Author |
Message |
heilig
|
Posted: Fri Aug 29, 2008 3:59 pm |
|
Joined: Sat May 12, 2007 2:41 am
Posts: 51
Location:
Southern Germany
|
Hi all,
sorry for being so quiet, I was on a holiday the last two weeks. We came back a few days ago, the myth box had been running all the time (with a forced reboot evey day at 05:00 am)
Since I had a few days off, i decided to give mythwelcome another try, and guess what happened: Suddenly it just works.
I have no idea what caused this. Before going on holidays, I was working on the issue, tested daily, and it never worked. Somehow something magical must have happened...
I made a number of modifications (I was desperate...) before going on holidays:
Downgraded the BIOS, reset all BIOS settings, upgraded BIOS again to the most recent version, set ACPI=force in lilo.conf, set HWCLOCKACCESS=no in hwclock.sh.
One of these things triggered it. I just don't know what exactly.
Thanks for all the hints here! Finally, our Knoppmyth box is working perfectly again, and WAF is back at 1.0
jens
|
|
Top |
|
|
heilig
|
Posted: Fri Aug 29, 2008 4:03 pm |
|
Joined: Sat May 12, 2007 2:41 am
Posts: 51
Location:
Southern Germany
|
tophee wrote: Unless I've missed it, now guess helper puts nvram-wakeup.conf in /etc automatically? Yes?
I did not when I tried it. Had to move it to /etc manually. tophee wrote: When I exit out of su mode and run Mythwelcome-test.sh I get an error. It reads: Code: sh: /tmp/alarm.log:Permission denied sh: /tmp/alarm.log:Permission denied nvram-wakeup retcode: 1Fatal: open /boot/map: Permission denied Execute the following command sudo reboot So sudo reboot gives me a system reboot. Running mythwelcome-test-wakeup.sh as root doesn't work either.
For nvram-wakeup mythwelcome-test-wakeup.sh must be run as root. When I tried it, it worked (well, almost: it set the wakeup time, but did not wake up...).
What error does it report?
jens
|
|
Top |
|
|
brfransen
|
Posted: Sun Sep 21, 2008 7:37 pm |
|
Joined: Fri Jul 21, 2006 11:12 pm
Posts: 1194
Location:
SC
|
I have a slight variation on this problem. My Dell's BIOS only shows hour and minutes for wakeup time. Using mythwelcome-test-wakeup.sh works correctly and will wake the machine and if I schedule something 30 mins out and then shutdown in MythWelcome the machine will boot at the correct time. The problem comes the next day. It doesn't boot for a scheduled wakeup time or scheduled recording. When I run mythwelcome-test-wakeup.sh, which worked the day before, it now doesn't start up. The only way to get it to start working again is to set the BIOS manually a few minutes ahead and then let it automatically boot. At that point mythwelcome-test-wakeup.sh will work again but just for that day. The next day it is hosed up again.
I am thinking that it has to do with the day but I haven't found any other reports of this or my googlefoo is lacking.
Any ideas?
Thanks,
Britney
|
|
Top |
|
|
alien
|
Posted: Mon Sep 22, 2008 1:20 am |
|
Joined: Mon Jun 21, 2004 5:28 am
Posts: 700
Location:
Germany
|
|
Top |
|
|
tophee
|
Posted: Mon Sep 22, 2008 3:00 am |
|
Joined: Tue Sep 13, 2005 10:48 am
Posts: 852
Location:
London, UK
|
Heilig,
I'd sort of forgoten about updating where I'd got to on this as it now works.
For me, there seemed to be an almost magical "suddenly it's working" moment too. After modifying the mythwelcome-test.sh with the touch command as described in the wiki I decided to start afresh.
After a quick backup, I did a fresh install of 5.5 and followed marc.aronson's suggestions on poking the BIOS and and adding the HWCLOCKACCESS=NO. (seen here)
It's been working fine for several weeks now.
I think the current WAF factor is at an all time high.
_________________ Version:R8 Intel C2D 7400, Nvidia 5600 via HDMI to Samsung B37B650TW (PAL), Asus P5QL-E mobo, 4Gb PC6400 DDR2 ram, Samsung Spinpoint 500 Gb & 1Tb drive, Nova-HD-S2 (x2)
|
|
Top |
|
|
brfransen
|
Posted: Mon Sep 22, 2008 8:51 am |
|
Joined: Fri Jul 21, 2006 11:12 pm
Posts: 1194
Location:
SC
|
alien wrote: Check out the last point in "Troubleshooting" at the bottom of: http://www.knoppmythwiki.org/index.php?page=MythWelcome
I do have a Daily Wake Period defined to have the box be on starting at 7am and shut off at 11:59pm. It is really only off for 7 hrs but it never wakes up.
