on sale for $258 until end of day today (it runs a Pentium G640 @2.8Ghz) -- going to be used as an HD300 replacement (in a trial-run setup)

[osx-addict]

Ok -- took the board out of the case, tried to remove the Intel CPU cooler -- got 2 of the 4 posts unlocked -- the other two wouldn't budge.. The pins were still sticking through the bottom.. I tried using various tools to push them back up through the holes but no luck.. After dicking around for about 40 minutes, I started cutting the plastic tabs off the bottom -- eventually I was able to remove the cooler -- albeit without the option of reusing it.. Stupid design.. Why cant they use small screws or something to hold them in place.

So, off to Fry's once again (2nd time today) -- picked up a CoolerMaster replacement cooler ($15) that would fit the LGA 1155 CPU.. Got it home -- it looked like it was direct from the factory -- not tampered with or a returned item (so common with many of their items at Fry's).. However, upon opening the parts bag it became obvious that all of the Intel mounting brackets were missing -- all 4 of them!

I'm starting to see a theme here -- seems like something is just not working in my favor!

Anyway, another day has come and gone.. I'll be heading over to Microcenter tomorrow AM to get the replacement board -- and will check with them that I can return it IF I find my cpu-reseating fixes the problem and I dont need it.

I'm hoping I can put this to rest sometime tomorrow.. Sheesh..

[davephan]

I would start the troubleshooting isolation process differently. You need to first isolate the problem as a software or hardware problem.

It would be much easier to take an image now, then recover back to an image before this problem began. If the computer functions normally, you know it was a software problem. If it does not, you know you have a hardware problem. You could also recover to a spare hard drive instead (to protect the data on the old hard drive). However, if the system works after the recovery, then the problem is isolated between the software and the old hard drive or the rest of the hardware. Recovering onto the same hard drive will isolate the problem as a software or hardware problem.

If you don't have an image of your computer in a "known good" state, then remove the hard drive and install a spare hard drive. Install the OS (plus probably over 100 patches) and SageTV on the spare hard drive. If the computer functions normally, you know it was a software problem (or a problem with old hard drive). If the problems continue, you know it is a hardware problem. The scratch rebuild process will probably take many hours or days. Your wasted time and downtime is the price you pay if you don't use imaging. The time savings could have easily paid for the imaging software many times over.

It is far easier to first determine if the problem is a software or hardware problem. Approaching it from the assumption that it is a hardware problem, without knowing that for sure, will cause you to try throwing things at the problem to fix it, like a power supply, memory, system board, CPU. The hardware might be good if you have a software problem.

This is precisely the reason for taking periodic images of the computer system. Since your household depends on the SageTV computer working with very little downtime, you cannot afford to operate the system without imaging and a quick way to recover the system.

If the problems are not resolved by recovering to a "known good" image (which could also be done on a spare hard drive to preserve the data on the old drive), then you can start throwing different hardware at the problem.

If you have a spare computer system, then you can take a "known good" image of your SageTV, and recover onto the spare hardware. Some imaging products support recoveries onto different hardware, some imaging products require identical or very similar hardware. I purchased two identical computer systems at the same time to give me this type of recovery option. One computer I use for my SageTV computer, the other I use for a general purpose, mostly web browsing computer. The spare hardware option gives you the ability to quickly resolve the "production" computer hardware problems. The hardware problems on the bad system can then be resolved over a longer time frame without effecting the "production" system. I have many other computers that could be used as general purpose systems, so that avoids downtime on the general purpose computer system too.

If you prepare for the recovery in advance, by imaging, and owning spare hardware, your downtime could be limited to about an hour, or several hours at the most. The downtime does not have to be days, weeks, or months. I've been doing disaster recovery preparation and actual recoveries at work for 15 years and longer at home. Your downtime can be reduced significantly if you are prepared to do the disaster recovery work in advance.

Dave

[osx-addict]

Ok.. Just a quick follow up and what I'm planning on doing differently in the future.. I am writing this post from the server in question that was having problems..

I was able to go visit Fry's this morning and exchanged the CoolerMaster CPU fan for another identical part -- luckily we opened that one up (the only one left on the shelf that wasnt a returned item -- it still had a fresh seal) -- it was missing the same Intel brackets as well.. So, got store credit and bought a different brand -- that one worked like a charm.. I installed it, and plugged everything back together (reinstalled mobo in case after vacuuming all dust from fans,etc).. Anyway, it booted windows right up..

Now, one caveat -- for whatever reason my brain had a major brain fart and I don't know WHY I recalled this machine running XP which it was NOT running -- it was W7 Home Premium.. I guess that is what you get when you almost NEVER login to the headless server -- remotely or otherwise.. I apparently need a major injection of Ginko Biloba to refresh my brain! Anyway, the server is happy as a clam and is running without issue.

