Content here is by Michael Still All opinions are my own.
See recent comments. RSS feed of all comments.

Fri, 04 Jan 2013

OpenStack at 2013

    As some of you might know, I'm the Director for 2013. I've tried really hard to not use my powers for evil and make the entire conference about OpenStack -- in fact I haven't pulled rank and demanded that specific content be included at all. However, the level of interest in OpenStack has grown so much since LCA 2012 that there is now a significant amount of OpenStack content in the conference without me having to do any of that.

    I thought I'd take a second to highlight some of the OpenStack content that I think is particularly interesting -- these are the talks I'll be going to if I have the time (which remains to be seen):

    • Cloud Infrastructure, Distributed Storage and High Availability Miniconf: while not specifically about OpenStack, this miniconf is going to be a good warm up for all things IaaS at the conference. Here's a list of the talks for that miniconf:
        Delivering IaaS with Apache CloudStack - Joe Brockmeier
      • oVirt - Dan Macpherson
      • Aeolus - Dan Macpherson
      • Ops: From bare metal to cloud space - Phil Ingram
      • VMs on VLANs on Bridges on Bonds on many NICs - Kim Hawtin
      • OpenStack Swift Overview - John Dickinson
      • JORN and the rise and fall of clustering - Jamie Birse
      • MongoDB Replication & Replica Sets - Stephen Steneker
      • MariaDB Galera Cluster - Grant Allen
      • The Grand Distributed Storage Debate: GlusterFS and Ceph going head to head - Florian Haas, Sage Weil, Jeff Darcy

    • The OpenStack Miniconf: this is a mostly-clear winner for Tuesday. Tristan Goode has been doing a fantastic job of organizing this miniconf, which might not be obvious to people who haven't been talking to him a couple of times a week about its progress like me. I think people will be impressed with the program, which includes:
      • Welcome and Introduction - Tristan Goode
      • Introduction to OpenStack - Joshua McKenty
      • Demonstration - Sina Sadeghi
      • NeCTAR Research Cloud: OpenStack in Production - Tom Fifeld
      • Bare metal provisioning with OpenStack - Devananda van der Veen
      • Intro to Swift for New Contributors - John Dickinson
      • All-around OpenStack storage with Ceph - Florian Haas
      • Writing API extensions for Nova - Christopher Yeoh
      • The OpenStack Metering Project - Angus Salkeld
      • Lightweight PaaS on the NCI OpenStack Cloud - Kevin Pulo
      • Enabling Compute Clusters atop OpenStack - Enis Afgan
      • Shared Panel with Open Government
    • The Open Government Miniconf: this is the other OpenStack relevant miniconf on Tuesday. This might seem like a bit of a stretch, but as best as I can tell there is massive interest in government at the moment in deploying cloud infrastructure, and now is the time to be convincing the decision makers that open clouds based on open source are the right way to go. OpenStack has a lot to offer in the private cloud space, and we need to as a community make sure that people are aware of the various options that are out there. This is why there is a shared panel at the end of the day with the OpenStack miniconf.

      There aren't any OpenStack talks on Wednesday, but I am really hoping that someone will propose an OpenStack BoF via the wiki. I'd sure go to a BoF.

    • Playing with OpenStack Swift by John Dickinson
    • Ceph: Managing A Distributed Storage System At Scale by Sage Weil

    • Openstack on Openstack - a single management API for all your servers by Robert Collins
    • Heat: Orchestrating multiple cloud applications on OpenStack using templates by Angus Salkeld and Steve Baker
    • How OpenStack Improves Code Quality with Project Gating and Zuul by James Blair
    • Ceph: object storage, block storage, file system, replication, massive scalability, and then some! by Tim Serong and Florian Haas

    So, if you're interested in OpenStack and haven't considered 2013 as a conference you might be interested in, now would be a good time to reconsider before we sell out!

