Re: Wake via serial port?
On Sat, 13 Jul 2013, Craig R. Skinner wrote: Lady Di, Ahh, that is so cute. glad you clarified your original short reply. g.day diana Past hissy-fits are not a predictor of future hissy-fits. Nick Holland(06 Dec 2005)
Re: Wake via serial port?
Appreciate the clarification. Now someone searching past posts will have a better idea of what happened. g.day diana Past hissy-fits are not a predictor of future hissy-fits. Nick Holland(06 Dec 2005)
Re: Wake via serial port?
On 2013-07-12 Fri 17:39 PM |, Diana Eichert wrote: > > What you are asking only makes sense, unfortunately > Craig appears to be like a lot of malling list > subscribers. They are "takers" not "givers". > Lady Di, It's gracious to be respectful of other's timezone's & life schedules. Cheers, -- Craig Skinner | http://twitter.com/Craig_Skinner | http://linkd.in/yGqkv7
Re: Wake via serial port?
On 2013-07-12 Fri 23:12 PM |, Thomas Reiter wrote: > > would you mind to share how you have solved the problem? > otherwise someone has to ask the same question some day. > Of course Thomas, but as each piece of hardware is different, I doubt what worked in this case will be transferable. I first began by trying Wake on LAN with pointers from https://calomel.org/wakeonlan.html Although the NIC's WOL cable was connected & the BIOS claimed WOL support, it wouldn't bring the box up. Eventually I discovered ifconfig(8)'s notes re WOL & established that the cheap card didn't actually support WOL: $ ifconfig rl0 hwfeatures | fgrep hwfeatures hwfeatures=10 hardmtu 1500 $ sudo ifconfig rl0 wol ifconfig: SIOCSIFXFLAGS: Not supported So I started experimenting via the serial port & then the power button. With Nick's suggestions about the power button, I thought that'd be the quickest route, so opened the box again, removed the WOL cable & noticed some motherboard settings. That jogged my mind to Nick's comments about just because the BIOS claims to do something, doesn't mean it will. With random inspiration I decided to have another look at the BIOS, & now there was an option to have the box auto power on. So whether I nudged a DIP switch or otherwise jiggled something, I can't be sure. Anyway, Nick was right, as "I'll tell you how to figure it out." Job done, in unspectacular style. Onwards, -- Craig Skinner | http://twitter.com/Craig_Skinner | http://linkd.in/yGqkv7
Re: Wake via serial port?
Sure, Nick did explain his solution, which was the same one I would have done since I've been doing this since 300 baud was the norm. But is this what Craig did? Who knows? diana On Fri, 12 Jul 2013, patrick keshishian wrote: On Fri, Jul 12, 2013 at 4:39 PM, Diana Eichert wrote: Thomas What you are asking only makes sense, unfortunately Craig appears to be like a lot of malling list subscribers. They are "takers" not "givers". Nick already explained and outlined all the necessary steps. Did he not? --patrick
Re: Wake via serial port?
On 07/12/13 20:05, patrick keshishian wrote: > On Fri, Jul 12, 2013 at 4:39 PM, Diana Eichert wrote: >> Thomas >> >> What you are asking only makes sense, unfortunately >> Craig appears to be like a lot of malling list >> subscribers. They are "takers" not "givers". > > Nick already explained and outlined all the necessary steps. Did he not? > > --patrick > yeah, but people usually nod off after the second or third sentence. :) Nick. (Curing Insomnia since 1983)
Re: Wake via serial port?