For the future, I'll be getting Norton Ghost or similar to make a full disk image of the entire boot disk so I've got some backup should a similar disaster strike in the future..

Now, what caused the issue in the first place?? Who knows.. I do know that the stock Intel CPU cooler was looser than it probably ought to have been and perhaps was causing the CPU to misbehave over time?? Just my speculation.. Anyway, now I've got some spare parts at least (PSU, thermal paste,etc) and will be better prepared for next time..

Thanks all for all of the suggestions and thoughts on my problem.. I'm just happy to put it behind me..

[RedR]

If the thermal paste did the trick, then it was a overheat issue. Chances are you can stick w/ the stock cooler and use the paste you bought.

Side note, run something like Speedfan to help monitor the temp of the computer. This will give you insight now as to the normal running temp of the CPU/system (today), and can set off alarms when certain thresholds are exceeded (6-7 months down the road).
Glad its up and running again for ya!!

[osx-addict]

Thanks for the suggestions..

Interestingly enough, the problems are not completely gone.. I do think it may be something specific to the SATA stuff -- the boot drive on this machine is an SSD (64Gb) and it has crashed about 3 times this afternoon/evening.. Each time it did, it was unable to see the SSD and tried to boot the old drive containing the remnants of an old XP installation that I had when that drive was on another machine.. Of course that fails to load since most files are deleted now and it just hangs.. Once I reboot I go into the BIOS and the SSD is missing until a hard power-off..

Anyone see anything like this before? I'm thinking of ditching the SSD and use a spare 320Gb drive I've got lying around instead..

[davephan]

That's really strange! I never heard of that before. Maybe the drive is going bad. It might be worth wiping the drive with drive wiping software, then re-install everything from scratch, including the imaging software. Or, just replace the boot drive and install everything from scratch.

I bought 3 copies of Norton Ghost 15.0 on Amazon, since it was many times cheaper than the Symantec web site. The software purchased on Amazon instead of the Symantec web site worked and registered perfectly. I found Norton Ghost 15.0 for about $14 on Amazon a few minutes ago. If you bought 2 copies, or something else that qualifies for free shipping, to get the total over $25, you could get free shipping too. There's really old Norton Ghost versions for sale on Amazon too. I would avoid those old versions. Norton Ghost version 15.0 is very close to the GUI and features that Symantec System Recovery Server Edition 2011 that we use at work for servers.

http://www.amazon.com/Norton-Ghost-1...s=norton+ghost

Dave

[RedR]

Again, assuming you can keep the machine up long enough, download http://www.hdtune.com/ (the free ver is more than enough to find out what you need to know), or use the provided utility from the SSD to check the drives health. Sounds like the SSD is dying. One nice thing being 6-7 months old, chances are its still under warranty.

If you can get the drive to last long enough, use your 320gb drive, get the OS going and then try HDtune on the SSD. If it is dying, once you get it returned you can always put it back in place as the main OS drive.

[jerryt]

Disappearing SSD

1) I work think is a SSD firmware problem

or

Motherboard BIOS issue.

I would verify if there are later versions of these availalble and consider upgrading them.


2) Did you load the motherboard manufacturer's ACHI drivers or leave it using Microsoft's ACHI drivers????

Google how to add registry entries for both ACHI and IDE boot drive and try switching to IDE and see if it will last, then later you can go back to ACHI with the latest MB drivers

[osx-addict]

I'm using the MS drivers for AHCI -- I will look into the other issues as well and try downloading an AHCI driver from Asus and see how it plays.. I can get the machine running for perhaps a bit over 30-45 minutes before it dies.. It's currently powered off and I will look at it later today.. I'll keep you posted.

P.S. I checked the AHCI driver version -- I think it's generic MS item but not sure -- claims to be an Intel driver version 9.2.0.1011 -- the official Asus driver is entitled "Intel(R) Rapid Storage Technology Driver Software V10.6.0.1002 for Windows" -- much different versioning.. I'm downloading it now.. Could this sort of issue cause a problem with Memtest86 though after 10 hours? Seems unlikely..

P.S. Is there a site that goes over what all of the blue screen codes mean similar to the 7C one I had earlier? I saw it crash last night (took about 10 minutes after it stopped responding to mouse requests to blue screen and write a crash dump) and wasn't able to jot down the code before it rebooted.. It was NOT 7C (or whatever the earlier code was)..

P.P.S.. This sounds similar to my experience?!?!? Check this out -- I'm using an M4 Crucial SSD!

[jerryt]

I would certainly install the latest SSD firmware.

Be sure to copy the disc first.