My system (both F38 and now F39) reboots itself every morning at 8am, yet nothing in the cron configuration is telling it to do that and I don't see anything obvious in the journal to cause it.
How can I figure out what is triggering this? I know the description is vague, but there it is.
poc
On 14 Nov 2023, at 13:52, Patrick O'Callaghan pocallaghan@gmail.com wrote:
My system (both F38 and now F39) reboots itself every morning at 8am, yet nothing in the cron configuration is telling it to do that and I don't see anything obvious in the journal to cause it.
How can I figure out what is triggering this? I know the description is vague, but there it is.
Check for systemd timers as well. systemctl list-timers.
Does the journal show system shutdown was requested?
Barry
poc _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
On Tue, 2023-11-14 at 14:26 +0000, Barry wrote:
On 14 Nov 2023, at 13:52, Patrick O'Callaghan pocallaghan@gmail.com wrote:
My system (both F38 and now F39) reboots itself every morning at 8am, yet nothing in the cron configuration is telling it to do that and I don't see anything obvious in the journal to cause it.
How can I figure out what is triggering this? I know the description is vague, but there it is.
Check for systemd timers as well. systemctl list-timers.
None of them are scheduled for near that time.
Does the journal show system shutdown was requested?
No. Here is a snippet of today's journal:
Nov 14 07:51:03 Bree systemd[1]: NetworkManager-dispatcher.service: Deactivated successfully. Nov 14 07:51:17 Bree systemd[1]: plocate-updatedb.service: Deactivated successfully. Nov 14 07:51:17 Bree systemd[1]: Finished plocate-updatedb.service - Update the plocate database. Nov 14 07:51:17 Bree systemd[1]: plocate-updatedb.service: Consumed 9.667s CPU time. Nov 14 07:52:29 Bree sshd[73704]: fatal: Timeout before authentication for 43.153.185.216 port 51704 -- Boot c04ce0f9f0b94587b88a1f8c0a2c4659 -- Nov 14 08:00:27 Bree kernel: microcode: updated early: 0x19 -> 0x21, date = 2019-02-13 Nov 14 08:00:27 Bree kernel: Linux version 6.5.11-300.fc39.x86_64 (mockbuild@d23353abed4340e492bce6e111e27898) (gcc (GCC) 13.2.1 20231011 (Red Hat 13.2.1-4), GNU ld version 2.40-1> Nov 14 08:00:27 Bree kernel: Command line: BOOT_IMAGE=(hd2,gpt2)/vmlinuz-6.5.11-300.fc39.x86_64 root=UUID=8e1f7af4-c0bf-434e-b1c4-a9af2c810d56 ro rootflags=subvol=root modprobe.bl> Nov 14 08:00:27 Bree kernel: BIOS-provided physical RAM map:
The last entry before reboot is irrelevant (I run a small web server and fail2ban is constantly fending off attempted logins, mainly from China.
Every day I'm getting the same thing at the same time.
Full disclosure: I have a nightly hibernation/resume procedure which wakes the machine at 07:50, using a smart power plug. The journal shows that this is working as expected, and AFAIK is unrelated to the reboot. I've triple-checked the smart plug settings and toggled it manually to make sure.
poc
That is not a reboot. That is a crash/power off or something else.
You might move your wake up the machine to say 7:40 and see if the machine still does it at 8am or then does it at 7:50 (10 min after reboot).
If it does it 10 minutes after reboot that is often the default for the watchdog reset timer that will force power the machine (in bios) if the OS does not touch the watchdog every so often.
If the machine has the watchdog enabled, there is a decent chance it is not getting correctly restarted after hibernation.
On Tue, Nov 14, 2023 at 10:28 AM Patrick O'Callaghan pocallaghan@gmail.com wrote:
On Tue, 2023-11-14 at 14:26 +0000, Barry wrote:
On 14 Nov 2023, at 13:52, Patrick O'Callaghan pocallaghan@gmail.com wrote:
My system (both F38 and now F39) reboots itself every morning at 8am, yet nothing in the cron configuration is telling it to do that and I don't see anything obvious in the journal to cause it.
How can I figure out what is triggering this? I know the description is vague, but there it is.
Check for systemd timers as well. systemctl list-timers.
None of them are scheduled for near that time.
Does the journal show system shutdown was requested?
