|
|
MSOSUG Past Meeting Details, Materials and A/V12th of September 2007:
This meeting was a great start for the group. Huge attendence, lots of interaction and heaps of questions. Boyd's Live Upgrade talk showed all who are NOT yet using it why they should. Nathan's N2 talk whet everyone's appetite for the upcoming Niagara 2 platform. (Update — It's released now!) 17th of October 2007:
What a wild ride this was! With many questions from the audience leading the talk in a number of interesting and unexpected directions, the whole crowd (including Nathan!) were treated to many weird permutations of ZFS and iSCSI, including presenting a multiple ZVOLs from a Zpool living on a USB stick as iSCSI targets, then using said targets to create a Zpool. This also aptly demonstrated a number of the golden reasons you don't want to be using non-replicated storage from the ZFS perspective. Doing unnatural things begets natural consequences… Panics, for instance. :) As Nathan was the only speaker for the night, he was forgiven for talking for 2 hours. :) 21st of November 2007:
Once again, the importance of reliable, well resourced hardware was demonstrated. With two panics for the night, we enjoyed not only ZFS but the rough and ready pain of an unreliable set of iSCSI targets. Unfortunately, time prevented us from determining what was happening, but given that they were being served up by a Sunblade 100 with only 256MB of memory, some of us were not particularly surprised. :) Boyd's excellent discussion of the ZFS layout, block update mechanism, snapshots and send/receive were well received. Questions, as usual, caused us to run quite late, with the night wrapping at 9:30 and the last of us finally exiting, 'stage left', at around 10:30pm! Boyd also took back all the bad things he said about Nathan for running too long in the first meeting!!! 19th of December 2007:
Andre's SSGD talk was great. We saw Solaris, Linux and Windows applications deployed to any desktop. As is always the case, Windows took some time to get in the way of a perfect run, but that's what you get for running windows. Even over the questionable network connection we had on the night, SSGD was great. (Gotta love the whole statelessness thing. Gold. Atul's chat on Lustre got us all thinking about what we could use such a setup for. I'm sure even now, we have folks thinking about new and interesting ways they could implement an storage system, and what they could do with that much available storage bandwidth and capacity… :) Thanks, Atul. Nathan's chat on Xen went down well, though there were some grunts about it not being 'available' yet. :) heh. Bring on Xen in a supported Solaris! We saw Windows, Linux, Solaris and Solaris Paravirtualised, on a whitebox system with only 4GB memory. All ran well, though we could have used a little more grunt from the CPU. But – how much is it fair to expect from a sub $100 (AUS) CPU? :) 16th of January 2008:
Andre's jumpstart talk was a real conversation. We covered very simple jumpstart, then swung straight into JET, JET capabilities, configuration and benefits. I'm sure those not using JET yet will be looking at it real soon… Nathan's chat on LDOMs was lively with lots of jumping and hand waving. heaps of questions and lots of great discussion. We say LDOM's created, booted, rebooted, and even a Service/Control domain reboot that showed that the LDOMS would pause if they needed to do IO, but recovered as soon as the service LDOM was back. Also saw why it's good to check out your SSH config before the meeting… ahem… :) 20th of February 2008:
David's talk on N1SPS and JET was an eye-opener. Talk about cool – We look at a production deployment, and jumped a system to see the end to end process of server provisioning. Dan's Extended Attribute chat got lots of folks thinking about how they could use extended attributes to do cool and funky things. Some considered the evil they could do with them… An interesting chat with lots of good examples. Nathan's look at virtual box was a simple dome of creating a new virtual system, and starting up the windows installer. Also demo'd a running Windows system. I'm sure we are all looking forward to more from the Innotek guys in this space. 19th of March 2008:
Although Richard Smith had pre-warned some folks that his talk might be a little esoteric, it turns out that it was very interesting to pretty much everyone there. We were transfixed with cache miss statistics, IPC (Instructions per Cycle) and other magnificence. Richard Spindler's talk on Jumpstart troubleshooting was also great. He covered the most likely things that go wrong, and even showed some of the oldies a new trick or two! 16th of April 2008:
Blaise hit us hard with a great explanation of the resource control capabilities now in Solaris, and showed how they could be applied with and to Zones. Nathan and Andre's LDOM's talk helped us all to appreciate what it takes to use LDOMS, and both how hard it might be, and how easy it can be. :) 21st of May 2008:
Andre and Nathan hit the group with the news of the month, then Andre broke into an S9 containers talk. We covered all of the fun stuff, including showing the actual commands required to get you there. :) For the Maramba talk, we covered the content for the box itself, then, Lots of tech detail, including the breaking out of a compiler, assembly language and source code. Some folks may have immediately thought of running screaming from this uncommon excursion into the unknown of SPARC assembly, by the end, everyone was able to see the impact of compiling both 64 bit and using optimization. Good fun all round. 19th of June 2008:
16th of July 2008:
20th August 2008:
17th September 2008:
15th October 2008:
19th November 2008:
|