    Tags for this post: openstack conference lca2013 rackspace
    Related posts: Yet more lca2013 setup; I give you Fidel, the slightly questionable LCA2013 organizer; Call for presentations for the 2014 OpenStack mini-conference; Moving on; Contact details for the Canberra LCA 2013 bid; Are you in a LUG? Do you want some promotional materials for LCA 2013?

posted at: 13:07 | path: /openstack | permanent link to this entry

Sat, 22 Dec 2012

Image handlers (in essex)

    George asks in the comments on my previous post about loop and nbd devices an interesting question about the behavior of this code on essex. I figured the question was worth bringing out into its own post so that its more visible. I've edited George's question lightly so that this blog post flows reasonably.
    Can you please explain the order (and conditions) in which the three methods are used? In my Essex installation, the "img_handlers" is not defined in nova.conf, so it takes the default value "loop,nbd,guestfs". However, nova is using nbd as the chose method.
    The handlers will be used in the order specified -- with the caveat that loop doesn't support Copy On Write (COW) images and will therefore be skipped if the libvirt driver is trying to create a COW image. Whether COW images are used is configured with the use_cow_images flag, which defaults to True. So, loop is being skipped because you're probably using COW images.
    My ssh keys are obtained by cloud-init, and still whenever I start a new instance I see in the nova-compute.logs this sequence of events:
    qemu-nbd -c /dev/nbd15 /var/lib/nova/instances/instance-0000076d/disk 
    kpartx -a /dev/nbd15 
    mount /dev/mapper/nbd15p1 /tmp/tmpxGBdT0 
    umount /dev/mapper/nbd15p1 
    kpartx -d /dev/nbd15 
    qemu-nbd -d /dev/nbd15 
    I don't understand why the mount of the first partition is necessary and what it happens when the partition is mounted.
    This is a bit harder than the first bit of the question. What I think is happening is that there are files being injected, and that's causing the mount. Just because the admin password isn't being inject doesn't mean that other things aren't being injected still. You'd be able to tell what's happening by grepping your logs for "Injecting .* into image" and seeing what shows up.

    Tags for this post: openstack loop nbd libvirt file_injection rackspace
    Related posts: Some quick operational notes for users of loop and nbd devices; OpenStack at 2013; Call for presentations for the 2014 OpenStack mini-conference; Moving on; Faster pip installs; Upgrade problems with the new Fixed IP quota

posted at: 15:51 | path: /openstack | permanent link to this entry

Sat, 15 Dec 2012

The Forever War (again)

posted at: 21:58 | path: /book/Joe_Haldeman | permanent link to this entry

Some quick operational notes for users of loop and nbd devices

posted at: 16:28 | path: /openstack | permanent link to this entry

Thu, 13 Dec 2012


posted at: 18:45 | path: /book/Robert_L_Forward | permanent link to this entry

Sat, 08 Dec 2012

Moving on

posted at: 12:56 | path: /openstack | permanent link to this entry

Wed, 14 Nov 2012

Fuzzy Nation

    ISBN: 9780765367037
    Yet another excellent Scalzi novel. This one took me a while to really warm up to, but it was worth the patience. The ending is fast paced and excellent.

    Tags for this post: book john_scalzi sentience mining colony

posted at: 22:51 | path: /book/John_Scalzi | permanent link to this entry


posted at: 22:43 | path: /book/Suzanne_Collins | permanent link to this entry

Fri, 21 Sep 2012

On conference t-shirts

    Conference t-shirts can't be that hard, right? I certainly don't remember them being difficult when Canberra last hosted in 2005. I was the person who arranged all the swag for that conference, so I should remember. Yet here I am having spent hours on the phone with vendors, and surrounded with discarded sample t-shirts, size charts and colour swatches. What changed?

    The difference between now and then is that in the intervening seven years the Australian Linux community has started to make real effort to be more inclusive. We have anti-harassment policies, we encourage new speakers, and we're making real efforts to encourage more women into the community. 2013 is making real efforts to be as inclusive as possible -- one of the first roles we allocated was a diversity officer, who is someone active in the geek feminism community. We've had serious discussions about how we can make our event as friendly to all groups as possible, and have some interesting things along those lines to announce soon. We're working hard to make the conference a safe environment for everyone, and will have independent delegate advocates available at all social events, as well as during the conference.