No. Here is a snippet of today's journal:
Nov 14 07:51:03 Bree systemd[1]: NetworkManager-dispatcher.service: Deactivated successfully. Nov 14 07:51:17 Bree systemd[1]: plocate-updatedb.service: Deactivated successfully. Nov 14 07:51:17 Bree systemd[1]: Finished plocate-updatedb.service - Update the plocate database. Nov 14 07:51:17 Bree systemd[1]: plocate-updatedb.service: Consumed 9.667s CPU time. Nov 14 07:52:29 Bree sshd[73704]: fatal: Timeout before authentication for 43.153.185.216 port 51704 -- Boot c04ce0f9f0b94587b88a1f8c0a2c4659 -- Nov 14 08:00:27 Bree kernel: microcode: updated early: 0x19 -> 0x21, date = 2019-02-13 Nov 14 08:00:27 Bree kernel: Linux version 6.5.11-300.fc39.x86_64 (mockbuild@d23353abed4340e492bce6e111e27898) (gcc (GCC) 13.2.1 20231011 (Red Hat 13.2.1-4), GNU ld version 2.40-1> Nov 14 08:00:27 Bree kernel: Command line: BOOT_IMAGE=(hd2,gpt2)/vmlinuz-6.5.11-300.fc39.x86_64 root=UUID=8e1f7af4-c0bf-434e-b1c4-a9af2c810d56 ro rootflags=subvol=root modprobe.bl> Nov 14 08:00:27 Bree kernel: BIOS-provided physical RAM map:
The last entry before reboot is irrelevant (I run a small web server and fail2ban is constantly fending off attempted logins, mainly from China.
Every day I'm getting the same thing at the same time.
Full disclosure: I have a nightly hibernation/resume procedure which wakes the machine at 07:50, using a smart power plug. The journal shows that this is working as expected, and AFAIK is unrelated to the reboot. I've triple-checked the smart plug settings and toggled it manually to make sure.
poc _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
On Tue, 2023-11-14 at 10:36 -0600, Roger Heflin wrote:
That is not a reboot. That is a crash/power off or something else.
You might move your wake up the machine to say 7:40 and see if the machine still does it at 8am or then does it at 7:50 (10 min after reboot).
If it does it 10 minutes after reboot that is often the default for the watchdog reset timer that will force power the machine (in bios) if the OS does not touch the watchdog every so often.
If the machine has the watchdog enabled, there is a decent chance it is not getting correctly restarted after hibernation.
Good idea. I'll try that and see what happens.
poc
On Tue, 2023-11-14 at 10:36 -0600, Roger Heflin wrote:
That is not a reboot. That is a crash/power off or something else.
You might move your wake up the machine to say 7:40 and see if the machine still does it at 8am or then does it at 7:50 (10 min after reboot).
If it does it 10 minutes after reboot that is often the default for the watchdog reset timer that will force power the machine (in bios) if the OS does not touch the watchdog every so often.
If the machine has the watchdog enabled, there is a decent chance it is not getting correctly restarted after hibernation.
Tried that, but the machine still boots at 08:00. I logged the watchdog status in the hibernate resume script:
Wed 15 Nov 07:40:46 GMT 2023 Device: /dev/watchdog0 Identity: iTCO_wdt [version 2] Timeout: 30 seconds Timeleft: 2 seconds Pre-timeout: 0 seconds FLAG DESCRIPTION STATUS BOOT-STATUS KEEPALIVEPING Keep alive ping reply 1 0 MAGICCLOSE Supports magic close char 0 0 SETTIMEOUT Set timeout (in seconds) 0 0
Interestingly, the running watchdog status (after the forced boot) has a different identity field:
$ wdctl Device: /dev/watchdog0 Identity: iamt_wdt [version 1] Timeout: 120 seconds FLAG DESCRIPTION STATUS BOOT-STATUS KEEPALIVEPING Keep alive ping reply 1 0 SETTIMEOUT Set timeout (in seconds) 0 0 ALARMONLY Not trigger reboot 0 0
but I'm not sure that's relevant. Either way, the forced boot still happens at 08:00 and not after a 10-minute timeout, so the mystery continues.
poc
On Wed, 2023-11-15 at 12:17 +0000, Patrick O'Callaghan wrote:
Either way, the forced boot still happens at 08:00 and not after a 10-minute timeout, so the mystery continues.
Strategic attack kittens?
But being serious, is the hardware clock set correctly?
You might try not doing a hibernate and see if it still goes at 8am. That way you at least know it is not related to hibernate.
Cron, or something in the bios. I would have thought that if you had the bios set to power up the node at time X then it would be smart enough not to reboot/recycle the node if it was already up. But bioses are often badly written and/or tested and or badly deisgned and do less than ideal things.
And you checked all cron's? crontab -l and all /etc/cron.*/* files?
