Re: [DNG] Orange Pi Zero/Orange Pi PC
On Mon, 15 Jun 2020 21:35:40 +0100 tux...@sapo.pt wrote: > The devuan images are tested, > And so, I personally, advise always to stick with its kernels, but > anyone is free to mix what they want too :) The Armbian kernel works fine. Remember there is no Beowulf image yet for the OPi-Zero. > That in so many boards,... it could be a batch that went wrong, or > badly designed boards.. Must be someting like that. One of these board just stopped replying to pings and is not accessible through ssh, but each night it sends an email telling me there are updates for that board. I'm running Devuan Beowulf with an Armbian kernel and a "performance" governor. Someone said that switching to "performance" seems to resolve the issue. R. -- richard lucassen https://contact.xaq.nl/ ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] Orange Pi Zero/Orange Pi PC
Citando richard lucassen via Dng : No, even cronjobs do not work or give segmentation fault. It must be the hardware. There must be a batch with these issues, I'm not the only one BTW. I wonder if it has got something to do with the wifi chip. A crash occurs earlier if you load the kernel module. That seems to be some problem with the rtc oscillator or so..hardware probably( I got a problem with my rpiv2 with its 26Mhz oscillator, its similar in nature with what you describe.. The ASCII image is ok, but I updated this image to Beowulf and I used Armbian kernel with it. But I fear it will crash again. Anyway, I just got some Rasberry Pi Zero-W now and I think I will stay with these boards. Remember I have also Orange Pi PC, different CPU, suffering from the same issues. The devuan images are tested, And so, I personally, advise always to stick with its kernels, but anyone is free to mix what they want too :) That in so many boards,... it could be a batch that went wrong, or badly designed boards.. ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] Orange Pi Zero/Orange Pi PC
On Fri, 12 Jun 2020 04:41:17 -0700 Rick Moen via Dng wrote: > > Yes and no: > > > > -n Use SNTP (old RFC 2030, currently RFC 5905) instead of the RFC > > 868 time protocol. > > > > # rdate -n ntp.xs4all.nl > > Fri Jun 12 12:07:04 CEST 2020 > > Point taken. SNTP _is_ worthy of respect -- but I'd personally rather > have and use full NTP tools, and, given that the later are around and > standard, I don't also need one that does just SNTP (or that plus RFC > 868 Time Protocol). I run openntpd and on boards without RTC I made a /etc/init.d/setboottime using rdate that is started right after the network comes up. After that openntpd and all other scripts find a clock that is not at 1-1-1970 01:00. I know openntpd has an option to set the clock before it starts but at that time I had some problems with it so I stayed with rdate. I used to use ntp and ntpdate -B, but as I switched to openntpd I resolved the non-RTC issue with rdate. Remember that is just the clock of a small board, not a stratum 1 server ;-) R. -- richard lucassen https://contact.xaq.nl/ ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] Orange Pi Zero/Orange Pi PC
Quoting richard lucassen (mailingli...@lucassen.org): > Yes and no: > > -n Use SNTP (old RFC 2030, currently RFC 5905) instead of the RFC > 868 time protocol. > > # rdate -n ntp.xs4all.nl > Fri Jun 12 12:07:04 CEST 2020 Point taken. SNTP _is_ worthy of respect -- but I'd personally rather have and use full NTP tools, and, given that the later are around and standard, I don't also need one that does just SNTP (or that plus RFC 868 Time Protocol). ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] Orange Pi Zero/Orange Pi PC
On Fri, 12 Jun 2020 01:19:50 -0700 Rick Moen via Dng wrote: > > Have a look at "rdate": > > I'm glad it works for you, but no. > > RFC 868 Time Protocol was obsoleted when NTP became ubiquotous in the > 1980s. IMO, there's no point in keeping rdate(8) around when there > are multiple good NTP implementations available. Yes and no: -n Use SNTP (old RFC 2030, currently RFC 5905) instead of the RFC 868 time protocol. # rdate -n ntp.xs4all.nl Fri Jun 12 12:07:04 CEST 2020 :-) -- richard lucassen https://contact.xaq.nl/ ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] Orange Pi Zero/Orange Pi PC
Quoting richard lucassen (mailingli...@lucassen.org): > Have a look at "rdate": I'm glad it works for you, but no. RFC 868 Time Protocol was obsoleted when NTP became ubiquotous in the 1980s. IMO, there's no point in keeping rdate(8) around when there are multiple good NTP implementations available. ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] Orange Pi Zero/Orange Pi PC
On Thu, 11 Jun 2020 15:09:37 -0700 Rick Moen via Dng wrote: > If running ntpdate, depending on how much skew there is from correct > time, you might need the '-B' option. Quoting the man page: Have a look at "rdate": $ apropos rdate rdate (8)- set the system's date from a remote host I use it on boards without RTC, just after the network comes up and before openntpd takes over. No need for "fake-hwclock". -- richard lucassen https://contact.xaq.nl/ ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] Orange Pi Zero/Orange Pi PC
