On Thu, Apr 10, 2008 at 10:13:42AM +0200, Ola Lundqvist wrote: > I think this problem is actually with drbd. The reason is that normally > all filesystem parts are actually in /etc/rcS.d/. Shouldn't it be > even before S35mountall.sh? > > Or do you have an other opinion?
Well, the init.d script which comes with (upstream?) drbd package contains: # Required-Start: $network $syslog sshd I guess depending on syslog makes sense, as otherwise you might lose information on what was going on during drbd initialization, if the kernel log buffer overflows. On the other hand we're not doing that for other similar things like NFS mounts.. I think the dependency on ssh is there just in case drbd initialization hangs indefinitely - having ssh started already means you can log in and fix drbd. However, how likely a permanent hang is? Do the drbd maintainers care to comment? -- Marcin Owsiany <[EMAIL PROTECTED]> http://marcin.owsiany.pl/ GnuPG: 1024D/60F41216 FE67 DA2D 0ACA FC5E 3F75 D6F6 3A0D 8AA0 60F4 1216 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