    What I want to specifically talk about here is the conference t-shirts though. We started out with the following criteria -- we wanted to provide a men's cut, and a separate women's cut, because we recognize that unisex t-shirts are not a good solution for most women. We also need a wider than usual size range in those shirts because we have a diverse set of delegates attending our event. We also didn't really want to do black, dark blue, or white shirts -- mostly because those colours are overdone, but also because the conference is in January when the mean temperature is around 30 degrees Celsius.

    Surprisingly, those criteria eliminate the two largest vendors of t-shirts in Australia. Neither Hanes nor Gildan make any t-shirt that has both men's and women's cuts, in interesting colours and with a large size variety. So we went on the hunt for other manufacturers. However, I'm jumping a little ahead of myself here, so bear with me.

    First off we picked a Hanes shirt because we liked the look of it. We were comfortable with that choice for quite a while before we discovered that the range of colours available in both the men's and women's cut was quite small. Sure, there are heaps of colours in each cut, but the overlapping set of colours is much smaller than it first appears. At this point we knew we needed to find a new vendor.

    The next most obvious choice is Gildan. Gildan does some really nice shirts, and I immediately fell in love with a colour called "charcoal". However, once bitten twice shy, so we ordered some sample t-shirts for my wife and I to try out. I'm glad we did this, because the women's cut was a disaster. First off it didn't fit my wife very well in the size she normally wears, which it turns out is because the lighter cotton style of t-shirt is 10 centimeters smaller horizontally than the thicker cotton version! It got even worse when we washed the shirts and tried them again -- the shirt shrunk significantly on first wash. We also noticed something else which had escaped our attention -- the absolute largest size that Gildan did in our chosen style for women was a XXL. Given the sizing ran small, that probably made the largest actual size we could provide a mere XL. That's not good enough.

    Gildan was clearly not going to work for us. I got back on the phone with the supplier who was helping us out and we spent about an hour talking over our requirements and the problems we were seeing with the samples. We even discussed getting a run of custom shirts made overseas and shipped in, but the timing wouldn't work out. They promised to go away and see what other vendors they could find in this space. Luckily for us they came back with a vendor called BizCollection, who do soft cotton shirts in the charcoal colour I like.

    So next we ordered samples of this shirt. It looked good initially -- my shirt fit well, as did my wife's. However, we'd now learnt that testing the shirts through a few wash cycles was useful. So then my wife and I wore the shirts as much as we could for a week, washing them each evening and abusing them in all the ways we could think of -- using the dryer, hanging them outside in the sun, pretty much everything apart from jumping up and down on them. I have to say these shirts have held up well, and we're very happy with them.

    The next step is I'm going to go back and order a bunch more sample shirts and make my team wear them. The goal here is to try and validate the size charts that the vendor provides and make sure that we can provide as much advice about fit as possible to delegates. Also, I love a free t-shirt.

    After all this we still recognize that some people will never be happy with the conference's t-shirt. Perhaps they hate the colour or the design, or perhaps they're very tall and every t-shirt is too short for them. So the final thing we're doing is we're giving delegates a choice -- they can select between a t-shirt, a branded cap, or a reusable coffee cup. In this way we don't force delegates to receive something they don't really want and are unlikely to use.

    When you register for the conference, please try to remember that we've put a lot of effort as an organizing team into being as detail oriented as possible with all the little things we think delegates care about. I'm sure we've made some mistakes, but we are volunteers after all who are doing our best. If you do see something you think can be improved I'd ask that you come and speak to us privately first and give us a chance to make it right before you complain in public.

    Thanks for reading my rant about conference t-shirts.

    Tags for this post: conference lca2013 swag t-shirts canonical
    Related posts: Wow, qemu-img is fast; Reflecting on Essex; Moving on; Are you in a LUG? Do you want some promotional materials for LCA 2013?; Announcement video; Taking over a launch pad project

posted at: 14:42 | path: /conference/lca2013 | permanent link to this entry

Sat, 08 Sep 2012

The Tuloriad

posted at: 22:14 | path: /book/John_Ringo | permanent link to this entry

Wed, 22 Aug 2012

Yellow Eyes

posted at: 20:01 | path: /book/John_Ringo | permanent link to this entry

Sat, 04 Aug 2012

Watch on the Rhine

    ISBN: 9781416521204
    If you knew you were in deep trouble, had the technology to rejuvenate any soldier you wanted, and happened to be a late nineties Germany desperate for cannon fodder, would you return the SS to service? A harsh reality is that they're some of the only soldiers you have left with real combat experience, even if their politics is abhorrent. This book has an interesting underlying concept, but to a certain extent its ruined by the politics of the authors -- any concern for anything other that military strength is dismissed as another example of rampant nimbyism. However, the book tells a good story and made me think about some stuff I wouldn't have otherwise thought about, while being entertaining. So, overall a success I guess.

