Alan Hargreaves' Blog

The ramblings of a former Australian SaND TSC* Principal Field Technologist


Open Solaris

The Sydney Open Solaris User Group met last night in North Sydney. We held an abbreviated meeting for reasons that I’ll go into later.

First up I got to talk about setting up to build and bfu-ing. This got interesting as I mislaid my notes1, so my reference material was the Release Notes and my memory.

As a part of this talk I used a lab machine (dual proc v60x), accessed by a Sunray we had hooked up to the projector. We set up the build environment, discussed the various parts of the environment file (including the Vanity part of the build – putting your name into VERSION). I showed all parts of lead up to the build up to "nightly", but didn’t hit return as I “had one I’d prepared earlier”.

From here we went into a brief discussion of the nightly log and what you can expect to see in it.

We then went on to BFU. A short talk about what it is, how it works and how the bfu archives are generated for Open Solaris was followed by bfu-ing a live upgrade partition that I had setup earlier.

While this was running I brought up something that I’m actually surprised that no-one has really talked about. My source browser of choice (cscope-fast) is included in /opt/onbld (specifically in /opt/onbld/bin/`uname -p`/cscope-fast) and building the indexes is absolutely trivial. Built the indexes for usr/src/uts (make cscope.out) and we jumped around the code a bit looking for definitions, what calls things and where assignments are made to variables. For anyone interested, my options of choice for cscope-fast are "-dqrp 5".

By this stage, the bfu had finished, so we activated what I thought was the right partition and rebooted. You guessed it, I got the wrong partition (I had been doing something else with on this machine). Activated the right partition, rebooted and paniced in an assertion. By this stage everyone was pretty happy with the process and I didn’t need to go and add

set aok = 1

into /etc/system and try again, so we moved on to Boyd’s presentation (I can’t believe that my presentation went for about 80 minutes!).

Boyd did a great presentation on SMF using my notebook. Showing how to use various commands, disabling/enabling stuff, and also discussing a manifest that he’s done up for CUPS. I’m going to try to talk him into making his slides available and perhaps writing some words to accompany them. He’s teaching in this building this week, so I can get to hassle him a bit 😉

As mentioned earlier, we adjourned at about 8:15 to all head down to the Victoria Cross Bar as we managed to choose the same night as the State of Origin Rugby League decider. Almost all of us had a good time watching the game and sharing a drink.

We only had 11 people show up for this one, but I think we can put that down to the conflict with the game.

It was nice to present in an intimate atmosphere. The talks were both very interactive. It’s just a shame that we didn’t have a video camera last night.

1. This needs a little explanation. I did not lose a presentation on dead trees. My “notes” were about 20 lines of bullet points to remind me of what I had to cover. There were stored on my notebook. Unfortunately i both forgot the name of the file, and it appears that I was not in the directory that I thought I was when I put them together.

Technorati Tags:


Written by Alan

July 3, 2005 at 4:55 pm

Posted in OpenSolaris

2 Responses

Subscribe to comments with RSS.

  1. Alan:
    Thanks for sharing, and for organizing the Sydney OpenSolaris User Group. Did you take any still photos per chance? If so, please post to and tag with OpenSolaris… Thanks again.

    Claire Giordano

    July 8, 2005 at 12:33 am

  2. Sorry Claire, no-one had a camera and the video camera we used last meeting was on holidays with it’s owner. It’s a shame as I was pretty happy with the two talks given.

    Alan Hargreaves

    July 8, 2005 at 12:44 am

Comments are closed.

%d bloggers like this: