OpenBSD Journal

Journal from NYCBSDCon 2006

Contributed by jason on from the save-a-dead-tree dept.

Will Backman of bsdtalk gives a recount of his weekend at NYCBSDCon 2006:

I arrived in New York City on a sunny Friday afternoon, and after dropping my stuff off with family, headed up to the Amsterdam Restaurant with my brother for the informal pre-conference dinner. Most of the speakers were there, and I shared a table with .ike, Michael, my brother Tom, Okan, and Mark. After the meal, people headed downstairs to the noisy bar area for a few beers. I didn't stay too late because of the early start for the conference, but it was a great opportunity to put some faces to names.

On Saturday morning, with heavy wind and rain in full swing, I headed up to Columbia University. My brother and I had both assumed that the other knew where we were going, but luckily there were public computer terminals running Linux in one of the campus buildings. I resisted the urge to install BSD. After getting directions, we headed across campus, and ran into a very wet Marco and Bob on the way. Registration was easy and there were plenty of bagels and assorted beverages, although it took a while for the tables in the hall to fill up and for wet attendees to trickle in. The conference room was nice, and besides some microphone problems, the presentation technology worked perfectly. I liked the single track for presentations, although there were a few BOF sessions happening at the same time as some of the presentations. I wish I could have made it to the BOF on breaking captcha, but I did go to a really good BOF session by Adam Martin about his Google Summer of Code work on AutoFS for FreeBSD. The audio of the presentations was recorded by both Nikolai and Damien, and is available along with most of the slides at the NYCBSDCon website.

The first presentation was from Corey Benninger, who discussed Security with Ruby on Rails on BSD. I haven't had much exposure to Ruby on Rails, so it was great to learn about some of the basic secure coding tools and insecure defaults. I was also fascinated by the Tamper Data Firefox Extension which allows for the easy modification of HTTP variables on the fly.

Next was a talk by Brian A. Seklecki about the challenges of making NetBSD network appliances. He has put a lot of work into solving the challenges of working with minimal builds that run on Compact Flash. He also provided some of the nifty hardware on display at the NetBSD table.

Bob Beck gave an enlightening talk on PF and how he uses it to abstract his services away from their public IPs. A couple of CARPed routers allows him to transparently load balance, scale, and change his services without needing to notify the users. He succeeded in changing how I think about PF, and I can see how even a small network could benefit from the flexibility of CARP and a few rdr rules.

After a catered lunch, Bjorn Nelson presented his build system for FreeBSD servers. His formula uses a central build server, mergemaster, and some custom scripts to allow for quick updates and upgrades. He showed his bravery by connecting remotely to some servers and trying some builds.

Next, Johnny Lam discussed software dependency problems and how he tried to solve them with pkgsrc. What I learned from this presentation is that there are no easy answers and that pkgsrc is an interesting prototype solution. There doesn't seem to be a lot of love for the different build tools across operating systems, and trying to keep pkgsrc portable presents an additional set of challenges.

Marco Peereboom spoke about bio and sensors on OpenBSD, and it is obvious that a lot of work is going into supporting hardware without any help from the manufacturers. Bioctl and sensorsd have done a great job of replacing the complicated vendor supplied RAID management stacks with consistent management and monitoring tools. Marco encouraged people to demand unrestricted documentation from vendors and to avoid binary blobs and Non-Disclosure Agreements.

After the break, I attended the AutoFS BOF by Adam Martin and had to miss the presentation by Russell Sutherland. Adam is a great speaker and demonstrates the qualities that Google tries to foster with their Summer of Code. There was a lot of interest in the room for a quality AutoFS daemon, and I think this project will be a great benefit to the community.

Jason Dixon closed with a very funny and fast-paced satire about the death of BSD. It really has to been seen, and Jason has posted video of the slides synced with the audio. He spoke about the history of Unix, industry reports of its death, and future goals. Except for a few slides, it would make a fine presentation for management.

