ramus@WINDSAIL.NERSC.GOV (Joe Ramus) (03/12/91)
>> What version of SunOS are you running? There is little problem in running >> xntpd under 4.1, the question is whether it will run under 4.0.x. I am having a problem with a Sparc 1+ running SunOS 4.1. I built all of the software and then tried the ntpq command like this: {ROOT:2 } ntpq -i ntpq: ntp/udp: unknown service I must have udp running because I have just the standard OS and I do a lot of network stuff such as NFS, telnet, ftp, rsh, Yellow Pages, etc. Joe Ramus Lawrence Livermore Lab ramus@nersc.gov
wales@CS.UCLA.EDU (Rich Wales) (03/12/91)
Joe -- You said you got the following when you tried running "ntpq": {ROOT:2 } ntpq -i ntpq: ntp/udp: unknown service This means you don't have (but need to have) a line such as the follow- ing in your "/etc/services" file: ntp 123/udp # Network Time Protocol If you are running YP/NIS, add this line to the /etc/services file on your YP/NIS master site, then remake the "services" map. Rich Wales <wales@CS.UCLA.EDU> // UCLA Computer Science Department 3531 Boelter Hall // Los Angeles, CA 90024-1596 // +1 (213) 825-5683
ramus@WINDSAIL.NERSC.GOV (Joe Ramus) (03/12/91)
>> This means you don't have (but need to have) a line such as the follow- >> ing in your "/etc/services" file: >> >> ntp 123/udp # Network Time Protocol >> >> If you are running YP/NIS, add this line to the /etc/services file on >> your YP/NIS master site, then remake the "services" map. Thanks to several people who came up with this idea. My current entry has TCP in it so I will change it to UDP. Here is the current entry from /etc/services: ntp 123/tcp # Network Time Protocol I assume that the standard software distribution from SUN for SunOS 4.1 must have the tcp entry. NOTE: How about listing this possible problem in the README file for the NTP distribution so other people will not have to solve the same problem again. Joe Ramus Lawrence Livermore Lab ramus@nersc.gov
ramus@WINDSAIL.NERSC.GOV (Joe Ramus) (03/14/91)
HELP. Is there a proper way to stop the xntpd server and then start it again later? I have SunOS 4.1 on a Sparc. I have the latest XNTP stuff from the LOUIE machine. I built it all and set up a configuration file and had the server running for more than 24 hours. It was great. Then I killed it by: kill -9 pid Now it will not run again. It seems to start and I can do a PS and see it but then it goes away real soon. Here is a sample line: root 871 0.0 4.9 80 332 ? S < 18:49 0:00 xntpd -c /etc/ntp.conf I tried rebooting my system but that did not help. I run the TICKADJ program. I tried the NTPQ program and it works. I also tried NTPDATE and it gives a message: Alarm clock Thanks in advance for any help you can give me. Joe Ramus Lawrence Livermore Lab ramus@nersc.gov
katz@MERIT.EDU (Dave Katz) (03/14/91)
Sounds like somebody probably recompiled the program while you weren't looking. I get the same "alarm clock" message from ntpdate when I comple on an RS/6000 without the bsd compatibility library. Perhaps some bad library routines got linked in? Check the last change date on xntpd. From ntp-relay@trantor.umd.edu Wed Mar 13 22:16:32 1991 Received: Wed, 13 Mar 91 22:16:28 EST from trantor.umd.edu by merit.edu (5.59/1.6) Received: by trantor.umd.edu (5.65/1.34) id AA14927; Wed, 13 Mar 91 22:04:06 -0500 Received: from merlin.Nersc.gov by trantor.umd.edu (5.65/1.34) id AA14922; Wed, 13 Mar 91 22:04:01 -0500 Received: from windsail.nersc.gov by merlin.nersc.gov (4.1/NERSC-1.18a) id AA05941; Wed, 13 Mar 91 19:04:12 PST Received: by windsail.nersc.gov (4.1/SMI-4.1) id AA00896; Wed, 13 Mar 91 19:04:06 PST Date: Wed, 13 Mar 91 19:04:06 PST From: ramus@windsail.nersc.gov (Joe Ramus) Message-Id: <9103140304.AA00896@windsail.nersc.gov> To: ntp@trantor.umd.edu Subject: NTP & Sparcstation Status: RO HELP. Is there a proper way to stop the xntpd server and then start it again later? I have SunOS 4.1 on a Sparc. I have the latest XNTP stuff from the LOUIE machine. I built it all and set up a configuration file and had the server running for more than 24 hours. It was great. Then I killed it by: kill -9 pid Now it will not run again. It seems to start and I can do a PS and see it but then it goes away real soon. Here is a sample line: root 871 0.0 4.9 80 332 ? S < 18:49 0:00 xntpd -c /etc/ntp.conf I tried rebooting my system but that did not help. I run the TICKADJ program. I tried the NTPQ program and it works. I also tried NTPDATE and it gives a message: Alarm clock Thanks in advance for any help you can give me. Joe Ramus Lawrence Livermore Lab ramus@nersc.gov
ramus@WINDSAIL.NERSC.GOV (Joe Ramus) (03/15/91)
Thanks to those who helped solve my problem. I had xntpd running once but then it would not run after a reboot. The problem is that the Sparc has both a BSD & a System V development environment. Since I work with System V, I had my path set that way yesterday when I recompiled the xntpd program. The earlier version that worked must have been compiled with my path set for BSD. So I changed the line in the Config file to look like this: COMPILER= /usr/ucb/cc NOTE: How about mentioning in the README file that SUN machines require the use of the BSD environment. --------------------------------------------------------------------------- >> I get the same "alarm clock" message from ntpdate when I comple >> on an RS/6000 without the bsd compatibility library. Perhaps >> some bad library routines got linked in? Check the last change >> date on xntpd. >> >> >> You've compiled your xntp stuff with /usr/5bin/cc. Use /usr/bin/cc instead. >> signal(2) works in different ways on System V and BSD systems. >> System V resets the signal handler (in most cases), BSD does not. >> Joe Ramus Lawrence Livermore Lab ramus@nersc.gov