    Tags for this post: book john_ringo tom_kratman aliens combat personal_ai rejuv legacy_of_the_aldenata germany
    Related posts: The Tuloriad; Yellow Eyes; Hell's Faire; A Hymn Before Battle; Cally's War; Gust Front

posted at: 18:13 | path: /book/John_Ringo | permanent link to this entry

Sat, 28 Jul 2012


    ISBN: 044100105x
    While it isn't immediately obvious, this book is quite similar to Johnny Mnemonic. The brain damage aspect is played up a bit, and gets repetitive, but the overall story is interesting and fun, even if the ending is a bit obvious from about half way through.

    Tags for this post: book william_c_dietz combat corporations

posted at: 23:34 | path: /book/William_C_Dietz | permanent link to this entry

Tue, 10 Jul 2012

A first pass at glance replication

    A few weeks back I was tasked with turning up a new OpenStack region. This region couldn't share anything with existing regions because the plan was to test pre-release versions of OpenStack there, and if we shared something like glance then we would either have to endanger glance for all regions during testing, or not test glance. However, our users already have a favorite set of images uploaded to glance, and I really wanted to make it as easy as possible for them to use the new region -- I wanted all of their images to magically just appear there. What I needed was some form of glance replication.

    I'd sat in on the glance replication session at the Folsom OpenStack Design Summit. The NeCTAR use case at the bottom is exactly what I wanted, so its reassuring that other people wanted something like that too. However, no one was working on this feature. So I wrote it. In fact, because of the code review process I wrote it twice, but let's not dwell on that too much.

    So, as of change id I7dabbd6671ec75a0052db58312054f611707bdcf there is a very simple replicator script in glance/bin. Its not perfect, and I expect it will need to be extended a bunch, but its a start at least and I'm using it in production now so I am relatively confident its not totally wrong.

    The replicator supports the following commands at the moment:

    glance-replicator livecopy fromserver:port toserver:port
        Load the contents of one glance instance into another.
        fromserver:port: the location of the master glance instance.
        toserver:port:   the location of the slave glance instance.

    This is the main meat of the replicator. Take a copy of the fromserver, and dump it onto the toserver. Only images visible to the user running the replicator will be copied if you're using Keystone. Only images active on fromserver are copied across. The copy is done "on-the-wire", so there are no large temporary files on the machine running the replicator to clean up.

    glance-replicator dump server:port path
        Dump the contents of a glance instance to local disk.
        server:port: the location of the glance instance.
        path:        a directory on disk to contain the data.

    Do the same thing as livecopy, but dump the contents of the glance server to a directory on disk. This includes meta data and image data, and this directory is probably going to be quite large so be prepared.

    glance-replicator load server:port path
        Load the contents of a local directory into glance.
        server:port: the location of the glance instance.
        path:        a directory on disk containing the data.

    Load a directory created by the dump command into a glance server. dump / load was originally written because I had two glance servers who couldn't talk to each other over the network for policy reasons. However, I could dump the data and move it to the destination network out of band. If you had a very large glance installation and were bringing up a new region at the end of a slow link, then this might be something you'd be interested in.

    glance-replicator compare fromserver:port toserver:port
        Compare the contents of fromserver with those of toserver.
        fromserver:port: the location of the master glance instance.
        toserver:port:   the location of the slave glance instance.

    What would a livecopy do? The compare command will show you the differences between the two servers, so its a bit like a dry run of the replication.

    glance-replicator size 
        Determine the size of a glance instance if dumped to disk.
        server:port: the location of the glance instance.

