NVIDIA Boot Agent loads after Sage freezes; doesn't see hard drive

[popechild]

This could be related to the freezes I've discussed here, or it could just marginally be related. Sage has been freezing daily or more for me ever since around the time of the upgrade to v6.

Yesterday, for the first time, when Sage froze and I had to hard-reboot the system, instead of it booting back up fine it popped up a message about "Nvidia Boot Agent" trying to load and something about DHCP, then eventually it will say "PXE-E53: No boot filename received" and it stops without ever loading windows.

When I went rebooted and went into my BIOS, there were no harddrives listed at all, which initially made me think maybe it was a harddrive problem (I have two in the machine). But I read online about PXE meaning it was trying to boot from the network, so I disabled the onboard LAN in the BIOS and rebooted again, and this time it loaded windows fine, and reads both harddrives fine. Only problem is the network didn't work, so I had to uninstall and re-install the motherboard's network drivers. After I did that I was able to re-enable the LAN in the BIOS and all was back to normal.

Jump to 8 hours later. I wake up to find that Sage has frozen again. I re-boot again. It goes to the nvidia boot agent again. So this is apparently something that's going to happen every time it freezes now, which is not good.

The only change I've made recently has been a display driver update, but I've had a number of freezes since then without issue until yesterday, and the driver I'm on is generally reported to be pretty stable.

Does anyone have any idea what this nvidia boot agent thing is and why it's loading (and not seeing my harddrives to boot from) after the computer freezes? When everything's working right, my boot order is HD, CD, Floppy. When this boot agent thing pops up, my boot order is CD, Floppy, Boot Loader, with no harddrives listed.

Could this be a harddrive issue? Motherboard? Driver? Any other possible suggestions?

Thanks!

[GKusnick]

Originally Posted by popechild

So this is apparently something that's going to happen every time it freezes now, which is not good.

A more illuminating way to say it might be that it's going to freeze every time your hard drive goes missing. Similarly, your BIOS is going to try to boot from the net whenever it can't find a bootable hard drive. That part is normal.

The question is why is your drive disappearing? Could be a bad drive, a bad cable, a bad controller, or possibly a bad power supply. Swapping out the cable would be the first thing I'd try, if you have a spare handy. If not, try reseating the cable connectors to make sure they haven't worked loose. If you're using a PCI controller card instead of an onboard controller, try reseating that too. Try connecting the drive to a different power supply lead.

Check your drive manufacturer's website for diagnostic software and download and run that. The Seagate site has some generic diagnostics that should work on any drive.

[popechild]

I'm not trying to be difficult at all - in fact I would LOVE to find out that it's a bad hard drive, just so I could replace it and get past it! Help me understand something though. Would a bad/failing hard drive (or cable, connection, etc.) potentially only show itself while using Sage? Granted, I use Sage more than anything else on that box, and it's certainly more cpu intensive than anything else I do on it (web surfing, email, etc.), but my limited understanding would make me think that a failing hard drive/part would show itself across a variety of activities instead of just Sage playback. Is it normal that it might only show up when using Sage because it's more intensive? The boot drive isn't even the same drive that I record to/playback from.

Another thing that may or may not help: I noticed that when the "hard drive goes missing", I actually have two drives and they both go missing. When I go to the BIOS, there are no drives listed at all. Could this be because if the master is "missing" that it also wouldn't show the secondary? Or could this possibly point to a failing IDE cable that would make both disappear when it's not working?

One final note: I'm currently running the Seagate SeaTOOLS diagnostic on the boot drive. When I booted to the diagnostic tool, it saw both hard drives with no problem. I guess that makes me think it would not be the IDE cable.

[stanger89]

Originally Posted by popechild

Would a bad/failing hard drive (or cable, connection, etc.) potentially only show itself while using Sage?

PVR apps are one of the few apps that constantly exercise the disk. It's possible that it only drops out intermittently and that most of the time (when not using Sage) when it goes missing nothing notices and it comes back, but since Sage uses the drive more, it's more likely to "notice".

Another possibility is that Sage (or a decoder, windows, driver....) is not gracefully handling the drive disappearing.

Regardless, solving your drive disappearing issue is the first step.

You could try running... Well I was going to recommend defrag or scandisk, both would definitely exercise the drive, but I think both could fubar the volume if it vanishes in the middle. Maybe try running a video in WMP or something and see if it happens then.