On Wed, Nov 15, 2023 at 12:43 PM Tim via users users@lists.fedoraproject.org wrote:
On Wed, 2023-11-15 at 12:17 +0000, Patrick O'Callaghan wrote:
Either way, the forced boot still happens at 08:00 and not after a 10-minute timeout, so the mystery continues.
Strategic attack kittens?
But being serious, is the hardware clock set correctly?
--
NB: All unexpected mail to my mailbox is automatically deleted. I will only get to see the messages that are posted to the list.
The following system info data is generated fresh for each post:
uname -rsvp Linux 6.2.15-100.fc36.x86_64 #1 SMP PREEMPT_DYNAMIC Thu May 11 16:51:53 UTC 2023 x86_64 _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
On Wed, 2023-11-15 at 15:02 -0600, Roger Heflin wrote:
You might try not doing a hibernate and see if it still goes at 8am. That way you at least know it is not related to hibernate.
I'll do that. I've looked over historical journal entries and this behaviour started around the end of July. It wasn't happening before then as far as I know.
Cron, or something in the bios. I would have thought that if you had the bios set to power up the node at time X then it would be smart enough not to reboot/recycle the node if it was already up. But bioses are often badly written and/or tested and or badly deisgned and do less than ideal things.
And you checked all cron's? crontab -l and all /etc/cron.*/* files?
I have. Nothing stands out. There is one entry that runs 'apachectl graceful' at 08:00 but it's hard to see that as being the cause of the problem (running it manually is innocuous).
poc
On Thu, 2023-11-16 at 05:12 +1030, Tim via users wrote:
On Wed, 2023-11-15 at 12:17 +0000, Patrick O'Callaghan wrote:
Either way, the forced boot still happens at 08:00 and not after a 10-minute timeout, so the mystery continues.
Strategic attack kittens?
But being serious, is the hardware clock set correctly?
As reported by hwclock, yes.
poc
Tim:
Strategic attack kittens?
But being serious, is the hardware clock set correctly?
Patrick O'Callaghan:
As reported by hwclock, yes
Is everything else fine at 8? You're not getting a mains power dive from something else coming on via a timer?
On Thu, 2023-11-16 at 15:46 +1030, Tim via users wrote:
Tim:
Strategic attack kittens?
But being serious, is the hardware clock set correctly?
Patrick O'Callaghan:
As reported by hwclock, yes
Is everything else fine at 8? You're not getting a mains power dive from something else coming on via a timer?
No, it was the "smart" plug. See my reply to Samuel.
poc
On 11/14/23 05:52, Patrick O'Callaghan wrote:
My system (both F38 and now F39) reboots itself every morning at 8am, yet nothing in the cron configuration is telling it to do that and I don't see anything obvious in the journal to cause it.
How can I figure out what is triggering this? I know the description is vague, but there it is.
As mentioned, that isn't a reboot. It's a hard power-off. Have you tried being at your computer at that time to see what it's doing?
On Wed, 2023-11-15 at 14:20 -0800, Samuel Sieb wrote:
On 11/14/23 05:52, Patrick O'Callaghan wrote:
My system (both F38 and now F39) reboots itself every morning at 8am, yet nothing in the cron configuration is telling it to do that and I don't see anything obvious in the journal to cause it.
How can I figure out what is triggering this? I know the description is vague, but there it is.
As mentioned, that isn't a reboot. It's a hard power-off. Have you tried being at your computer at that time to see what it's doing?
Once I've tried Barry's suggestion of turning off the programmed hibernation, that'll be the next thing. It's usually a bad idea to change more that one variable at a time.
poc
On Wed, 2023-11-15 at 23:02 +0000, Patrick O'Callaghan wrote:
On Wed, 2023-11-15 at 14:20 -0800, Samuel Sieb wrote:
On 11/14/23 05:52, Patrick O'Callaghan wrote:
My system (both F38 and now F39) reboots itself every morning at 8am, yet nothing in the cron configuration is telling it to do that and I don't see anything obvious in the journal to cause it.
How can I figure out what is triggering this? I know the description is vague, but there it is.
As mentioned, that isn't a reboot. It's a hard power-off. Have you tried being at your computer at that time to see what it's doing?
Once I've tried Barry's suggestion of turning off the programmed hibernation, that'll be the next thing. It's usually a bad idea to change more that one variable at a time.
I should have said Roger's suggestion, not Barry's. Apologies.
poc
On 11/15/23 15:02, Patrick O'Callaghan wrote:
On Wed, 2023-11-15 at 14:20 -0800, Samuel Sieb wrote:
On 11/14/23 05:52, Patrick O'Callaghan wrote:
My system (both F38 and now F39) reboots itself every morning at 8am, yet nothing in the cron configuration is telling it to do that and I don't see anything obvious in the journal to cause it.