After the presentations finished, everyone headed over to the Amsterdam Restaurant for appetizers, food, and beer. The entire downstairs had been reserved for the Con, so people had a good opportunity to mingle. A few people were even hacking on code in the middle of it all. I spent a lot of time talking with people who use BSD at work, and I hope to catch up with these people at a later time for interviews. I also learned that not everybody likes Jack Daniels.

Day two started off with with better weather and Jason Wright's talk on the efforts to support Sparc64 hardware in OpenBSD. It was good to hear about code sharing between the different BSDs, and I also learned how many chips besides the CPU are involved in a port. Once again the topic of the lack of vendor documentation came up, along with a few offers of Sun hardware for testing. It is really amazing how much effort people are willing to put into creating demand for hardware from a vendor that doesn't want to help.

Kristaps Johnson came all the way from Latvia to talk about the sysjail project, which uses systrace to provide userland virtualization similar to FreeBSD jails. It should work on any operating system that supports systrace, although it is currently developed on OpenBSD. Testing sysjail is as easy as downloading a package, and the sysjail website has been running in it successfully for a long time. Kristaps even demonstrated Linux emulation running inside a sysjail, although this is just a preliminary implementation.

Wietse Venema gave an interesting talk on the design goals and development history of the Postfix MTA. Many of the techniques such as least privilege and a plug-in architecture have served to create a secure service that is able to grow with the feature creep of the Internet. He did show some willingness to revisit the licensing issues to possibly make it more BSD compatible, but stressed that the decision was not up to him.

Bob Beck gave an energetic post-lunch presentation on the spamd greylisting daemon. He described the characteristics of spam, the benefits of greylisting, and the design goals of spamd. It was very interesting to hear his real world statistics from the mail servers at the University of Alberta. He also described some new tools he is developing to cull additional greylist information out of the logs, and I look forward to trying them once they are released. I'm also looking forward to greylist synchronization between firewalls, and maybe even between organizations.

Murray Stokely gave the last talk of the Conference, where he spoke about the distributed mid-layer at Google and the BSD-related Summer of Code projects. The problem set that Google deals with is way beyond anything that I would have to deal with, but it is great to see a long-time BSD developer doing great work inside such an influential company. He did mention that every BSD related project in this year's Summer of Code made it through to the satisfaction of the mentors. I guess this it is quite unusual, so kudos to the BSD community.

During the conference, there were tables for [Open|Free|Net]BSD, New York Internet, Addison Wesley, Google, Everest Broadband/Promenet, and BSDCertification. Every table had stuff to buy or take for free, and the NetBSD table had the usual display of unusual hardware including a SEGA Dreamcast. Dru Lavigne had a bunch of BSDCertification DVDs that included the installers for the four major BSDs in addition to documentation.

The conference ended with a raffle, and they gave away a lot of great books and CDs. I won a 3rd Edition of the FreeBSD Handbook Users Guide. Between all the food and prizes, the cost of the conference was well worth it.

(Comments are closed)


Comments
  1. By Anonymous Coward (69.70.207.240) on

    I'd like to give a big THANK YOU to Will Backman for all his efforts and his contributions as a whole. Great work, good job - simply put, awesome stuff!

    Comments
    1. By George (71.249.91.105) georgeATnycbugDOTorg on http://www.nycbug.org

      > I'd like to give a big THANK YOU to Will Backman for all his efforts and his contributions as a whole. Great work, good job - simply put, awesome stuff!

      Absolutely. . . Will has made himself into a vital part of the BSD community, IMHO.

      Very resourceful approach to being a media outlet for the BSDs.

      And having Will and his brother at the con was great. It was doubly an honor to have him provide such a detailed overview. . .

