Short version: How do you (the multitudes) handle puppetized networking with nfs volumes?
Details: I'm tweaking work's puppetized networking for debian/ubuntu hosts. As part of this, I'd like better handling of networking changes when I have nfs mounts present on the server. Right now on all hosts (debian, ubuntu, centos) with puppetized networking will kill their networking and start it from scratch when there's a change. Conceptually it's a decent way of getting atomic networking changes. This is mostly fine (some less important daemons crash but monit takes care of that), but if there are nfs mounts the server will block hard and require rebooting. The problem is compounded by how if some daemon has an open filehandle to an nfs mount, we need to identify/stop the daemon in order to unmount the volume. The cleanest way I can think of handling this is to schedule a reboot right after laying down the new /etc/network/interfaces, but that seems excessive. (I figure on switching to http://forge.puppetlabs.com/razorsedge/network for the centos hosts, better than my current stuff.) -- You received this message because you are subscribed to the Google Groups "Puppet Users" group. To post to this group, send email to puppet-users@googlegroups.com. To unsubscribe from this group, send email to puppet-users+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/puppet-users?hl=en.
