Premiere upload - crash - reboot, resume upload - crash - CPU lock up - crash

[netringer]

My Premiere has crashed 4 times in a 14 hour span. It's obviously got an overloaded CPU. What am I doing to cause this? Uploading saved content from my HP Windows Media server with MRV. In les than 30 minutes it freezes, goes black and does a restart. I'm getting to hate that dancing TiVo guy.

At first I thought it was due to doing inbound transfer at the same time that kmttg is downloading. I killed kmttg. Crash.

I know it's the CPU because even playing stored content stutters and stops and it stops responding to the remote. The Now Playing list crawls. It's apparently managing the transfer that it too much of a load to bear.

The only amazing thing is that it seems to resume the transfer on restart even before it displays the UI and the content transferred is apparently not corrupted.

I wonder if we can ever expect a code fix from TiVo. My money says no.

[aaronwt]

How can you be sure it's the CPU?
I don't have those issues with my Premieres and tiVo desktop or kmttg. I can be uploading from my TiVo server(with TiVo Desktop or Kmttg), recording two programs, and watching a recorded program all concurrently and with no problems. Maybe there is a network issue causing the problem or worse the Premiere is defective. All five of my Premieres have no issue doing this.

[Thuye]

Originally Posted by netringer

My Premiere has crashed 4 times in a 14 hour span. It's obviously got an overloaded CPU. What am I doing to cause this? Uploading saved content from my HP Windows Media server with MRV. In les than 30 minutes it freezes, goes black and does a restart. I'm getting to hate that dancing TiVo guy.

At first I thought it was due to doing inbound transfer at the same time that kmttg is downloading. I killed kmttg. Crash.

I know it's the CPU because even playing stored content stutters and stops and it stops responding to the remote. The Now Playing list crawls. It's apparently managing the transfer that it too much of a load to bear.

The only amazing thing is that it seems to resume the transfer on restart even before it displays the UI and the content transferred is apparently not corrupted.

I wonder if we can ever expect a code fix from TiVo. My money says no.

When I hear spontaneous reboots, I start thinking heat. Is the box well ventilated?

[jcthorne]

I would look at heat and a failing hard drive for the solution to this. Its not CPU load.

[netringer]

Originally Posted by Thuye

When I hear spontaneous reboots, I start thinking heat. Is the box well ventilated?

Originally Posted by jcthorne

I would look at heat and a failing hard drive for the solution to this. Its not CPU load.

Yeah. It just occurred to me before I read this to look at adding a fan or something. Weakknees doesn't have a twin breeze for Premiers.

What's working is putting it standby while it does the uploads.

[netringer]

Originally Posted by Thuye

When I hear spontaneous reboots, I start thinking heat. Is the box well ventilated?

Originally Posted by jcthorne

I would look at heat and a failing hard drive for the solution to this. Its not CPU load.

Originally Posted by netringer

Yeah. It just occurred to me before I read this to look at adding a fan or something. Weakknees doesn't have a twin breeze for Premiers.

What's working is putting it standby while it does the uploads.

Update. I can see the failure coming on. It started freeze the video and become unresponsive to the remote. I threw a small desk fan behind the Premier. Other than making *me* cold it's helping.

I can see the temp go from 35C at idle up to 36C and 37C doing the transfer.

I stuck a cheapie USB laptop cooler I have on hand underneath. The USB on the Premeire will power the fan.

With the cooler and external fan it's holding at 36C as I torture it as of this writing.

I ordered a premium cooler from Amazon that I'll try.

I'd like to add a thermal (speed!) control for the fan(s) at some point. I wonder if I can find the thermal switch that comes in the OWC enclosure I use.

I'll update on what works.

[aaroncgi]

Originally Posted by netringer

Update. I can see the failure coming on. It started freeze the video and become unresponsive to the remote. I threw a small desk fan behind the Premier. Other than making *me* cold it's helping.

I can see the temp go from 35C at idle up to 36C and 37C doing the transfer.

I stuck a cheapie USB laptop cooler I have on hand underneath. The USB on the Premeire will power the fan.

With the cooler and external fan it's holding at 36C as I torture it as of this writing.

I ordered a premium cooler from Amazon that I'll try.

I'd like to add a thermal (speed!) control for the fan(s) at some point. I wonder if I can find the thermal switch that comes in the OWC enclosure I use.

I'll update on what works.

Where are you measuring these temperatures? There's no decent hard drive that should have a problem with 37C indefinitely, and certainly not one rated for A/V usage as would be inside the Premiere, assuming that's the actual drive temperature. Now if that were 37C case temperature, one would expect the actual drive temperature to be much higher, which could cause a problem. A two degree temperature differential from idle (ie just recording the two tuners) vs your transfer load, doesn't seem problematic or unusual at all.

