netfs service kills all processes in the system after terminating
Issue
- In certain situation, running
# service netfs stop
- This kills everything, including shell scripts that are supposed to bring the system down, when service netfs stop is called after switching to runlevel 6.
- On system with
psmisc
older thanpsmisc-22.6-19.el6_5
, one has buggy/sbin/fuser
- it expects PIDs to be only 5-digits long, longer are concatenated, giving something like:
$ fuser -m /home/test/ 2>/dev/null; echo
3188109050109166109167109168109169109170109171109172109173109174109175
$
/bin/kill
doesn't check errno after callingstrtol()
when converting argument to integer. This kills everything.
# service netfs stop
Unmounting NFS filesystems: umount.nfs: /nfs: device is busy
umount.nfs: /nfs: device is busy
umount.nfs: /home/test: device is busy
umount.nfs: /home/test: device is busy
[FAILED]
Read from remote host localhost: Connection reset by peer
Environment
- Red Hat Enterprise Linux 6.4
- util-linux-ng-2.17.2-12.9.el6_4.3.x86_64
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.