Britney
|
|
Top |
|
|
alien
|
Posted: Mon Sep 22, 2008 9:32 am |
|
Joined: Mon Jun 21, 2004 5:28 am
Posts: 700
Location:
Germany
|
If I understand correctly, mythwelcome works fine for one day, but stops working at midnight? Have you tested with mythwelcome only. I.e. Schedule something for 4-5pm, 8-9pm and 3-4am. The box wakes up and shutdown for the first two but not the third? I wanted to clarify because there are some minor differences between the test and mythwelcome itself.
One thing that does cause wakeups to stop working on a box like this is a complete powerdown (turn off or unplug the power supply). I know this is a remote possibility, but do you have something that maybe interrupting power at night?
Otherwise, you might try nvram-wakeup. It does something closer to a manual wakeup which seems to be working with your box. ACPI seems to be hit-and-miss with some MBs.
_________________ ASUS AT3N7A-I (Atom 330) TBS 8922 PCI (DVB-S2)
|
|
Top |
|
|
brfransen
|
Posted: Mon Sep 22, 2008 10:55 am |
|
Joined: Fri Jul 21, 2006 11:12 pm
Posts: 1194
Location:
SC
|
Allen, thanks for the help.
I can't say that it is exactly at midnight but it will be working in the afternoon and then it will shut down at 11:59pm but then not wake up for the 7am wake time. I have not tried scheduling something at 3am but I will tonight.
The Dell does shutdown completely but it is connected to a UPS and power is not interrupted at night.
I played with nvram-wakeup a little bit yesterday but it doesn't know what to do with my MB. So I used the guess-helper.sh script but that didn't seem to work either.
Britney
|
|
Top |
|
|
alien
|
Posted: Tue Sep 23, 2008 2:12 am |
|
Joined: Mon Jun 21, 2004 5:28 am
Posts: 700
Location:
Germany
|
Hmmmm.... A couple of other things you might try:
The 11:59pm is kind of like Y2K. If this is a daily wakeup, you might try something not quite so close to midnight. Say 12:05 or 11:45.
Nvram-wakeup is depends on the BIOS version you are using. Changing your BIOS might allow it to work. However, don't just upgrade to the latest which may not be supported yet. Check the source of nvram-wakeup-mb.c at the bottom which lists the MBs and BIOS versions that it will detect. You can download the source from here:
http://svn.debian.org/wsvn/pkg-vdr-dvb/ ... rev=0&sc=0
_________________ ASUS AT3N7A-I (Atom 330) TBS 8922 PCI (DVB-S2)
|
|
Top |
|
|
pvr500
|
Posted: Wed Sep 24, 2008 8:49 pm |
|
Joined: Thu Nov 17, 2005 9:23 am
Posts: 18
|
marc.aronson wrote: alien wrote: marc.aronson wrote: I run a process that is launched at boot time. It wakes up every 2 minutes and measures the amount of CPU time used by the front end process. If the amount of CPU time does not change for 15 checks in a row, it declares that the front end is idle. Nice Idea! I'm going set this up. I use my mythbox primarily for a jukebox and we sometimes to forget to turn it off after the playlist finishes. Alien, I've found that the mythfrontend under r5.5 will use some CPU cycles even when idle. This was not the case in R5F27. On my machine it will use 2-5 seconds during a 2 minute idle period. So I've made two adjustments to my approach: 1. 6 seconds or less of CPU time during a 120 second sleep cycle is considered idle. 2. I monitor LIRC for input and treat the system as busy if any events are received. I do this by running "IRW > logfile" in the background. Any change to logfile's timestamp means that an input event has been received. This does not prevent myth from also seeing that same event. Marc
Hello Marc,
I am interested in your mfemon.sh program - do have an updated script with these changes?
Thanks
PVR500
|
|
Top |
|
|
brfransen
|
Posted: Fri Oct 03, 2008 11:52 am |
|
Joined: Fri Jul 21, 2006 11:12 pm
Posts: 1194
Location:
SC
|
I seem to have finally made ACPI wakeup to work reliably on my Dell Dimension 4600. I rolled the BIOS back from A12 to A10 and it is working great.
Britney
|
|
Top |
|
|
sb90
|
Posted: Wed Nov 05, 2008 9:13 am |
|
Joined: Thu Jan 18, 2007 10:33 am
Posts: 1
|
After upgrading to R5.5 ACPI wakeup continued to work well (after some reconfiguration), but after a few weeks it broke. After some head scratching and hacking about, I realised that what was required was a reset of the motherboard CMOS.
