candlejack is correct. Do not delete the file, because processes are accessing it. instead reduce it to zero length by issuing
Code:
cat /dev/null > /var/adm/wtmp
The difference is that this way the file will keep its inode and hence alle the processes accessing it will still find the file where they look for it.
If you delete the file and it is created anew it would most probably get a different inode, which is a potential hazard.
Hi, guys, I have a big problem.
I've got a sun solaris 4.1.4 workstation, and the /var/adm/message file will add one row every few seconds. It becomes a large file in a short time.
I wander if there are some mistakes configuring the workstation.
the /var/adm/message is as follow:
... (3 Replies)
hi sirs
can u tell the difference between /var/log/syslogs and /var/adm/messages
in my working place i am having two servers.
in one servers messages file is empty and syslog file is going on increasing..
and in another servers message file is going on increasing but syslog file is... (2 Replies)
I have a solaris 10 machine, lately I had a drive error issue so the dmesg command would show a screen full, now that the drive issue is resolved I want to have a fresh start for my #dmesg command output.
The root cron looks like this
The /etc/logadm.conf file has this entry
If I manually... (2 Replies)
Do not know the reason y messages file is empty already restarted the syslog daemon but still its showing empty .
xxxxxxx# more /var/adm/messages
xxxxxx#
# ps -efo zone,pid,ppid,time,comm | grep syslog | grep global
global 11861 1 00:10 /usr/sbin/syslogd
svcs... (2 Replies)
Hi, The server is p570 with AIX@5300-10...
The messages file is missing...how do I get it back and start logging the messages??
Can I just create the file again? (2 Replies)
Hi all.
I have a strange case on one of my AIX boxes. /var/adm/wtmp on server01 is ~ 400MB large but it only has ~1200 lines. For example on server02 there are ~85000 lines and the file is ~158MB large. I check lines through 'last | wc -l'. But when I check line directly with 'wc -l... (2 Replies)
dear all
this attached photo is send to me from Arcsight admin can you please advice
ftpd failed to write /var/adm/wtmp not owner
ftpd failed to write /var/adm/wtmp error 0 (1 Reply)
Hi all,
I have F5 load balancer on my system and checking service status by opening an ftp session in every 30 seconds. These ftp sessions are being logged in /var/adm/wtmpx and filling up the file. when i run the last command most of the output is this ftp session. I was wondering if there is a... (1 Reply)
Discussion started by: cepxat
1 Replies
LEARN ABOUT DEBIAN
utmpset
utmpset(8) System Manager's Manual utmpset(8)NAME
utmpset - logout a line from utmp and wtmp file
SYNOPSIS
utmpset [ -w ] line
DESCRIPTION
The utmpset program modifies the user accounting database utmp(5) and optionally wtmp(5) to indicate that the user on the terminal line has
logged out.
Ordinary init(8) processes handle utmp file records for local login accounting. The runit(8) program doesn't include code to update the
utmp file, the getty(8) processes are handled the same as all other services.
To enable local login accounting, add utmpset to the getty(8) finish scripts, e.g.:
$ cat /service/getty-5/finish
#!/bin/sh
exec utmpset -w tty5
$
OPTIONS -w wtmp. Additionally to the utmp file, write an empty record for line to the wtmp file.
EXIT CODES
utmpset returns 111 on error, 1 on wrong usage, 0 in all other cases.
SEE ALSO sv(8), runsv(8), runit(8), runit-init(8)runsvdir(8), runsvchdir(8), chpst(8), svlogd(8), getty(8)
http://smarden.org/runit/
AUTHOR
Gerrit Pape <pape@smarden.org>
utmpset(8)