I rebuilt the HTPC a month or so ago, and it's been acting a little funny ever since. Stuff seems to record just fine just about all of the time, except it will reliably crash on Monday and Thursday nights. It's not being used at the time, everything is off but the HTPC itself, and even though it seems to be able to record any other day of the week flawlessly the recordings on Monday and Thursday nights seem to kill it. When I switch back over to the HTPC there's no video output (the HDMI can't establish a connection because the thing has crashed), so I'm forced to kill the power and bring it back online. Live TV always seems to work fine, and video playback (the bulk of it's usage) is flawless.
Does anyone have a suggestion on what to do? I've tried uninstalling and reinstalling the tuner drivers, and from what I can tell on the history it occurs on both of the NTSC tuners, on those days (QAM is used on Thursdays, but recordings occur at the same time as NTSC recordings). The RAM has passed multiple passes of memtest, and I'm positive the power is clean coming into the system.
Here's the setup:Gigabyte GA-H57M-USB3Intel i5 650Hauppauge HVR-1600 & HVR-1800 (set up for dual NTSC and dual QAM)2x2gb DDR3-1600 (running at 1333 with the 1600 timings)3xSeagate 1.5TB in 50GB primary and 2.7TB secondary RAID 5 arraysLite-On BD-ROMAntec Fusion, stock 430w PSU and LCD displayHK AVR-254 to InFocus X10 display7 Pro, 7 Codecs pack (Shark007) set up to use FFDshow and disable Windows Media foundation
Does anything show up in the event log?
Rob.
Win7, P5Q Pro Turbo, Q6600, GT430, BGT3595, Hauppauge Nova-hd-s2, DM500s, DVBLink.
www.thegreenbutton.tv
In terms of errors when the system goes down, none whatsoever; all events just stop in the neighborhood of when the crash occured. The best that I've got is an event when it comes back online saying that it lost power at (whatever time). In Media Center there are similar events for the partial recordings, saying that it recorded from (time) to (time) due to an unexpected loss of power, although it seems like it would probably give that error due to any hard crash.
Is this a new problem from when the last time you rebuilt it ? Do you have any scheduled tasks kicking off on any of these nights ?
Loss of power..hmmm..makes me think of a power supply issue (I have seen it before), as more tuners are in use, power usage goes up and causes a crash because of the lack of power. I could be wrong but, it could be the problem
-Dave
MCP, MCSA, MCSE 2003 Windows Vista Connected Exp:Home Theater for Technologists Windows Vista Connected Exp:Home Theater for Sales professionals
My Media Center Blog and fourms....
http://mc.anywherecool.com/Blog/
It's definitely a problem since I rebuilt, and there's nothing else going on; I went so far as to set the anti-virus to run in the early AM of Sunday and Wednesday and disable the update-checker of the built-in display, which probably wasn't even running. Besides a Blu-ray ripper from time to time, the only software that runs besides 7MC is AnyDVD, AVG, and the client software for the built-in display.
I considered that it might be a power issue too, but this doesn't quite fit the bill. When I rebuilt it, I was coming from a similar P5N7A-VM & E6850 setup, with two of the drives and both tuners. The chassis is new, but has the same model PSU as the prior, and I'm sure that even with the third hard drive being added I'm well below the power needed, not to mention below my previous power usage. It's an Antec 430W (nothing fancy), and I'm calculating the load to be below 300W. The weird part about it reporting a "power outage" is that the system at no time loses power. The lights stay on, including the LCD display (although it does freeze, which isn't surprising due to the lack of input).
I had considered that somehow my tuners had developed a problem, and was planning on removing them both and adding in the Ceton cablecard tuner but.. well, you guys know how well THAT plan panned out.
If a system were to have a hard crash, then be brought down, would it report a power failure like this? I'm accustomed to seeing real error logs, with at least something cryptic around the time when the system dies.
I just realized that I've only got the 4pin 12v aux connector plugged in, since I don't have the 8pin 12v aux connector. That shouldn't be a problem though, considering what little I've got plugged in, right?