How can I figure out what is triggering this? I know the description is vague, but there it is.
As mentioned, that isn't a reboot. It's a hard power-off. Have you tried being at your computer at that time to see what it's doing?
Once I've tried Barry's suggestion of turning off the programmed hibernation, that'll be the next thing. It's usually a bad idea to change more that one variable at a time.
I don't see how watching your computer is changing a variable. :-)
On 11/15/23 17:10, Samuel Sieb wrote:
On 11/15/23 15:02, Patrick O'Callaghan wrote:
On Wed, 2023-11-15 at 14:20 -0800, Samuel Sieb wrote:
On 11/14/23 05:52, Patrick O'Callaghan wrote:
My system (both F38 and now F39) reboots itself every morning at 8am, yet nothing in the cron configuration is telling it to do that and I don't see anything obvious in the journal to cause it.
How can I figure out what is triggering this? I know the description is vague, but there it is.
As mentioned, that isn't a reboot. It's a hard power-off. Have you tried being at your computer at that time to see what it's doing?
Once I've tried Barry's suggestion of turning off the programmed hibernation, that'll be the next thing. It's usually a bad idea to change more that one variable at a time.
I don't see how watching your computer is changing a variable. :-)
Indeed, I believe any quantum effects from being observed can be disregarded here.
Or hey, maybe your computer has fantasies about becoming a quantum computer!
But seriously, you might check whether there is some BIOS power control option that is doing this. You could try changing the hardware clock to/from UTC and see if that affects the time when this occurs.
On Wed, 2023-11-15 at 21:53 -0600, Robert Nichols wrote:
On 11/15/23 17:10, Samuel Sieb wrote:
On 11/15/23 15:02, Patrick O'Callaghan wrote:
On Wed, 2023-11-15 at 14:20 -0800, Samuel Sieb wrote: > > On 11/14/23 05:52, Patrick O'Callaghan wrote: > > > > My system (both F38 and now F39) reboots itself > > > > every morning > > > > at > > > > 8am, > > > > yet nothing in the cron configuration is telling it > > > > to do > > > > that and > > > > I > > > > don't see anything obvious in the journal to cause > > > > it. > > > > > > > > How can I figure out what is triggering this? I > > > > know the > > > > description is > > > > vague, but there it is. > > > > As mentioned, that isn't a reboot. It's a hard power- > > off. > > Have you > > tried being at your computer at that time to see what > > it's > > doing?
Once I've tried Barry's suggestion of turning off the programmed hibernation, that'll be the next thing. It's usually a bad idea to change more that one variable at a time.
I don't see how watching your computer is changing a variable. :-)
Indeed, I believe any quantum effects from being observed can be disregarded here.
Or hey, maybe your computer has fantasies about becoming a quantum computer!
But seriously, you might check whether there is some BIOS power control option that is doing this. You could try changing the hardware clock to/from UTC and see if that affects the time when this occurs.
Thanks Bob, but it turned out to be the F**king smart plug. See my reply to Samuel.
poc
On Wed, 2023-11-15 at 15:10 -0800, Samuel Sieb wrote:
On 11/15/23 15:02, Patrick O'Callaghan wrote:
On Wed, 2023-11-15 at 14:20 -0800, Samuel Sieb wrote:
On 11/14/23 05:52, Patrick O'Callaghan wrote:
My system (both F38 and now F39) reboots itself every morning at 8am, yet nothing in the cron configuration is telling it to do that and I don't see anything obvious in the journal to cause it.
How can I figure out what is triggering this? I know the description is vague, but there it is.
As mentioned, that isn't a reboot. It's a hard power-off. Have you tried being at your computer at that time to see what it's doing?
Once I've tried Barry's suggestion of turning off the programmed hibernation, that'll be the next thing. It's usually a bad idea to change more that one variable at a time.
I don't see how watching your computer is changing a variable. :-)
Heisenberg?
Seriously though, in the end I did watch it with hibernation disabled, and sure enough the "smart" power plug turned itself off and on again at 08:00.
I've checked and rechecked the settings on the thing, and it positively, definitely, absolutely does *not* have a scheduled power cycle at that time. In fact I'd completely disabled it during the experiment (but of course left it connected), yet it still decided to do this entirely of its own volition.
In other words, it's a bug in the plug.
Given that I actually do want to reboot (i.e. resume) at 08:00, I could just hibernate as normal and let it do its thing, but I don't like magic solutions. I may decide to replace it if I can get a refund (or even if not - it was cheap).