[netringer]

Originally Posted by aaroncgi

Where are you measuring these temperatures? There's no decent hard drive that should have a problem with 37C indefinitely, and certainly not one rated for A/V usage as would be inside the Premiere, assuming that's the actual drive temperature. Now if that were 37C case temperature, one would expect the actual drive temperature to be much higher, which could cause a problem. A two degree temperature differential from idle (ie just recording the two tuners) vs your transfer load, doesn't seem problematic or unusual at all.

It's the temperature listed in Account and System information.

I started getting freezes on playback and read the temp at the high of 37C.

I hacked a bit and put finally put my rotary desk fan so it blew air in where the vents are on the bottom of the left side as you view from the front. The case fan is on the right rear and blows out. I still had my minimal laptop cooler blowing air up from underneath.

In a few minutes I saw the temperature at 33C which was a low I have never seen before.

I'm doing back flips.








It locked up and crashed anyway during playback about 15 minutes later.

I have the 2TB WD AV drive installed. I dunno if the drive makes more heat or is more heat sensitive but this Premiere gets slow and is not stable when it does those inbound file transfers.

My only defense seems to be to let it cool down, preferably in standby, between transfer sessions.

[aaronwt]

37C is not very hot. The S3 boxes run much hotter and don't cause any issues. Although my Premieres are typically bewteen 33C and 36C.

Last Summer when I had water damage in my main viewing area there were dehumidifiers and fans running for several days straight. The temperature in the room rose to around ninety degrees and stayed there the entire time. I had three Premieres in there during that time. I have no idea what the temperatures were in the Premieres since I could not get to them but they recorded my shows daily and had no issues with me accessing the my shows list and transferring content to my Premieres in my other rooms.

I figure with my Premieres running between 33C and 36C in rooms between 65F and 70F, the Premieres had to be pretty warm when the room was 90F. And it didn't cause any problems.

[netringer]

I had slowdown-lock up-reboot last night at a cool 32C with no unusual loading going on.

I think my new 2TB drive is bad. I'm thinking it backs up read/write requests when it gets warm and the TiVo reboots when data isn't written in time.

I'm about to replace it. *sigh*

[jsjames]

Try running it without the network cable plugged in. I am on my 3rd Premiere and this one is rebooting all the time (like all others) but when I have my network cable unplugged, it seems fine.

[netringer]

Originally Posted by jsjames

Try running it without the network cable plugged in. I am on my 3rd Premiere and this one is rebooting all the time (like all others) but when I have my network cable unplugged, it seems fine.

Really? I'll have to try that. Back to my idea that the network transfer is what leads to the overload. Idea! I have a Gigabit Ethernet switch and this started when I brought the HP Mediasmart WHS server back to life. I wonder if it sends out too many Bonjour broadcasts. There is a lot of activity. I'll try unplugging it, too.

I'm only here at the computer because it just crashed again. I was only watching live while it recorded two shows. I was wondering how I can make it stop making new recordings. Even pulling the cable won't do it. I already disabled recording suggestions.

Yours is unmodded? I was thinking I'd replace the hard drive with another 2TB drive, but if that won't fix it. ...

[netringer]

My Premeire just stopped responding to the remote and did the freeze-reboot thing. No matter that I wanted the Amazon VOD download that was in hour 6 to finish.

Thanks to Rich's prodding, I took that restart as the opportunity to run all of the Kickstarts. I mistakenly began with Kickstart 5-4.

Guess what I got?


HMMMMMMMM??????

I ran Kickstart 5-7, then 5-4, then 5-8. They all went fairly quickly, considering I have a 2TB drive to check. I'm wondering if I didn't get into repair mode.
(I never saw the menu on 5-4 this time.)

I'll see how it holds up now. The way it's been going the next crash shouldn't be too far off if the problem is still there.

[netringer]

The software version is on my Series 4 Premiere is now 14.7-01-3-746

Is that new?

[richsadams]

I believe 14.7 came out late last year?

I'm not big on wagering, but I'd almost bet my very last dollar that you pressed 5-2 to get the installation screen pictured, not 5-4. KS54 would have started the SMART diagnostic.

With respect to the Kickstarts...52 simply installs the OS on the alternate boot partition. If it runs fine after that it's possible that the data was corrupt. That could be caused by a problematic download (although the check-sum feature should have caught it), a power issue (surge, brown-out, failure, etc.) or, gulp, the hard drive (bad sectors, head issue, etc.). If it is a hard drive issue it would likely mean that after the next update the problem may rear it's ugly head again...only time will tell...or you could run KS52 again. Now, all of that being said, what you described isn't likely to be caused by a corrupt boot partition, but more likely other problems.

KS 56 and 57 may or may not have done anything, it's very difficult to tell. If either do find corruption that they can repair, they do that. If it's more than they can repair TiVo simply reboots. In other words, if they ran only briefly it's possible they were able to repair problems (isolate bad sectors or whatever) quickly. It's also possible that they threw up their digital hands and said "whoa, that's above my pay grade!". Without studying the logs it's impossible to know what happened.

