You are on page 1of 4

https://www.linkedin.com/in/syedashrafcloud/ YouTube: https://youtu.

be/__SBjmHEBfY

Difference between horizontally scaling


and vertically scaling
Particular Horizontal Scaling Vertical Scaling
Concept involves adding more processing units refers to adding more

or physical machines to your server or resources

database. It involves growing the (CPU/RAM/DISK) to

number of nodes in the cluster, existing machine

Reasons to scale vertically includes increasing In order to reducing the

IOPS (Input / Ouput Operations), responsibilities of each

amplifying CPU/RAM capacity, as member node and providing

well as disk capacity additional end-points for

client connections. much

more used for high level of

computing and for

application and services

Databases In a database world, horizontal scaling In vertical scaling, the data

is usually based on the partitioning of lives on a single node and

Web site: https://cloudstable.blogspot.com/


https://www.linkedin.com/in/syedashrafcloud/ YouTube: https://youtu.be/__SBjmHEBfY

data (each node only contains part of scaling is done through

the data). multi-core, e.g. spreading

the load between the CPU

and RAM resources of the

machine.

Downtime In theory, adding more machines to the Vertical scaling is limited to

existing pool means you are not limited the capacity of one

to the capacity of a single unit, making machine, scaling beyond

it possible to scale with less downtime. that capacity can involve

downtime and has an upper

hard limit, i.e. the scale of

the hardware on which you

are currently running.

Concurrency Also described as distributed Actor model: concurrent

programming, as it involves programming on multi-core

distributing jobs across machines over machines is often

the network. Several patterns performed via multi-

Web site: https://cloudstable.blogspot.com/


https://www.linkedin.com/in/syedashrafcloud/ YouTube: https://youtu.be/__SBjmHEBfY

associated with this model: threading and in-process

Master/Worker*, Tuple Spaces, message passing.

Blackboard, MapReduce.

Message In distributed computing, the lack of a In a multi-threaded


passing
shared address space makes data scenario, you can assume

sharing more complex. It also makes the existence of a shared

the process of sharing, passing or address space, so data

updating data more costly since you sharing and message

have to pass copies of the data. passing can be done by

passing a reference.

Examples Cassandra, MongoDB, Google Cloud MySQL, Amazon RDS

Spanner

Requirement horizontal scaling use whenever a high Vertical scaling requires

availability of (server) services are for adding more resources

required (CPU/RAM/DISK) to your

server (database or

application server is still

Web site: https://cloudstable.blogspot.com/


https://www.linkedin.com/in/syedashrafcloud/ YouTube: https://youtu.be/__SBjmHEBfY

remains one) as on demand

Web site: https://cloudstable.blogspot.com/

You might also like