I appreciate everyone's efforts to help with this, even if in the end it had nothing to do with Fedora, or hibernation, or my computer.
Thanks guys.
poc
On Thu, 2023-11-16 at 09:59 +0000, Patrick O'Callaghan wrote:
Seriously though, in the end I did watch it with hibernation disabled, and sure enough the "smart" power plug turned itself off and on again at 08:00.
I've checked and rechecked the settings on the thing, and it positively, definitely, absolutely does *not* have a scheduled power cycle at that time. In fact I'd completely disabled it during the experiment (but of course left it connected), yet it still decided to do this entirely of its own volition.
In other words, it's a bug in the plug.
I have a few of those things, thankfully none of mine have done that (using a widespread Tuya chipset). I won't use them for anything critical, or risky. Just a few light bulbs and a pedestal fan. I'd never trust them on something that might be a fire risk (heaters), or requires stability (computing equipment).
I wish I could hack them to be controlled by Linux in my LAN, rather than some server on the internet which might decide they're wasting too much money supporting hardware you bought a few years back. Since I do have a PC on 24/7 it's certainly feasible in my situation. And other people have things like Raspberry Pies that they're happy to leave one running all the time.
On Thu, 2023-11-16 at 22:15 +1030, Tim via users wrote:
On Thu, 2023-11-16 at 09:59 +0000, Patrick O'Callaghan wrote:
Seriously though, in the end I did watch it with hibernation disabled, and sure enough the "smart" power plug turned itself off and on again at 08:00.
I've checked and rechecked the settings on the thing, and it positively, definitely, absolutely does *not* have a scheduled power cycle at that time. In fact I'd completely disabled it during the experiment (but of course left it connected), yet it still decided to do this entirely of its own volition.
In other words, it's a bug in the plug.
I have a few of those things, thankfully none of mine have done that (using a widespread Tuya chipset). I won't use them for anything critical, or risky. Just a few light bulbs and a pedestal fan. I'd never trust them on something that might be a fire risk (heaters), or requires stability (computing equipment).
I wish I could hack them to be controlled by Linux in my LAN, rather than some server on the internet which might decide they're wasting too much money supporting hardware you bought a few years back. Since I do have a PC on 24/7 it's certainly feasible in my situation. And other people have things like Raspberry Pies that they're happy to leave one running all the time.
I had installed this thing as a simple solution to waking up the hibernating machine without having to futz around with Wake On Lan, which I have no experience with and which apparently can be flaky depending on what the BIOS supports. A more reliable device might be the answer.
poc
On 11/16/23 04:03, Patrick O'Callaghan wrote:
I had installed this thing as a simple solution to waking up the hibernating machine without having to futz around with Wake On Lan, which I have no experience with and which apparently can be flaky depending on what the BIOS supports. A more reliable device might be the answer.
Depending on the MCU inside and hardware accessibility, you might look into something like https://esphome.io/ to replace the firmware. I'm involved with that project and it's very nice for getting local control of your own devices.
On Thu, 2023-11-16 at 14:13 -0800, Samuel Sieb wrote:
On 11/16/23 04:03, Patrick O'Callaghan wrote:
I had installed this thing as a simple solution to waking up the hibernating machine without having to futz around with Wake On Lan, which I have no experience with and which apparently can be flaky depending on what the BIOS supports. A more reliable device might be the answer.
Depending on the MCU inside and hardware accessibility, you might look into something like https://esphome.io/%C2%A0to replace the firmware. I'm involved with that project and it's very nice for getting local control of your own devices.
That would seem to be way outside my comfort zone/area of competence.
poc
You might look in the bios. Some of the bioses have an option to turn on at a specific time each day.
On Thu, Nov 16, 2023 at 4:00 AM Patrick O'Callaghan pocallaghan@gmail.com wrote:
On Wed, 2023-11-15 at 15:10 -0800, Samuel Sieb wrote:
On 11/15/23 15:02, Patrick O'Callaghan wrote:
On Wed, 2023-11-15 at 14:20 -0800, Samuel Sieb wrote:
On 11/14/23 05:52, Patrick O'Callaghan wrote:
My system (both F38 and now F39) reboots itself every morning at 8am, yet nothing in the cron configuration is telling it to do that and I don't see anything obvious in the journal to cause it.
How can I figure out what is triggering this? I know the description is vague, but there it is.
As mentioned, that isn't a reboot. It's a hard power-off. Have you tried being at your computer at that time to see what it's doing?
