Date: Sun, 27 Apr 2003 21:09:32 -0400 (EDT) From: "J. Seth Henry" <jshamlet@comcast.net> To: questions@freebsd.org Subject: Re: Odd behavior on 4.8-REL since installing KDE 3.1 Message-ID: <20030427210629.D27080@alexandria.gambrl01.md.comcast.net> In-Reply-To: <20030427200729.P26560@alexandria.gambrl01.md.comcast.net> References: <20030427200729.P26560@alexandria.gambrl01.md.comcast.net>
next in thread | previous in thread | raw e-mail | index | archive | help
Ok, I managed to find the source of the first problem. Too many *'s in the crontab. A shame cron didn't detect the problem, but oh well. I noticed the * in the email output, and that got me to thinking. A shame I didn't do more thinking before posting... ;) The other problem, though, is still baffling. I can't figure out what uptime is doing. I can understand it may be trying to read info on the terminal, but shouldn't it be looking at the virtual tty? Then, there is the strange fact that it doesn't clear after a reboot. Sorry for the confusion, Seth Henry On Sun, 27 Apr 2003, J. Seth Henry wrote: > Hello, > I recently updated my X app server to 4.8-REL, largely to gain easier > access to KDE 3.1, which is vastly improved (so far) over previous > released of KDE. However, since installing it, I have noticed strange > behavior systemwide. I'm not sure if this is really a FreeBSD question, > but I'm hoping someone else has seen this, or can diagnose where it is > coming from. > > The first odd thing is that scripts run by cron now return an error (by > email) saying: > > Date: Sun, 27 Apr 2003 20:05:00 -0400 (EDT) > From: Cron Daemon <root@alexandria.gambrl01.md.comcast.net> > To: root@alexandria.gambrl01.md.comcast.net > Subject: Cron <root@alexandria> * /usr/local/bin/sndate > > Desktop: not found > > This started a few hours ago, shortly after I logged off as root. This > appears to be causing the scripts to fail, so I would call this a bit > severe. > > The other is just mildly annoying. I have four dedicated X terminals, and > a Windows PC. They all use DHCP to get network info. Having blown all my > naming originality on the dedicated hosts, I just call them dhcp0 -> dhcp9 > (I added a few extras for guests) > > Now, when I run uptime, I get this: > > alexandria# uptime > uptime: /dev/dhcp0:0: No such file or directory > uptime: /dev/dhcp4:0: No such file or directory > 8:13PM up 20:41, 0 users, load averages: 0.08, 0.04, 0.01 > > I first noticed this after I ran it while logged into (you guessed it) > dhcp0. Later, I logged into my bedroom terminal, and noted that it added > /dev/dhcp4. It appears to be permanent. I have even rebooted the app > server, and it still appears. What is odd is that it even appears on > console, and via ssh. It also appears for all users, root included. > > Again, this appears to be kde related. This, too, started after I > installed the load. Unfortunately, I upgraded both the OS and the desktop > manager at the same time - but it looks suspicious. > > Other notes: > > The upgrade was a clean new install. I backed up important configuration > files, and restored where necessary. Only files such as fstab, /etc/ssh/*, > smb.conf, etc were restored. (although any user-specific KDE specific > stuff from 3.0 would have been retained with the user accounts) > > I use kdm as the login manager, with only a few modifications to the base > scripts (primarily to allow network access, and disallow console access), > but is running clean otherwise. > > As an attempted diagnostic tool, I modified one of the shell scripts run > by cron to echo $PATH, but I get no result. (which appears to confirm that > the scripts aren't being run. > > i.e.: > alexandria# cat /usr/local/bin/sndate > #!/bin/sh > echo $PATH > /usr/local/bin/statnet `date "+DATE=%m%d%y"` > /dev/null > sleep 1 > /usr/local/bin/statnet `date "+TIME=%H%M"` > /dev/null > sleep 1 > /usr/local/bin/statnet `echo "SN1 PMES1=Outdoor T: > "``/usr/local/bin/statnet r1s2 | cut -c 10-` > /dev/null > > This script updates the clock in my thermostat, and displays the > otherwise hidden outdoor temperature to the LCD. (it has five sensors, > only two of which are displayed on the primary LCD. > > I can't think of what else might be helpful. I use tcsh for all accounts, > including root. I have modified the scripts to clear the screen after > login, and print the uptime. They also clear the screen before logout, but > are otherwise unmodified. I can't see how this affects anything, as the > scripts call sh to run, but you never know. > > Any help would be appreciated, though I am probably going to follow up on > the KDE questions list as well. It's a shame, because KDE 3.1 is, > otherwise, a vast improvement over previous releases. I've been using it > fairly heavily, and no bombs yet! > > Thanks, > Seth Henry >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030427210629.D27080>