wip-crush-linear

2015-01-29 Thread Sage Weil
Hi Yujie, I've rebased your congruential crush series on top of the other recent CRUSH changes (mainly straw2 buckets) and made a few small changes and cleanups. Can you please take a look and see if I may have missed anything? It would also be helpful if you could verify that the final

wip-crush

2012-11-26 Thread Sage Weil
Hi Caleb- Do you mind looking at wip-crush today? First we should demonstrate with the crushtool tests that the change is in fact a Good Thing. Setting lots of osds out inside a host and then doing chooseleaf over host should do it. Then the compatibility bit hijinx need to be reviewed

Re: wip-crush

2012-09-05 Thread Sage Weil
On Fri, 31 Aug 2012, Sage Weil wrote: Okay, it's trivial to change 'pool' to 'root' in the default generated crush map, and update all the docs accordingly. The problem is that some stuff built on top of ceph has 'pool=default' in there, including our chef cookbooks and those that other

Re: wip-crush

2012-08-31 Thread Sage Weil
Okay, it's trivial to change 'pool' to 'root' in the default generated crush map, and update all the docs accordingly. The problem is that some stuff built on top of ceph has 'pool=default' in there, including our chef cookbooks and those that other people are building. I see a few options:

Re: wip-crush

2012-08-22 Thread Florian Haas
On 08/22/2012 03:10 AM, Sage Weil wrote: I pushed a branch that changes some of the crush terminology. Instead of having a crush type called pool that requires you to say things like pool=default in the ceph osd crush set ... command, it uses root instead. That hopefully reinforces that

Re: wip-crush

2012-08-22 Thread Gregory Farnum
On Wed, Aug 22, 2012 at 9:33 AM, Sage Weil s...@inktank.com wrote: On Wed, 22 Aug 2012, Atchley, Scott wrote: On Aug 22, 2012, at 10:46 AM, Florian Haas wrote: On 08/22/2012 03:10 AM, Sage Weil wrote: I pushed a branch that changes some of the crush terminology. Instead of having a crush