    The size command will tell you how much disk is going to be used by image data in either a dump or a livecopy. It doesn't however know about redundancy costs with things like swift, so it just gives you the raw number of bytes that would be written to the destination.

    The glance replicator is very new code, so I wouldn't be too surprised if there are bugs out there or obvious features that are lacking. For example, there is no support for SSL at the moment. Let me know if you have any comments or encounter problems using the replicator.

    Tags for this post: openstack glance replication multi-region canonical
    Related posts: Further adventures with base images in OpenStack; Openstack compute node cleanup; Wow, qemu-img is fast; Reflecting on Essex; Moving on; Are you in a LUG? Do you want some promotional materials for LCA 2013?

posted at: 16:09 | path: /openstack | permanent link to this entry

Sun, 20 May 2012

Got Something to Say? The LCA 2013 CFP Opens Soon!

    The call for presentations opens on 1 June, which is only 11 days away! So if you're thinking of speaking at the conference (a presentation, tutorial, or miniconference), now would be a good time to start thinking about what you're going to say. While you're thinking, please spare a thought for our web team, who are bringing up the entire zookeepr instance so that the CFP will work properly.

    We've been getting heaps of stuff done over the past few months. We've had a "ghosts" meeting (a meeting with former LCA directors), found conference and social venues, and are gearing up for the Call For Presentations.

    We've signed a contract for the keynote venue, which I think you will all really enjoy. We have also locked in our booking for the lecture theatres, which is now working its way through the ANU process. For social events, we've got a great venue for the penguin dinner, and have shortlisted venues for the speakers' dinner and the professional delegates' networking session. We're taking a bit of extra time here because we want venues that are special, and not just the ones which first came to mind.

    The ghosts meeting went really well and I think we learnt some important things. The LCA 2013 team is a bit unusual, because so many of us have been on a LCA core team before, but that gave us a chance to dig into things which deserved more attention and skip over the things which are self-evident. We want to take the opportunity in 2013 to have the most accessible, diverse and technically deep conference that we possibly can, and there was a lot of discussion around those issues. We've also had it drummed into us that communications with delegates is vitally important and you should expect our attempts to communicate to ramp up as the conference approaches.

    I'm really excited about the progress we've made so far, and I feel like we're in a really good state right now. As always, please feel free to contact the LCA2013 team at if you have any questions.

    Tags for this post: conference lca2013 cfp canonical
    Related posts: Call for papers opens soon; Wow, qemu-img is fast; Reflecting on Essex; Moving on; Are you in a LUG? Do you want some promotional materials for LCA 2013?; Announcement video

posted at: 20:44 | path: /conference/lca2013 | permanent link to this entry

Thu, 10 May 2012

Catching Fire

posted at: 04:55 | path: /book/Suzanne_Collins | permanent link to this entry

Sun, 22 Apr 2012


posted at: 19:37 | path: /book/Ben_Bova | permanent link to this entry

Mon, 16 Apr 2012

The Hunger Games

posted at: 14:04 | path: /book/Suzanne_Collins | permanent link to this entry

Sat, 14 Apr 2012

The Android's Dream

posted at: 13:48 | path: /book/John_Scalzi | permanent link to this entry

Logos Run

    ISBN: 0441015360
    This is the continuation from Runner, and continues the story of the attempt to re-enable the star gates. It has the comicly incompetent Technosociety once again, as well as series of genetically engineered protagonists. I am bothered by why the star gate power supplies cause people to fall ill -- you'd think in a highly advanced society capable of building star gates they might have spent some time on shielding. Or did the shielding somehow fail on all the power sources sometime over the thousands of years of decay? The has a disappointing ending, but was a fun read until then. I find it hard to suspend disbelief about how the AIs present themselves, but apart from that the book was solid. This one is probably not as good as the first.

    Tags for this post: book william_c_dietz religion combat space_travel decay courier engineered_human genetic_engineering runner_series
    Related posts: Runner; Friday ; The Accidental Time Machine ; Rendezvous With Rama; Buying Time; Marsbound

posted at: 13:45 | path: /book/William_C_Dietz | permanent link to this entry

Previous page