If you do start having trouble again, I know it would be a PIA, but you might try reinstalling your original hard drive and let it run for a while...see if anything goes south. If it does, it's probably something other than your new hard drive. If it doesn't...well...it's not rocket science as they say.

Hope things are good now...fingers crossed!

[netringer]

No joy.

It crashed and rebooted just because I was searching for the A/V receiver code for the remote.

I ran KS 5-7 a few cycles. KS 5-8.

This time I got some indication that the diags were running...and I'm an idiot because that "Installing Service Update" is one of the things you get.

It just froze up real good after leaving the remote setup screen but LO! This time it recovered and became responsive without crashing and rebooting. Could that be a ray of hope?

I suspect that I've had too many retries on my Amazon VOD download. I may have to call them to have it sent to the S3.

[netringer]

It's dead, Jim.

The glitches were gone enough that I had convinced myself that the Kickstarts must done some good unscrambling the inodes.

Then the last few nights I could tell by the freezes it was about to go. I put it in standby and amazingly was able to transfer shows out and watch them on the S3.

This morning it rebooted twice just playing Macbreak weekly TiVocast. (It looks like h.264 videos and uploads are the heavy tasks.)

I did a Kickstart 5-7 on the second reboot in hopes it would clean up the disk and it went into a GSOD loop. It would display the GSOD on the second "please wait" and crash. It did about 7 loops before I pulled the plug.

I put the original OEM hard drive back in. It turns out that it had been exactly one month since I swapped the drive.

It had no current guide data but it's been getting it - slowly. There have been some freezes and hangs but it's only because the disk is churning with the guide data and recordings to make.

I'll leave the factory drive in place for a week or two while I get the 2TB drive replaced and I upgrade again.

I lost a bunch of programs on the dead drive but a lot had been offloaded. I won't mourn too much.

[richsadams]

Originally Posted by netringer

It's dead, Jim.

Well, that's aggravating, but I guess it happens. That the original drive has had some freezes and hangs is a bit alarming though, even if it's doing some housekeeping. There may be something more to this but only time will tell. Keep us posted!

[netringer]

Originally Posted by richsadams

Well, that's aggravating, but I guess it happens. That the original drive has had some freezes and hangs is a bit alarming though, even if it's doing some housekeeping. There may be something more to this but only time will tell. Keep us posted!

Ya think because this drive was flopping around loose in the open, unsealed shipping box with the padding removed by whoever at FexEd Ground checks out my stuff see if it's worth stealing had anything to do with it? (I've had a slew of my shipment boxes arrive with the tape off or loose.)

I checked the drive and found that it was working OK and told myself it was fine. Thinking about it I got some glitches and hangs from the get-go but it worked fine and that's to be expected, e.g. this S4TP with the OEM drive hung right away but more severely stopped responding to the remote and froze up thoroughly with the moving spotlight screen after I forced a call last night. I let it live and turned off the TV and decided to see if came to life by this morning. It did.

The lesson re-learned is that loading and indexing the guide data is a heavy and disk intensive load so you need to be patient and take it easy in the first few hours.

As I mentioned my S3OLED rebooted once right after I upgraded to 1.3TB with a 1.5TB drive and was pushing it - also by playing a TWiT TiVocast. It's been rock solid since, he says, while he knocks on wood.

[richsadams]

Originally Posted by netringer

Ya think because this drive was flopping around loose in the open, unsealed shipping box with the padding removed by whoever at FexEd Ground checks out my stuff see if it's worth stealing had anything to do with it? (I've had a slew of my shipment boxes arrive with the tape off or loose.)

I checked the drive and found that it was working OK and told myself it was fine. Thinking about it I got some glitches and hangs from the get-go but it worked fine and that's to be expected, e.g. this S4TP with the OEM drive hung right away but more severely stopped responding to the remote and froze up thoroughly with the moving spotlight screen after I forced a call last night. I let it live and turned off the TV and decided to see if came to life by this morning. It did.

The lesson re-learned is that loading and indexing the guide data is a heavy and disk intensive load so you need to be patient and take it easy in the first few hours.

As I mentioned my S3OLED rebooted once right after I upgraded to 1.3TB with a 1.5TB drive and was pushing it - also by playing a TWiT TiVocast. It's been rock solid since, he says, while he knocks on wood.

I know hard drives are supposed to be almost impervious to being knocked about anymore, but I can't believe how they get shipped sometimes. Amazon's been doing a great job of packing them for a while and even Newegg has improved. You'd think after taking back so many that some of the other sellers would get a clue.

Running the extended diagnostic on a new drive isn't a guarantee that things will be good but it's saved me from installing a buggy drive more than a couple of times.

Hopefully life will be good soon!