Once I've tried Barry's suggestion of turning off the programmed hibernation, that'll be the next thing. It's usually a bad idea to change more that one variable at a time.
I don't see how watching your computer is changing a variable. :-)
Heisenberg?
Seriously though, in the end I did watch it with hibernation disabled, and sure enough the "smart" power plug turned itself off and on again at 08:00.
I've checked and rechecked the settings on the thing, and it positively, definitely, absolutely does *not* have a scheduled power cycle at that time. In fact I'd completely disabled it during the experiment (but of course left it connected), yet it still decided to do this entirely of its own volition.
In other words, it's a bug in the plug.
Given that I actually do want to reboot (i.e. resume) at 08:00, I could just hibernate as normal and let it do its thing, but I don't like magic solutions. I may decide to replace it if I can get a refund (or even if not - it was cheap).
I appreciate everyone's efforts to help with this, even if in the end it had nothing to do with Fedora, or hibernation, or my computer.
Thanks guys.
poc _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
Am 16.11.23 um 14:39 schrieb Patrick O'Callaghan:
On Thu, 2023-11-16 at 07:08 -0600, Roger Heflin wrote:
You might look in the bios. Some of the bioses have an option to turn on at a specific time each day.
The plug has no accessible BIOS, if that's what you mean.
I guess, the OP meant your PC.
Some PCs have some kind of a "timed wake up" in their BIOS.
Ralf
On Thu, 2023-11-16 at 19:09 +0100, Ralf Corsépius wrote:
Am 16.11.23 um 14:39 schrieb Patrick O'Callaghan:
On Thu, 2023-11-16 at 07:08 -0600, Roger Heflin wrote:
You might look in the bios. Some of the bioses have an option to turn on at a specific time each day.
The plug has no accessible BIOS, if that's what you mean.
I guess, the OP meant your PC.
Some PCs have some kind of a "timed wake up" in their BIOS.
Sorry if I wasn't clear. I'd already discarded that explanation as I could see the LED on the plug turning off and on again at the time, i.e. it was the plug itself that was causing the problem.
poc
No, we both mean use the bios timed wake up to wake up your machine instead of the plug for anything.
On Thu, Nov 16, 2023 at 4:14 PM Patrick O'Callaghan pocallaghan@gmail.com wrote:
On Thu, 2023-11-16 at 19:09 +0100, Ralf Corsépius wrote:
Am 16.11.23 um 14:39 schrieb Patrick O'Callaghan:
On Thu, 2023-11-16 at 07:08 -0600, Roger Heflin wrote:
You might look in the bios. Some of the bioses have an option to turn on at a specific time each day.
The plug has no accessible BIOS, if that's what you mean.
I guess, the OP meant your PC.
Some PCs have some kind of a "timed wake up" in their BIOS.
Sorry if I wasn't clear. I'd already discarded that explanation as I could see the LED on the plug turning off and on again at the time, i.e. it was the plug itself that was causing the problem.
poc
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
On Thu, 2023-11-16 at 16:25 -0600, Roger Heflin wrote:
No, we both mean use the bios timed wake up to wake up your machine instead of the plug for anything.
Ah, I misunderstood. I hadn't realised that was even possible.
For now, I've factory reset the plug and will see what happens tomorrow morning. If the problem persists, I'll look into the BIOS option.
Thanks.
poc
I checked a mb bios from 10 years ago and it is in it.
And I checked a 2 year old one and both have it.
RTC Alarm Power On Allow the system to be waked up by the real time clock alarm. Set it to By OS to let it be handled by your operating system.
web page for setting the timer from linux: https://www.linux.com/training-tutorials/wake-linux-rtc-alarm-clock/
On Thu, Nov 16, 2023 at 4:29 PM Patrick O'Callaghan pocallaghan@gmail.com wrote:
On Thu, 2023-11-16 at 16:25 -0600, Roger Heflin wrote:
No, we both mean use the bios timed wake up to wake up your machine instead of the plug for anything.
Ah, I misunderstood. I hadn't realised that was even possible.
For now, I've factory reset the plug and will see what happens tomorrow morning. If the problem persists, I'll look into the BIOS option.
Thanks.
poc
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
On Thu, 2023-11-16 at 17:47 -0600, Roger Heflin wrote:
I checked a mb bios from 10 years ago and it is in it.
And I checked a 2 year old one and both have it.
Yes, I just verified that mine has it (and it's around 10 years old).
RTC Alarm Power On Allow the system to be waked up by the real time clock alarm. Set it to By OS to let it be handled by your operating system.
web page for setting the timer from linux: https://www.linux.com/training-tutorials/wake-linux-rtc-alarm-clock/
That's very helpful, thanks.