On Fri, Jul 12, 2013 at 4:39 PM, Diana Eichert wrote: > Thomas > > What you are asking only makes sense, unfortunately > Craig appears to be like a lot of malling list > subscribers. They are "takers" not "givers". Nick already explained and outlined all the necessary steps. Did he not? --patrick > g.day > > diana > > On Fri, 12 Jul 2013, Thomas Reiter wrote: > >> On 07/12/2013 10:38 PM, Craig R. Skinner wrote: >>> >>> On 2013-07-12 Fri 10:42 AM |, Nick Holland wrote: but I'll tell you how to figure it out. [ wise words of practical relevance ] >>> >>> Solved! >>> >>> Thanks, >>> >> >> would you mind to share how you have solved the problem? >> otherwise someone has to ask the same question some day. >> >> >> best, >> thomas
Re: Wake via serial port?
Thomas What you are asking only makes sense, unfortunately Craig appears to be like a lot of malling list subscribers. They are "takers" not "givers". g.day diana On Fri, 12 Jul 2013, Thomas Reiter wrote: On 07/12/2013 10:38 PM, Craig R. Skinner wrote: On 2013-07-12 Fri 10:42 AM |, Nick Holland wrote: but I'll tell you how to figure it out. [ wise words of practical relevance ] Solved! Thanks, would you mind to share how you have solved the problem? otherwise someone has to ask the same question some day. best, thomas
Re: Wake via serial port?
On 07/12/2013 10:38 PM, Craig R. Skinner wrote: > On 2013-07-12 Fri 10:42 AM |, Nick Holland wrote: >> >> but I'll tell you how to figure it out. >> >> [ wise words of practical relevance ] >> > > Solved! > > Thanks, > would you mind to share how you have solved the problem? otherwise someone has to ask the same question some day. best, thomas
Re: Wake via serial port?
On 2013-07-12 Fri 10:42 AM |, Nick Holland wrote: > > but I'll tell you how to figure it out. > > [ wise words of practical relevance ] > Solved! Thanks, -- Craig Skinner | http://twitter.com/Craig_Skinner | http://linkd.in/yGqkv7
Re: Wake via serial port?
On 07/12/2013 09:45 AM, Craig R. Skinner wrote: I've a box that won't self start after a power failure. The BIOS docs shows: Remote Ring On This allows you to wake up the system from a serial port modem. How could this be done from another OpenBSD box connected via a serial cross over cable + cu/tip/etc? The serial link is operational & I get the console on the non-self-starting box after I manually press the power button. Thoughts? not going to give you a "do this and all will work", but I'll tell you how to figure it out. 1) verify that this "feature" really works on this machine... a) Get a nine volt battery and a battery clip ending in two wires. b) Connect the battery between the Ring Indicator (RI) pin and the ground pin. c) If it doesn't turn on, swap the red and black wires, and try again. d) If it still doesn't work, it's a left over feature in the bios, your hardware doesn't actually support this. 2) Find a line you can control on the terminal machine. man 4 tty, man 4 termios appear to be useful. a) Get/build an RS232 monitoring plug, and figure out what RS232 handshake line you can control (std two-pin, red/green LEDs and 1K resistors do just fine here) b) the pin you can control should default to the right polarity for what you wish to accomplish. 3) Make it work Build a custom cable which connects the line you can control to the ring detect line. Note that standard null modem cables don't generally pass the RI pin, so you will be building one. Maybe easier: just strap the RI pin to a level that causes the machine to light up on its own. An old cell phone charger or other wall wart may be usable to do this. A stupidly simple trick to make a box auto-start after a power failure, and I think I can credit Henning@ with suggesting it to me, is to put a capacitor across the power button lines. On power-up, the capacitor is discharged, so passes current, acting like someone was pushing the power button. It quickly charges up, and now it acts as if someone released the button. IIRC, 100uF worked pretty well on one machine I did this with, your results will vary. Make sure you get the cap polarity right, or it won't last very long!! I found it good to put a bleeder resistor across the cap/switch combo, too, as otherwise the power had to be off too long to auto-start when it came back up (the capacitor was still charged!), you will have to experiment with this. The bleeder resistor should be as low in resistance as doesn't cause the machine to think the button is pushed, maybe try 1k, 10k, 100k, 1M values. Nick.
