Hi, I encounterd an issue involving automount start. As part of our tests we reboot our machines and test their state after they have recovered. Sometimes these reboots include the --force flag and even --no-sync. At other times we are forces to use more drastic commands like "echo 1 > /proc/sys/kernel/sysrq; echo b > /proc/sysrq-trigger". Once in a while, after a reboot, we detect autofs is not running. After some investigating we discovered that the pid file in /run/autofs-running hasn't been cleaned before reboot and autofs reports it is already running: [root@autofs-6 ~]# pidof automount [root@autofs-6 ~]# automount --debug automount: program is already running. [root@autofs-6 ~]# cat /run/autofs-running 3394 [root@autofs-6 ~]# rm /run/autofs-running rm: remove regular file ‘/run/autofs-running’? y [root@autofs-6 ~]# automount --debug [root@autofs-6 ~]# pidof automount 6362 [root@autofs-6 ~]# cat /run/autofs-running 6362 This seems very odd to me. Why doesn't automount check if the process is really running (like with pidof or any other way) and how do we solve this issue? some guesses to why this is happening include: an issue with the tmpfs /run directory or a mount race. We are trying to reproduce and collect some data. We are using autofs 5.1.2. Thanks, Doron -- To unsubscribe from this list: send the line "unsubscribe autofs" in
