On Tuesday, April 03, 2012 12:38:42 PM Erik Christiansen did opine: > On 03.04.12 09:23, gene heskett wrote: > > They run very well indeed, until you quit an app. At which point > > there seems to be a 5 to 10% possibility of the whole box going away > > about 1 second after the apps screen goes away, leaving a black > > screen & frozen mouse cursor. Hit the reset or power buttons to > > recover. OTOH, if its X crashing, maybe I could try typing startx? > > I'll see if I can induce another such incident and try that just for > > S&G. > > Is the host networked, Gene?
Yes. > Linux should not be seriously > inconvenienced by an app crashing, and I usually ssh in, then kill gdm > or similar. Then it's only necessary to restart X. Looking at an output > of ps on this box, I see: > > root 745 1 0 15:03 ? 00:00:00 gdm-binary > root 841 745 0 15:03 ? 00:00:00 > /usr/lib/gdm/gdm-simple-slave --display-id /org/gnom > e/DisplayManager/Display1 root 850 841 3 15:03 tty7 > 00:18:47 /usr/bin/X :0 -nr -verbose -auth /var/run/gdm/auth-f > or-gdm-FK2czT/database -nolisten tcp vt7 > > If it's late at night, I'm not always completely scientific, and just > whack these things till the jammed X goes away, then do a startx. > (But gdm-binary has init (PID == 1) as parent, and is momma to the > other guys, so that's the bullseye.) > TBT I haven't tried to ssh into it when its in that state. I'll try it next time. The normal ssh session is ssh -Y, which I finally got fixed as I posted a day or so back. > As for the cause - there are a lot of us using ubuntu and a recent > edition of the X11 suite, with nary a problem, so a driver might be the > culprit - it may not be too happy with your hardware? Is this with > on-board graphics, and do you have another you could plug in? Hardware on these D525MW boards is fixed, video is i915 intel, built into the motherboard. No slot to plug in another video card. > I can understand that the 5-10% rate of falling over would get on your > wick! That is a nice way of saying it. ;) > Erik > > P.S. I had mains power go once, in the middle of typing a post in vim > within mutt. So I lost that one because mutt's tmp file was in > /tmp, and was deleted on reboot. Changing the tmpdir fixed that. That is something that I wish I could disable. Heyu is hard coded, and goes away when cron cleans up in /tmp. So it turns on the Christmas lights, cron cleans up /var/tmp, heyu goes away till I notice the lights are on a day or 4 later. This box has a lot of duties, email is just one of many. > I think you might have been better off with vim. ;-) Probably, but this morning I am doing it with a stick & a pack of blue- line. I'll make it purty later, if I live till later. :( I was dumb I guess, I let the bios put in the "optimized defaults", and now the z motor is running like the latencies are in milliseconds the instant it starts to dither the pulses, and it will freeze up any time I set it to go faster than about 12 IPM. I can grab the motor and feel it running rougher than a corn sheller. Humm, go check, I'll bet that turned on hyperthreading again. Yup, sure did. So much for that silk purse from wallmart feature. Turning that off ought to be a jumper on the motherboard so it could be permanent! Cheers, Gene -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) My web page: <http://coyoteden.dyndns-free.com:85/gene> Houston, Tranquillity Base here. The Eagle has landed. -- Neil Armstrong ------------------------------------------------------------------------------ Better than sec? Nothing is better than sec when it comes to monitoring Big Data applications. Try Boundary one-second resolution app monitoring today. Free. http://p.sf.net/sfu/Boundary-dev2dev _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
