lets edit it so we don't have an angry mob on our hands :-) On Sep 20, 2011, at 1:52 PM, Hunter Matthews wrote:
I put that on the slide just to have SOME kind of config listed for the VM's. I'm starting to regret that.
On Sep 20, 2011, at 1:42 PM, Narayan Desai wrote:
That is exactly what we are planning out right now. The main issue is that the existing system will be without an operations team come 10/1. We aren't going to turn off the current stuff, but we can't guarantee that it will be reliable either. There will definitely be a flag day in terms of data and so forth due to the way that the infrastructure is laid out behind the scenes.
There will also be some security things to work out as well. (was reminded by Gene stopping by my office)
What do you mean 1G images? -nld
On Sep 20, 2011, at 1:00 PM, Robert Olson wrote:
Having all that hardware unavailable for many months seems nonoptimal however; can't the machine be split and part upgraded/rebuilt and the rest left running, then upgraded after the first portion is rebuilt? Magellan in its current state works very well.
In any event, we cannot run the production services on a bunch of 1G images.
--bob
On Sep 20, 2011, at 12:52 PM, Narayan Desai wrote:
Yeah, we learned a lot from the initial deployment of openstack on Magellan. There are a bunch of new software features that we should be able to use with newer versions of openstack as well. -nld
On Sep 20, 2011, at 12:11 PM, Robert Olson wrote:
oh, it has to be remade first?
On Sep 20, 2011, at 12:07 PM, Rick Stevens wrote:
That is the eventual plan, once we remake magellan.
On Sep 20, 2011, at 11:55 AM, Robert Olson wrote:
How does ANL magellan fit into this? Seems it would be natural to host the VMs there and have much more resources available than just 1G/100G images.
--bob
On Sep 20, 2011, at 10:37 AM, Hunter Matthews wrote:
Attached is a draft of my deployment drawing in electronic form (omnigraffle) <kbase-hw-rollout.pdf> On Sep 20, 2011, at 10:10 AM, Brettin, Thomas S. wrote:
There are two things on the agenda for Thursday that I'm on the hook for.
1) hardware deployment strategy (Hunter and Tom) (20mins) 2) KB store (Bob and Tom) (30mins 4:30pm-5pm)
-- Hunter Matthews Unix Administrator Office: Bldg 221 Room B240 Argonne National Labs, MCS Key: F0F88438 / FFB5 34C0 B350 99A4 BB02 9779 A5DB 8B09 F0F8 8438 Never take candy from strangers. Especially on the internet.
_______________________________________________ Kbase-hw mailing list [email protected] https://lists.kbase.us/mailman/listinfo/kbase-hw
_______________________________________________ Kbase-hw mailing list [email protected] https://lists.kbase.us/mailman/listinfo/kbase-hw
Rick Stevens Associate Laboratory Director Computing, Environment and Life Science Argonne National Laboratory
Professor of Computer Science Senior Fellow Computation Institute Senior Fellow Institute for Genomics and Systems Biology The University of Chicago
[email protected] [email protected]
(v) 630-252-3378 (f) 630-252-6333
_______________________________________________ Kbase-hw mailing list [email protected] https://lists.kbase.us/mailman/listinfo/kbase-hw
_______________________________________________ Kbase-hw mailing list [email protected] https://lists.kbase.us/mailman/listinfo/kbase-hw
-- Hunter Matthews Unix Administrator Office: Bldg 221 Room B240 Argonne National Labs, MCS Key: F0F88438 / FFB5 34C0 B350 99A4 BB02 9779 A5DB 8B09 F0F8 8438 Never take candy from strangers. Especially on the internet.
_______________________________________________ Kbase-hw mailing list [email protected] https://lists.kbase.us/mailman/listinfo/kbase-hw
Rick Stevens Associate Laboratory Director Computing, Environment and Life Science Argonne National Laboratory Professor of Computer Science Senior Fellow Computation Institute Senior Fellow Institute for Genomics and Systems Biology The University of Chicago [email protected] [email protected] (v) 630-252-3378 (f) 630-252-6333