I may play with it, however the factory reset on the plug seems to have worked. 8am came and went without incident, so it looks like I can use the plug as intended. However the BIOS option is there if it proves unreliable.
poc
Roger Heflin wrote:
No, we both mean use the bios timed wake up to wake up your machine instead of the plug for anything.
Patrick O'Callaghan:
Ah, I misunderstood. I hadn't realised that was even possible.
Way back in the day of my brief torture of Win98SE, I used to explore the BIOS to see what options there were, and which stupid ones were preselected. Wake on timer has been around for a very long time.
I haven't a use for it, but various businesses use it if they don't leave PCs on 24/7. Have the main server wake up by itself before the staff lob in, then progressively booting up each client PC with a minutes delay stops having a sudden mains power load, and staggers things that poll the server for everything under the sun.
On Fri, 2023-11-17 at 14:50 +1030, Tim via users wrote:
Roger Heflin wrote:
No, we both mean use the bios timed wake up to wake up your machine instead of the plug for anything.
Patrick O'Callaghan:
Ah, I misunderstood. I hadn't realised that was even possible.
Way back in the day of my brief torture of Win98SE, I used to explore the BIOS to see what options there were, and which stupid ones were preselected. Wake on timer has been around for a very long time.
I haven't a use for it, but various businesses use it if they don't leave PCs on 24/7. Have the main server wake up by itself before the staff lob in, then progressively booting up each client PC with a minutes delay stops having a sudden mains power load, and staggers things that poll the server for everything under the sun.
The logic is clear. I just hadn't ever been aware of it, thanks.
poc
Given it is the same time each day, I am guessing "feature" rather than bug.
It may also be that the power plug resets/reboots itself once a day and that the reboot does a quick power blip that is quick enough to not matter for non-computer devices.
Device makers love to find and implement half-assed "solutions" to problems. I would guess they had some sort of memory leak (most software does) and will stop working after a week or 2, and instead of finding and fixing the leak chose to reboot it daily so that the leak is not an issue. And on a reboot they don't have a way to maintain power during the reboot and decide a quick blip is ok.
On Thu, Nov 16, 2023 at 4:00 AM Patrick O'Callaghan pocallaghan@gmail.com wrote:
On Wed, 2023-11-15 at 15:10 -0800, Samuel Sieb wrote:
On 11/15/23 15:02, Patrick O'Callaghan wrote:
On Wed, 2023-11-15 at 14:20 -0800, Samuel Sieb wrote:
On 11/14/23 05:52, Patrick O'Callaghan wrote:
My system (both F38 and now F39) reboots itself every morning at 8am, yet nothing in the cron configuration is telling it to do that and I don't see anything obvious in the journal to cause it.
How can I figure out what is triggering this? I know the description is vague, but there it is.
As mentioned, that isn't a reboot. It's a hard power-off. Have you tried being at your computer at that time to see what it's doing?
Once I've tried Barry's suggestion of turning off the programmed hibernation, that'll be the next thing. It's usually a bad idea to change more that one variable at a time.
I don't see how watching your computer is changing a variable. :-)
Heisenberg?
Seriously though, in the end I did watch it with hibernation disabled, and sure enough the "smart" power plug turned itself off and on again at 08:00.
I've checked and rechecked the settings on the thing, and it positively, definitely, absolutely does *not* have a scheduled power cycle at that time. In fact I'd completely disabled it during the experiment (but of course left it connected), yet it still decided to do this entirely of its own volition.
In other words, it's a bug in the plug.
Given that I actually do want to reboot (i.e. resume) at 08:00, I could just hibernate as normal and let it do its thing, but I don't like magic solutions. I may decide to replace it if I can get a refund (or even if not - it was cheap).
I appreciate everyone's efforts to help with this, even if in the end it had nothing to do with Fedora, or hibernation, or my computer.
Thanks guys.
poc _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
On Thu, 2023-11-16 at 07:15 -0600, Roger Heflin wrote:
Given it is the same time each day, I am guessing "feature" rather than bug.
It may also be that the power plug resets/reboots itself once a day and that the reboot does a quick power blip that is quick enough to not matter for non-computer devices.
Device makers love to find and implement half-assed "solutions" to problems. I would guess they had some sort of memory leak (most software does) and will stop working after a week or 2, and instead of finding and fixing the leak chose to reboot it daily so that the leak is not an issue. And on a reboot they don't have a way to maintain power during the reboot and decide a quick blip is ok.
I've emailed the manufacturer (TP-Link), though without much expectation of a reply.