Simply make a note of any important non-standard settings, shutdown, disconnect power, remove backup battery or use the reset button on the board. Put everything back together and test. A few people have posted that re-installing, upgrading or downgrading the bios has worked for them but I think a CMOS reset might be all that is needed.
Steve.
|
|
Top |
|
|
nicom
|
Posted: Tue Jan 27, 2009 5:08 pm |
|
Joined: Mon Aug 14, 2006 8:24 pm
Posts: 320
Location:
Melbourne, Australia
|
I have just managed to get my ACPI wake working again after upgrading from R5F1. I have drawn heavily on the knowledge shared in this post so I will describe my experiences.
The machine woke correctly by ACPI using the old /proc/acpi/alarm method in R5F1.
Going through the grasky's procedure the machine failed to restart after the mythwelcome-test-wake script. I then used the troubleshooting hints in the wiki The machine would wake if the time was set manually in the bios. I had tried toggling the bios interface a couple of times to no effect. Increasing the test shutdown time to 2 hours also did not work. I then tried a number of changes suggested in this post:
1. Include ACPI=force in /etc/lilo.conf - Did not wake
2. Include HWCLOCKACCESS=no in /etc/init.d/hwclock.sh - Did not wake
3. Enable Wake System from S5 in bios -- Success
4. Remove HWCLOCKACCESS=no in /etc/init.d/hwclock.sh - Did not wake
5. Include HWCLOCKACCESS=no in /etc/init.d/hwclock.sh - Did not wake
6. Toggle Wake System from S5 to disable, reboot, then enable --Success
7. Manually set 5 recordings (5 minutes long) in Mythtv to occur overnight at hourly intervals and shutdown in Mythwelcome -- All recordings successful
What I conclude from this is that disabling the hwclock updates is the most important thing. What I am not sure of is what is the effect of disabling it. I see alien says the lack of updates could cause the clock to drift
Quote: Note that if you have to completely disable the hwclock updates, any drifts in the hwclock will never be corrected. You will probably have to correct the BIOS clock manually on occasion.
Is the "HWCLOCKACCESS=no" disabling updates completely or only when hwclock.sh is invoked at shutdown? If it does disable it completely, could it be updated by a cron job? If so, how?
_________________ Intel DG965WH, Dvico DVB-T Lite x2, HDHR, Gigabyte GT220, KingstonSSD, WD20EARS version=latest
|
|
Top |
|
|
alien
|
Posted: Wed Jan 28, 2009 4:56 am |
|
Joined: Mon Jun 21, 2004 5:28 am
Posts: 700
Location:
Germany
|
nicom wrote: Is the "HWCLOCKACCESS=no" disabling updates completely or only when hwclock.sh is invoked at shutdown? If it does disable it completely, could it be updated by a cron job? If so, how? HWCLOCKACCESS only affects hwclock.sh. You can try setting the HW clock in a cron job, but I suspect you will find it creates the same problem. (I may be wrong, so it is probably still worth a try)
If your HW clock is fairly accurate, it probably won't be too much of a problem. If you notice shows recording a little early or a little late, just update the time in the BIOS by hand.
_________________ ASUS AT3N7A-I (Atom 330) TBS 8922 PCI (DVB-S2)
|
|
Top |
|
|
marc.aronson
|
Posted: Wed Jan 28, 2009 12:58 pm |
|
Joined: Tue Jan 18, 2005 2:07 am
Posts: 1532
Location:
California
|
alien wrote: nicom wrote: Is the "HWCLOCKACCESS=no" disabling updates completely or only when hwclock.sh is invoked at shutdown? If it does disable it completely, could it be updated by a cron job? If so, how? HWCLOCKACCESS only affects hwclock.sh. You can try setting the HW clock in a cron job, but I suspect you will find it creates the same problem. (I may be wrong, so it is probably still worth a try) If your HW clock is fairly accurate, it probably won't be too much of a problem. If you notice shows recording a little early or a little late, just update the time in the BIOS by hand.
My old ASUS board requierd me to disable hardware updates. I don't recall the exact amount of drift, but after operating this way for 6 months it was enough of a problem that it effecting the timing of my recordings. My work-around was to re-sync the O/S clock on every boot up. The hardware clock still drifted, but I have my machine boot up 5-minutes before a recording is scheduled to start, so the drift wasn't enough to mess up the timing of the "wakeup from sleep", and with a forced re-sync at boot-up, the clock was always accurate by the time the myth back end loaded.
Marc
_________________ Marc
The views expressed are my own and do not necessarily reflect the views of my employer.
|
|
Top |
|
|