Thanks John,
I was leaning towards '2 is not quite enough' for parity, but wanted to
get a 2nd opinion. The level of detail and discussion in your answer is
very helpful, much appreciated!
Mark
On 05/04/18 08:25, John Dickinson wrote:
The answer always starts with "it depends...". Depends
The answer always starts with "it depends...". Depends on your hardware, where
it's physically located, the durability you need, the access patterns, etc
There have been whole phd dissertations on the right way to calculate
durability. Two parity segments isn't exactly equivalent to three replic
Hi Andrew!
AFAIK there's no difference between Liberty and Mitaka nova endpoints
(we're using newton and $(tenant_id)s is still present in the nova
endpoints and it's working properly). It seems to me that the
difference you're showing was introduced in ocata, you can compare the
official doc inst
...hearing crickets - come on guys, I know you have some thoughts about
this :-) !
On 29/03/18 13:08, Mark Kirkwood wrote:
Hi,
We are looking at implementing EC Policies with similar durability to 3x
replication. Now naively this corresponds to m=2 (using notation from
previous thread). Howev
Hi,
AFAIU the get_attr function does not use the values you've passed to Heat
in the resource definition, instead it fetches their actual values from
Neutron (basically making a 'port show' API call), and Heat does nothing
wrt to ordering afterwards.
Btw AFAIR this is exactly why heat requires a
