rpc.statd starting before portmap

One problem gone, another one turns up. When rpc.statd (nfs-common) tries to start before portmap, it’s gonna result in failure. Now, the logfile (/var/log/daemon.log) is gonna print a rather cryptic error message:

After fixing the start order (I really hate *SUSE*/Debian* for not having init-script dependencies — like Gentoo’s baselayout/Roy’s openrc does have), everything is like it should be and I’m able to put the /srv/xen mount into the fstab

portmap hanging on shutdown

Here’s yet another post about my compute cluster. It’s (obviously) running NFS and that works quite well. Up till now, I would always have trouble with portmap hanging on shutdown/reboot. After spending some time thinking about the problem, looking at the init script and googling, I stumbled upon this Ubuntu bug on portmap.

As noted in the bug, a pmap_dump would hang indefinitely. After taking another look at our nfs-root configuration (in regard to the first comment on the bug), it turns out it’s exactly that. We didn’t setup lo which seems vital for some things.

After adding the lines

to /etc/network/interfaces, portmap stops just fine …

OFED packages for Debian

As I mentioned yesterday, I’m currently doing some project work. Said project includes InfiniBand technology.

Apparently we bought a “cheap” InfiniBand switch, which comes without a subnet manager. So, in order to communicate between the nodes, you need to install the subnet manager (opensm in my case) on each node.

In order to utilize the InfiniBand interface you need to do a few things first though:

  1. Obviously install the opensm package
  2. Add ib_umad and ib_ipoib to /etc/modules

After installing opensm on the host as well as the NFS root, opensm comes up just fine and the network starts automatically. Only trouble right now is, that ISC’s DHCP doesn’t support InfiniBand, otherwise I could even utilize DHCP to distribute the IP addresses.