  2. Comments
    1. By Jason L. Wright (134.20.35.80) jason@openbsd.org on http://www.thought.net/jason

      > From spamd presentation, the url to greyscan gives a 404.
      > http://www.ualberta.ca/~beck/greyscan


      Try: http://www.ualberta.ca/~beck/nycbug06/

      There's even a new script that hooks into greyscanner which will do local address validation based on a "default" install of OpenBSD (/etc/passwd + /etc/mail/aliases + /etc/mail/local-host-names)

  3. By Anonymous Coward (128.171.90.200) on

    Cheers for posting a link for Jason Dixon's presentation

    Comments
    1. By Anonymous Coward (24.82.182.29) on

      > Cheers for posting a link for Jason Dixon's presentation

      Agreed, that was actually really funny.

      Comments
      1. By Anonymous Coward (38.113.22.50) on

        > > Cheers for posting a link for Jason Dixon's presentation
        >
        > Agreed, that was actually really funny.

        I thought it was a nice way to end the day. A funny, nontechnical talk before everyone went off to the pub (which, unfortunately, I was unable to attend). Kudos to NYCBUG for putting on a solid conference. I had a great time and I learned a lot.

        Comments
        1. By Jeroen Janssen (213.84.80.73) on http://efantasy.be

          > > > Cheers for posting a link for Jason Dixon's presentation
          > >
          > > Agreed, that was actually really funny.
          >
          > I thought it was a nice way to end the day. A funny, nontechnical talk before everyone went off to the pub (which, unfortunately, I was unable to attend). Kudos to NYCBUG for putting on a solid conference. I had a great time and I learned a lot.

          Wasn't there, but oh boy, what a nice ending. Thanks Jason Dixon for fixing my bad off-day...

  4. By Sevan / Venture37 (82.134.90.244) venture37 (AT) hotmail (DONT) com on www.geeklan.co.uk

    Nice one Will, good stuff, though I have to say I am a little disappointed with one thing in your write up:

    > I also learned that not everybody likes Jack Daniels.

    tut tut, shocking stuff, who are these people! ;)

    Jason Dixons presentation made it onto digg a couple of days ago
    http://video.google.com/videoplay?docid=7833143728685685343

    Comments
    1. By Bob Beck (129.128.11.43) beck@openbsd.org on

      > Nice one Will, good stuff, though I have to say I am a little disappointed with one thing in your write up:
      >
      > > I also learned that not everybody likes Jack Daniels.
      >
      > tut tut, shocking stuff, who are these people! ;)
      >

      Some of us like proper whiskey[1], not that vile corn mash which
      is an excuse to make something out of corn that your cows wouldn't
      eat.

      [1] Which is made only from barley [2]
      [2] ... and must come from Scotland.

      8-P

      Bob




      Comments
      1. By CODOR (67.158.67.218) on

        > Some of us like proper whiskey[1], not that vile corn mash which
        > is an excuse to make something out of corn that your cows wouldn't
        > eat.
        >
        > [1] Which is made only from barley [2]
        > [2] ... and must come from Scotland.

        No whiskey comes from Scotland. Whisky, however, does...

  5. By CS (216.220.116.154) on

    Not to play favorites or anything, but Bob Beck kicks ass. The spamd/greylisting talk was the best overview I've ever seen. Plus the fact that he's running this stuff in production on a big chunk of machines gives me the warm fuzzies.

    I'm mostly a FreeBSD person but all the OBSD people I briefly met were extremely friendly and smart in that scary sort of way. These people so aren't the NYLUG nerds.

  6. By Anonymous Coward (195.29.148.251) on

    I'm totally impressed with Bob Beck's work with pf, loadbalancing, CARP and pfsync. Using these goodies it is possible to have ip_sticky-tcp_preserving-redundant loadbalancer all in a few pf.conf lines.

    Great design !!

Latest Articles

Credits

Copyright © - Daniel Hartmeier. All rights reserved. Articles and comments are copyright their respective authors, submission implies license to publish on this web site. Contents of the archive prior to as well as images and HTML templates were copied from the fabulous original deadly.org with Jose's and Jim's kind permission. This journal runs as CGI with httpd(8) on OpenBSD, the source code is BSD licensed. undeadly \Un*dead"ly\, a. Not subject to death; immortal. [Obs.]