**edit**I checked the manual, and it looks like I'm ok on the 8pin connector:"Use of a power supply providing a 2x4 12v power connector is recommended by the CPU manufacturer when using an Intel Extreme Edition CPU (130W)."
nmech123 I just realized that I've only got the 4pin 12v aux connector plugged in, since I don't have the 8pin 12v aux connector. That shouldn't be a problem though, considering what little I've got plugged in, right? **edit**I checked the manual, and it looks like I'm ok on the 8pin connector:"Use of a power supply providing a 2x4 12v power connector is recommended by the CPU manufacturer when using an Intel Extreme Edition CPU (130W)."
I just downloaded your motherboard manual and it says this:
• Use of a power supply providing a 2x4 12V power connector is recommended by the CPU manufacturer when using an Intel Extreme Edition CPU (130W).
I would suggest you get a supply that has a 2x4 12V power connector and at least 500W or greater. I realize that you are not running a Extreme Edition CPU but even my old Q6600 quad core CPU needed the 2X4 power connector. It's better to have a PSU that is over powered rather than under. Under powered PSU's will put a huge strain on your components and generate a lot of heat plus you could actually damage your CPU or motherboard. So if you only have only one 4 pin connector your CPU probably isn't getting all the power it needs. I could be wrong but a newer core i5 i would think would need more power. If your case has enough room for a full size power supply i would recommend a Corsair Power Supply. I have 2 of these and they are silent and rock solid. I just recently built my HTPC and my 2x4 plug had a cap on 1/2 of the plug. I removed the plug and plugged in both of the 12V CPU Power Connectors and my PC is running perfect.
Thanks for the research, but I'm pretty positive that it isn't a power supply sizing issue. I went back over the power supply sizer, courtesy of Antec, and they peg the current system at 280W (figuring in all of the drives, the fans, CPU, RAM, display and a total of four cards (I'm over-guesstimating for the pair of dual-tuner cards)), not to mention that the system that I had going before on the same model PSU was pegged at 288W. Granted, I'm sure that the 430W PSU isn't going to give me a full 430W, but even when the drives are going bananas my Kill-a-watt checks in around 100-110W (not volts, I checked) and around 90W idle. I suppose there's a possiblity that I've got a bum PSU, so I'll try to yank the old one out and install it in the new case, but the old HSF blocks the removal of the 4pin 12v cable so it may not happen for a bit.
nmech123 Thanks for the research, but I'm pretty positive that it isn't a power supply sizing issue. I went back over the power supply sizer, courtesy of Antec, and they peg the current system at 280W (figuring in all of the drives, the fans, CPU, RAM, display and a total of four cards (I'm over-guesstimating for the pair of dual-tuner cards)), not to mention that the system that I had going before on the same model PSU was pegged at 288W. Granted, I'm sure that the 430W PSU isn't going to give me a full 430W, but even when the drives are going bananas my Kill-a-watt checks in around 100-110W (not volts, I checked) and around 90W idle. I suppose there's a possiblity that I've got a bum PSU, so I'll try to yank the old one out and install it in the new case, but the old HSF blocks the removal of the 4pin 12v cable so it may not happen for a bit.
I did not mean a power supply sizing issue, I was meaning signs that the power supply is starting to fail. If the system worked fine before and a new problem after a rebuild, it tends to make me think, hardware is starting to fail or it's software based. I might try to swap like your saying to try it on the nights with problems.
Do you take backups ? What is the chance you can backup your current build and restore from your last build, to see if this problem is still there ?
There has to be a way to rule out it's not software, unless you want to try to rebuild to attempt this..(but, nothing seems like it would be the problem besides the PSU). If you deside to go this route, just run MC, no other software (no Anti-virus even), setup a night to record like crazy, like your problem nights and see if you can crash it.
I do stress tests on my system every once and a while... last week I did mine, recordee 4 HD channels(QAM) and 2 analog channels while watching a ripped high bandwith Blu-ray (26gb)...flawless. Maybe you can try this on your system now, to just stress test it a little....
