History source of Computer Problem-Solution (Open)(No. 17)

This is a page of "How to" solve some computer problems that may happen.

P = problem;
S = solution

#contents

** P: (2010.04.29) set-pathname-coding-system function does not exists in new xemacs lisp which existed in flex. [#t79f4d54]
** P: (2010.01.19) Managing ntp [#p100119]
*** S: ntp and ntpdate package have to be configured in some tricky ways [#s100119]
- ntp.conf file
 [root@tube40 ~]# emacs /etc/ntp.conf
The /etc/ntp.conf file is the main configuration file for Linux NTP in which we place the IP addresses of the servers we want to use. Here we only use one server:
 server  ntp.tohoku.ac.jp    # A stratum 1 server at tohoku.ac.jp
 #server  ntp.redhat.com     # A stratum 2 server at redhat.com
We have to restart the NTP process every time we make a change to the configuration file for the changes to take effect on the running process.

- To get NTP configured to start at boot, use the line: 
 [root@tube40 ~]# chkconfig ntpd on

- To start, stop and restart NTP after booting, follow these examples:
 [root@tube40 ~]# /etc/init.d/ntpd start
 [root@tube40 ~]# /etc/init.d/ntpd stop
 [root@tube40 ~]# /etc/init.d/ntpd restart

- For the synchronization, ntpd should be stopped and ntpdate should be run at least three times, then ntpd can be run again:
 [root@tube40 ~]# /etc/init.d/ntpd stop
 Shutting down ntpd:                                        [  OK  ]
 [root@tube40 ~]# /usr/sbin/ntpdate -u ntp.tohoku.ac.jp
 19 Jan 16:11:08 ntpdate[23079]: adjust time server 130.34.11.117 offset 0.016601 sec
 [root@tube40 ~]# /usr/sbin/ntpdate -u ntp.tohoku.ac.jp
 19 Jan 16:11:09 ntpdate[23080]: adjust time server 130.34.11.117 offset 0.015737 sec
 [root@tube40 ~]# /usr/sbin/ntpdate -u ntp.tohoku.ac.jp
 19 Jan 16:11:13 ntpdate[23081]: adjust time server 130.34.11.117 offset 0.014081 sec
 [root@tube40 ~]# /etc/init.d/ntpd start
 Starting ntpd:                                             [  OK  ]

** P: (2010.01.13) The problem of 2010.1.6 is solved but response is slow. [#cc7513d1]

*** S: The problem is due to the fact that /etc/mtab is on root (not snapshot) [#j127f4e3]

Since we do not know what is the process of mount the file.
we can not change mtab to snapshot. 
For a moment, we force mount /home1 and /rsaito on /tmp/home1, /tmp/rsaito

 /etc/rc5.d/S28autofs stop
 mkdir /tmp/home1
 mkdir /tmp/rsaito
 mount -n 192.168.2.100:/home1 /tmp/home1
 mount -n 192.168.2.100:/rsaito /tmp/rsaito
 df

/rsaito/bin/pxe-mount

 

** P: (2010.1.6) /rsaito is not automountd on PXE client. [#m92944ad]

*** S: edit and add flexS:/rsaito information on [#q5d5939e]

 tube30:/nfsboot/root/etc/mtab

** P: (2009.10.10) Gnorm logout is slow for Xming etc. [#y54d795e]

This is a firewall problem of windows.

*** S: Open port 16001 tcp [#c4bf3cc3]

control panel -> firewall -> add new port -> 

 Name Esound
 Port 16001
      TCP

** P: (2009.07.17) Load average of FLEX is very high. [#ba0d0fa4]
This is due to mounting problem again (mount NFS). After checking mount status, it is found that a network harddisk was not unmounted nicely before turning off.
*** S: Re-mount (turn on) the harddisk, and do some safety procedure. [#l1c8b137]
- check mount status:
 # mount
- umount once (assuming mount point is /mnt/backup):
 # umount -f /mnt/backup
- synchronize several times:
 # sync
 # sync
 # sync
- check df or mount status again
- if the harddisk has been nicely unmounted, then it is safe to turn it off.

** P: (2009.07.02) Flex is unstable for logining or e-mails. [#ac90f02f]
*** S: mount to tube:/home1-old was not allowed any more. [#e2a3634b]

 comment out one line in /etc/auto.misc 
 # home1    172.17.4.134:/home1-old

 ./S28autofs forcerestart

 I also delete in /etc/exports 

 the requests from 172.17.4.184 (nuguraha) 172.17.4.188 (mahabubu)
 and 172.17.4.127 (furu).

 exportfs -a

 we also need to restart nfs
 /etc/rc5.d/S60nfs restart

** P: (2009.06.03) Today we were failed to run ether-wake command on new-PXE server/client machines using new Hub/router. [#m53007b8]
*** S: We manually turn those on and off. After that ether-wake (wake-on-Lan) command was successful. [#a08cf22e]

** P: PXE (network boot) is now installing 2009.5.1 [#e8f71204]
*** S: finished. status [[PXE Server (Open)]] [#p74ace39]

** P: We move /etc/passwd then we can not be root. No more even shutdown. [#oaefbc51]
*** S: cntl-alt-del and reboot. Then when OS selection, push any arrow. [#v36584e8]
Then many selection of OS appears. Select one of OS, and press E.
Then select kernel and press E again. Then you can edit the command for starting  kernel and add "single" in the last of the command. Finally select "b" for booting and single user mode appears. Make /etc/passwd files and exit and runlevel 5 is recovers. 

Even if we used NIS, we should not delete /etc/passwd /etc/group and /etc/shadow.

** P: linux emacs ispell error (no language "eu") 2009.4.24 [#xf61128d]
*** S: solved. "M-x ispell-change-dictionary" and select english [#t33f8f93]
*** S: solved. Edit .emacs to add (setq ispell-dictionary "english") [#l1a50648]

** P: How to update the files of remote to the files of local [#lb59159d]
*** S: solved. "rsync -avc -e ssh saito@remote:/home/saito/book09/ /rsaito/book09"[#f25e1654]
History source of Computer Problem-Solution (Open) (No. 17)
