John Furrier43
Q1: What does "scale out" mean & where does the term apply?
jeff dinisco
the ability to scale performance linearly, and predict costs in the same manner
Chris Harney
Non-disruptive addition of additional nodes through the same management interface and the ability for the same data to grow without moving it to the new hardware
Chris Dwan
It usually describes a system where you add resources (I/O, CPU, Storage capacity) in a balanced way, which helps to stave off bottlenecks.
Andrew Miller
linear scalability of performance as add more resources without performance, administrative, or capacity bottlenecks
Patrick Rogers
Opposite of scale-up. It means linear horizontal growth in performance, capacity without logical limits.
Stuart Miniman
we used to scale-up (grow until you reached max allowed configuration), the challenge of our day is distributed architectures, which leverage scale-out deployments. smaller components that are highly scalable.
Dave Vellante
the ability to increase capacity and performance by adding resources horizontally in the form of nodes
Jeff Hughes
I think of it as the same term as horizontal scaling in the web world...adding units of scale seamlessly as the workload increases
Andrew Miller
in some ways it's more easily defined to people in "negative terms". That is, are you used to a dual controller system? Are you used to having to purchase more CPU/RAM capacity in a storage array upfront so can add disk later?
I am John White
scale out = anything that can incrementally grow and in turn shrink. Things like infrastructure, apps, and my waist line are all things that can scale out
Andrew Miller
The term applies in multiple areas - hadoop clusters, general storage applications, hyperconverged (Nutanix, etc.), even backup and recovery more recently.
jeff dinisco
@andriven agree, when explaining it, it almost always starts with an explanation of a head and sled architecture and then a comparison
Chris Dwan
@fdmts That said, there is no such thing as a system without limits. When we say "no limits," it just means we don't know them yet.
Kenneth Hui @rubrikInc HQ
These days, scale-out also implies the ability to add capacity and performance non-disruptively.
Andrew Miller
Also applies to cloud resources in that most cloud providers have written code to make all their internal resources scale-out. Customers don't see that but assume it's the case.
Patrick Rogers
It means never having to do a forklift upgrade ever again!
John Furrier
cloud is horizontally scalable therefore #devops software scaling out means APIs drive all the action in cloud
Andrew Miller
when talking "cloud" and diving in to the elasticity and flexibility characteristics, often there's some scale-out technologies under the covers to enable that
John Furrier
also #TruePrivateCloud for on prem is key to software models that operationally work across enterprises and clouds
Chris Harney
Purchase only what you need at this time and grow with your data / performance needs
I am John White
Anymore it seems like if we are talking scale out as opposed to up due to a change in application demand. Micro-services do not require big fat servers anymore
Stephen Pao
Good news is that we all seem consistent with the industry terms. Scale-out is horizontal scaling. Scale-up is vertical scaling.
John Furrier
applications that require data might want scale out and scale up operations
Andrew Miller
@dinisco +100 on cost - customers need predicable budget as grow. Scale-out technically isn't always paired with a similarly linear pricing model.
Stuart Miniman
@patrickrogersca isn't that assuming that you stick with a solution forever? Migration and API compatibility as they relate to scale-out environments.
Andrew Miller
Scale-out = removing technical barriers to growing your environment. Spend more money faster. ;)
Patrick Rogers
Cloud is not really horizontally scalable. It lacks a global namespace without logical limits.
John Furrier
@steve_pao great point on scale up for apps bc data is application centric but horizontal scale is how data moves
Andrew Miller
@johna_white the "shrinking" part is very interesting - it's often assumed but not always the case. Sometimes it's technically possible but not sufficiently tested/QA-ed for real world use.
John Furrier
@steve_pao great point on scale up for apps bc data is application centric but horizontal scale is how data moves