That's the problem though, the rebuilt involved more than just the OS. Here's the transition:
Original:C2D E6850Asus P5N7A-VM8500GT (for Hybrid-SLI with onboard video)2x2gb DDR2-8002x1.5TB SeagateHVR-1600&1800 tunersSilver Antec Fusion, w/430W PSU7 Pro x64
Now:i5 650Gigabyte GA-H57M-USB3 (onboard video)2x2gb DDR2-1600 (running 1333)HVR-1600&1800 tuners3x1.5TB Seagate (same two from before, added a third for RAID 5)Black Antec Fusion, w/430W PSU (same BLOCKED EXPRESSION case/PSU, bought it new to fit my rack color scheme)7 Pro x64
So yeah, about the only thing that stayed the same when I rebuilt was the tuners and OS. I'm sure that the capture cards are good (they've survived 3 motherboards now), the array reports good, the memory tests good, so about all that's left for questionable hardware is the motherboard/cpu. The only backups I did before the rebuild was the media, since before it was a RAID 1 on an nVidia controller, and now a RAID 5 on an Intel controller, so it wouldn't have done me any good.
Tomorrow I'll fire up the projector (doesn't do much good to test without a real video output) and try and get beat it up like you mentioned. I'm pretty sure that it's the NTSC tuner that's making it crash, as QAM tuning doesn't seem to do much. What really bugs me is why it only happens on those two days! Two long-running NTSC programs recorded Friday of last week, two HD QAM programs recorded fine today.. but a single NTSC recording on Monday kills it. Weird.
Well, I got it going today, and the results are pretty inconclusive. Here are the tests:
1) 2 QAM tuners recording HD, Blu-ray rip running. 1hr OK2) 2 QAM tuners recording HD, 1 NTSC tuner recording, Blue-ray rip running. 45min OK3) 2 NTSC tuners recording, 1hr OK4) 2 QAM tuners recording HD, 2 NTSC tuners recording. In progress (25 min elapsed thus far)
At no time did the the temp of either of the cores get past 43c (according to Realtemp), and CPU utilization was about 33% when playing back the rip, below 10% for the tuner-only tests. I also had the Kill-a-watt hooked up, and the power usage at the plug was below 120W, averaging around 113W.
So, what's so special about Monday and Thursday night? I'm double checking the logs, and the actually error in Media Center is "While recording (whatever is recording), the computer lost power or experienced a temporary failure". The corresponding error in the event log is "The system has rebooted without cleanly shutting down first. This error could be caused if the system stopped respoding, crashed, or lost power unexpectedly." So yeah, I goofed it up by limiting the scope of the error to just the power early on, and I apologize for that.
Where do I go from here? I COULD reinstall Windows.. but man I really don't feel like it. Especially since the RAID 5 means the thing is dog slow moving files back on.
xegroeg For fun, is there anything in scheduled tasks under control panel? When troubleshooting crashes you want to check the event viewer for STOP messages. It may also be useful to use a debugger to read the dump files (if present) check http://support.microsoft.com/kb/315263 for the tools / instructions.
For fun, is there anything in scheduled tasks under control panel? When troubleshooting crashes you want to check the event viewer for STOP messages. It may also be useful to use a debugger to read the dump files (if present) check http://support.microsoft.com/kb/315263 for the tools / instructions.
Nope, no scheduled tasks. There are some active tasks, but it looks like a bunch of Microsoft stuff.
xegroegTry running perfmon /rel from a command prompt to open the reliability monitor, it may provide some insight to what's going on.
I took a look, and it's just a long list of reminders that the thing keeps failing. However, it did point out that it's not just Mondays and Thursdays!
In other very related news, the last test I did running all 4 tuners finished up fine (total time of around 70min), and then crashed around 10:22pm. One QAM tuner was active, along with two NTSC tuners, and died after about 27min of recording (I program in a five minute lead). So it isn't days of the week, as much as it's just piss-poor stability based on recording. I'm feeling pretty strongly about it being the NTSC though.