Open Stack

hi,

i was doing some googling on the current state of time series databases 
to see if there was a worthwhile db to implement as a Gnocchi driver and 
it seems we're currently in the same state as before: there are flaws 
with all open source time series databases but if you have lots of 
money, Splunk is legit[1].

diving into each of the existing tsdb, there was a common theme that 
kept cropping up -- each of the tsdb claimed to do <insert all buzzwords 
here> really, really well. by association, as Gnocchi captures time 
series data, it's assumed Gnocchi should do <insert all buzzwords here> 
really, really well. this is obviously incorrect and setting Gnocchi up 
to fail.

as i get asked what Gnocchi is/does pretty often, i usually point them 
to docs[2]. that said, do we think it's necessary to define and list 
explicit use cases for Gnocchi? Gnocchi is 'a resource indexing and 
metric storage service' is a nice one-liner but would a comparison of 
Gnocchi make it easier to understand[3]. just wondering if there's a  
way we can make Gnocchi easier to understand and consumable.

[1] https://news.ycombinator.com/item?id=10530983
[2] http://docs.openstack.org/developer/gnocchi/
[3] http://prometheus.io/docs/introduction/comparison/

cheers,

-- 
gord