Oh, couple more thoughts, yes, it's definitely conceivable that losing either master or slave will take out both. Also could be the HDD controller.

[popechild]

Originally Posted by stanger89

PVR apps are one of the few apps that constantly exercise the disk. It's possible that it only drops out intermittently and that most of the time (when not using Sage) when it goes missing nothing notices and it comes back, but since Sage uses the drive more, it's more likely to "notice".

Another possibility is that Sage (or a decoder, windows, driver....) is not gracefully handling the drive disappearing.

Regardless, solving your drive disappearing issue is the first step.

You could try running... Well I was going to recommend defrag or scandisk, both would definitely exercise the drive, but I think both could fubar the volume if it vanishes in the middle. Maybe try running a video in WMP or something and see if it happens then.

Oh, couple more thoughts, yes, it's definitely conceivable that losing either master or slave will take out both. Also could be the HDD controller.

Okay, I ran the Seagate diagnostic on both disks. The primary (boot) disk passed with flying colors. The secondary (recordings) disk found 2 bad sectors. I chose to repair them and both seemed to repair fine. Re-ran the test on that drive and it passes now.

I don't know if 2 bad sectors is bad or insignificant - probably depends on which sectors they are I guess. What are the odds that this could have been the problem?

One final thought: The one component that I KNOW has been having problems is my dvd drive. Basically just not always recognizing that there's a disk, or having problems successfully reading a disk. In order to run the diagnostics I've temporarily replaced it with a CD drive. I've been assuming it's just some sort of laser problem, and it's on it's own IDE channel, but is it possible that the problems I'm having with that drive are related to the freezing/missing disk issues?

Thanks so much for the help!

[stanger89]

Were the Seagate diagnostics run outside Windows? If so, it might be a driver issue.

[GKusnick]

If the BIOS isn't seeing the drives at boot time, that says to me that it's a hardware issue, not a driver issue.

[stanger89]

Yeah, but weird things can happen, especially on a warm start. It's conceivable at least that if the drivers in Windows fubar something and the drives get dropped, that the chipset/controler isn't reset properly and doesn't see the drives again until a proper restart.

Did the Seatools analysis take as long as it takes for the drives to disappear when Sage is running? If not, is there a longer diagnostic that can be run?

[popechild]

Originally Posted by stanger89

Were the Seagate diagnostics run outside Windows? If so, it might be a driver issue.

Yes, it was run off of a bootable cd.

Originally Posted by GKusnick

If the BIOS isn't seeing the drives at boot time, that says to me that it's a hardware issue, not a driver issue.

That's what's such a bugger about this little thing. It's not behaving consistently. For instance, I meantioned that if I disable the onboard LAN in the bios, I could re-start Windows, only without LAN. After I do that, I can re-install the mobo drivers (including LAN) and it still works fine. So right now for instance, it's running great, I can reboot, turn off/on, etc. and it all boots up fine every time and it sees the drives just fine. It's just that the last two times it's frozen, when I've tried to reboot, THEN it doesn't see either of the drives in the BIOS and tries to do the PXE boot. It seems so strange that it will recognize them no problem just like that by disabling the on-board LAN. And re-enabling the LAN once it's recognized them again doesn't automatically make them "disappear" again. Grrrr!

Originally Posted by stanger89

Yeah, but weird things can happen, especially on a warm start. It's conceivable at least that if the drivers in Windows fubar something and the drives get dropped, that the chipset/controler isn't reset properly and doesn't see the drives again until a proper restart.

Did the Seatools analysis take as long as it takes for the drives to disappear when Sage is running? If not, is there a longer diagnostic that can be run?

It's hard to say. It's definitely frozen before in the amount of time that it took the diagnostic to run, but it's also taken up to a day or two for it to freeze before. It's hard to pin down because it's only ever frozen during playback (live tv or pre-recorded).

For now, I've replaced the DVD drive with a CD drive, and I've fixed the two errors that the Seagate diagnostic found on the recording drive. We'll see how this goes and whether or not it freezes again, and if it does, if it still has the same problem finding the drives.

If this doesn't work, I may try moving my recordings folder to the boot drive (the one that didn't have any errors) and wait again to see if it freezes.

It's just so dang difficult to troubleshoot because I have to wait with each new try for up to a day or two to see if it will still freeze or not... Ahh, the joys of homebrew DVRs I guess.