If it is as you say, it's a design fault. The power cycle is a definite Off, wait a second or so, then On.
poc
On Thu, Nov 16, 2023 at 8:43 AM Patrick O'Callaghan pocallaghan@gmail.com wrote:
I've emailed the manufacturer (TP-Link), though without much expectation of a reply.
If it is as you say, it's a design fault. The power cycle is a definite Off, wait a second or so, then On.
I have TP-Link KASA EP10 smart plugs and I don't see this behavior, and at least one plug controls a device that would definitely be impacted by such a power cycle.
For the sake of completeness you might double/triple/quadruple check that the plug doesn't have a schedule programmed. Assuming it doesn't have its own schedule, is there any chance the plug is a member of a group/scene/etc such that while the *plug* may not have a schedule, it is impacted by something else that does? Maybe try a factory reset on the plug (assuming there is such a reset)?
On Thu, 2023-11-16 at 11:04 -0500, Go Canes wrote:
On Thu, Nov 16, 2023 at 8:43 AM Patrick O'Callaghan pocallaghan@gmail.com wrote:
I've emailed the manufacturer (TP-Link), though without much expectation of a reply.
If it is as you say, it's a design fault. The power cycle is a definite Off, wait a second or so, then On.
I have TP-Link KASA EP10 smart plugs and I don't see this behavior, and at least one plug controls a device that would definitely be impacted by such a power cycle.
This is a PM110, which is a low-end device. I'll take a look at the EP10 as a possible alternative. Of course it's always possible that I have a faulty one,
For the sake of completeness you might double/triple/quadruple check that the plug doesn't have a schedule programmed.
I already did that.
Assuming it doesn't have its own schedule, is there any chance the plug is a member of a group/scene/etc such that while the *plug* may not have a schedule, it is impacted by something else that does?
There are no other smart plugs on the network. There are several Amazon Echo devices, which I think also implement the Matter protocol and can in theory talk to the TP-Link, but I've never configured them to do this and they don't have any scheduled activities.
Maybe try a factory reset on the plug (assuming there is such a reset)?
The docs say there is a factory reset, so I might try that.
poc
On Tue, 2023-11-14 at 13:52 +0000, Patrick O'Callaghan wrote:
My system (both F38 and now F39) reboots itself every morning at 8am, yet nothing in the cron configuration is telling it to do that and I don't see anything obvious in the journal to cause it.
How can I figure out what is triggering this? I know the description is vague, but there it is.
Just to close this off:
- The problem was being caused by a flaky smart plug - Restoring factory settings to the plug fixed it - I decided to follow some recommendations in the thread and use the built-in motherboard RTC instead of the smart plug as a way to re- activate the system at a set time. This is now working (touch wood).
My hibernation script now includes this snippet, if anyone's interested:
set_wakeup() { # See https://www.linux.com/training-tutorials/wake-linux-rtc-alarm-clock/ if [[ $(date +"%H") -lt 8 ]]; then datestring=$(date '+%s' -d '08:00 today') else datestring=$(date '+%s' -d '08:00 tomorrow') fi echo "$datestring" > /sys/class/rtc/rtc0/wakealarm }
Thanks again to all who chipped in.
poc
On Sat, 2023-11-18 at 12:59 +0000, Patrick O'Callaghan wrote:
On Tue, 2023-11-14 at 13:52 +0000, Patrick O'Callaghan wrote:
My system (both F38 and now F39) reboots itself every morning at 8am, yet nothing in the cron configuration is telling it to do that and I don't see anything obvious in the journal to cause it.
How can I figure out what is triggering this? I know the description is vague, but there it is.
Just to close this off:
- The problem was being caused by a flaky smart plug
- Restoring factory settings to the plug fixed it
- I decided to follow some recommendations in the thread and use the
built-in motherboard RTC instead of the smart plug as a way to re- activate the system at a set time. This is now working (touch wood).
My hibernation script now includes this snippet, if anyone's interested:
Corrected version. It turns out that the wakeup alarm has to be reset before setting:
set_wakeup() { # See https://www.linux.com/training-tutorials/wake-linux-rtc-alarm-clock/ if [[ $(date +"%H") -lt 8 ]]; then datestring=$(date '+%s' -d '08:00 today') else datestring=$(date '+%s' -d '08:00 tomorrow') fi echo 0 > /sys/class/rtc/rtc0/wakealarm echo "$datestring" > /sys/class/rtc/rtc0/wakealarm }
-
Barry -
Go Canes -
Patrick O'Callaghan -
Ralf Corsépius -
Robert Nichols -
Roger Heflin -
Samuel Sieb -
Tim