The upgrades from Ubuntu v.older to 9.04 went quite smoothly on every machine including the ones where I had to do several stepped upgrades since on various machines I skipped one or more upgrades prior to 9.04.

Rather than return to my old ways of just hosing packages (and, worse, “setup.py develop” symlinks) into /usr/local I’ve decided to use python-virtualenv to create some non-root-owned Python environments to work on my various Python-y apps. The main difference to me is being able to keep things I’m working on separate from one another.

We’ll see how it goes.

I’m afraid all these years of using package-managed software has made me soft — now a piece of software not having a nice package means I need to think a bit harder about if it’s worth dealing with the hassle of having that software outside of the package manager. (That applies to when I’d like to be using a more recent version of the package, too, though generally on Ubuntu I run into that a lot less frequently than I did on Debian.) I have mixed feelings about setuptools and easy_install.

I am pleased that now mod_wsgi is in a package so I can switch to that version rather than the one I installed by hand. mod_wsgi was worth it to install by hand to use until there was a nice package available. It’s Just Better than my previous mechanisms for running WSGI apps in, under, or behind Apache.

I investigated a little and found some evidence that Xen had some Issues with some motherboard chipsets when the machine was under load (as when it’s running the backups, for example). Curious, though, since I ran backups for an entire day before the last migration without a problem. Sadly, this happened as I ran out of energy for doing home sysadmin work, so my response after brief investigation was: migrate everything important back to the still-running “normal” Linux box and shut the Xen machine off.

Xen 3.0.3 was supposed to be out at he end of July and apparently will contain a fix for the bug that might have been causing my crashes. It might also have been LVM2/snapshot-related — I never relaly got the warm fuzzy feeling about LVM2 and snapshots after I had some trouble getting it working and it was only reported to work in very recent kernel revisions. I can’t believe the LVM folks rewrote everything (LVM -> LVM2) but broke snapshots for several releases.

Anyway, my Xen experiment has come to a close for the time being.

The machine shut down safely and came back up properly…. and the VNC viewer stayed connected. The state of the domain running the vncserver was restored as if the host machine hadn’t just been powered off. It’s one thing to know something is supposed to work a given way and another to actually see it happen.

I made a Debian sarge root image and swap following Creating a Debian VM with debootstrap. Then I cloned it and edited the hostname so I had two of them. Then, I configured the MAC address of their virtual adapters so I could set up static DHCP mappings and DNS entries on my LAN for them.

I wanted shared disk between my dom0 host and domX domains. NFS was, apparently, too easy for me, so I did some research into clustered filesystems to play with. I found two free possibilities that both looked reasonably mature and easy enough to install. GFS is from RedHat and OCFS2 is from Oracle. Both are open source. OCFS2 shipped with the kernel I already had built. GFS didn’t. So I decided to try OCFS2. I installed ocfs2-tools from Debian sarge. ocfs2_notes was a handy reference except I used ocfs2-tools from Debian sarge and the ocfs2 kernel module I built with the 2.6.16-xen kernel rather than installing from source as that guide shows. I set up a cluster named “ocfs2″ with my two Xen domains and the host in it.

It seemed to work when running against a loopback file. I then brought the machine into single user and used parted to resize one of the partitions to make some free space to use LVM2 against. I created a logical volume to put the OCFS2 partition that will be shared between the Domain0 host and some of the DomainUs.

My plan is to set up some domains to continue to test Xen over a longer period of time. Assuming it all goes well I’ll eventually migrate the services from the other machine(s) they currently run on into domains.