Quoting tux...@sapo.pt (tux...@sapo.pt): > problems with rtc clock are not uncommon, the need for a power > cycle, seems to be.. > > This discrepancies, in the date, .. how many days it takes to > happen, does you have any Idea? > does the Boards are properly powerup?I mean suficient levels of > Voltage/Current.. > > Try to update the date with something like 'ntpdate ntp_server_ip' > And see is that helps, if yes then you can run it via a cronjob, > from time to time.. If running ntpdate, depending on how much skew there is from correct time, you might need the '-B' option. Quoting the man page: -B Force the time to always be slewed using the adjtime() system call, even if the measured offset is greater than +-128 ms. The default is to step the time using settimeofday() if the offset is greater than +-128 ms. Note that, if the offset is much greater than +-128 ms in this case, that it can take a long time (hours) to slew the clock to the correct value. During this time, the host should not be used to synchronize clients. I should also mention that ntpdate being increasingly unmaintained and crufty means that the 'sntp' or 'ntpd' utilities with appropriate options are now better long-term tools. https://support.ntp.org/bin/view/Dev/DeprecatingNtpdate I think 'ntpd -gq' will do the trick. Or 'sntp -s '. 'ntpd -gq' will fail if the ntp daemon is binding to the service port, but ISTR you can supply '-u' to work around that. (Above is from memory, not tested.) ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] Orange Pi Zero/Orange Pi PC
On Thu, 11 Jun 2020 03:50:54 +0100 tux...@sapo.pt wrote: > > As I need to apply a power cycle I suppose it is not a software > > issue (is that right?). I'm just trying the Ascii image to see what > > will happen, but I'm not very hopefull. > > > > Has anyone seen such issues with these boards? > > problems with rtc clock are not uncommon, the need for a power > cycle, seems to be.. > > This discrepancies, in the date, .. how many days it takes to > happen, does you have any Idea? > does the Boards are properly powerup?I mean suficient levels of > Voltage/Current.. > > Try to update the date with something like 'ntpdate ntp_server_ip' > And see is that helps, if yes then you can run it via a cronjob, > from time to time.. No, even cronjobs do not work or give segmentation fault. It must be the hardware. There must be a batch with these issues, I'm not the only one BTW. I wonder if it has got something to do with the wifi chip. A crash occurs earlier if you load the kernel module. > For beowulf only orangepi_zero_plus2 image is available.. > > Try to install the Devuan ascii image: > https://mirror.leaseweb.com/devuan/devuan_ascii/embedded/README.txt The ASCII image is ok, but I updated this image to Beowulf and I used Armbian kernel with it. But I fear it will crash again. Anyway, I just got some Rasberry Pi Zero-W now and I think I will stay with these boards. Remember I have also Orange Pi PC, different CPU, suffering from the same issues. R. -- richard lucassen https://contact.xaq.nl/ ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] Orange Pi Zero/Orange Pi PC
Hello richard, Citando richard lucassen via Dng : As I need to apply a power cycle I suppose it is not a software issue (is that right?). I'm just trying the Ascii image to see what will happen, but I'm not very hopefull. Has anyone seen such issues with these boards? problems with rtc clock are not uncommon, the need for a power cycle, seems to be.. This discrepancies, in the date, .. how many days it takes to happen, does you have any Idea? does the Boards are properly powerup?I mean suficient levels of Voltage/Current.. Try to update the date with something like 'ntpdate ntp_server_ip' And see is that helps, if yes then you can run it via a cronjob, from time to time.. For beowulf only orangepi_zero_plus2 image is available.. Try to install the Devuan ascii image: https://mirror.leaseweb.com/devuan/devuan_ascii/embedded/README.txt Best Regards, tux ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
Re: [DNG] Orange Pi Zero/Orange Pi PC
On Mon, 1 Jun 2020 20:56:30 +0200 richard lucassen via Dng wrote: Oh, BTW, kudos for the one who made these images. Just the strict necessary. No bloat or junk. -- richard lucassen http://contact.xaq.nl/ ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
[DNG] Orange Pi Zero/Orange Pi PC
Hello list, I just installed the Ascii image for the Orange Pi Zero (H2 CPU). I have a bunch (20) of these OPi Zero's, running a SysV version of Debian. Anyway, in practice *all* of these units are unstable (some services stop working, clock at a weird time like 1937 or 2134). Even a reboot does not help. To get things back in order a power cycle is necessary. The Orange Pi PC also suffers from this phenomenon (H3 CPU). Switching on the wireless interface speeds up the appearance of the problem. When it's idle there's no problem. I also tried to set the max CPU freq to 800 MHz, but no way. These devices have different tasks and thus different user space software. As I need to apply a power cycle I suppose it is not a software issue (is that right?). I'm just trying the Ascii image to see what will happen, but I'm not very hopefull. Has anyone seen such issues with these boards? R. -- richard lucassen http://contact.xaq.nl/ ___ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng