‘WHAT’ we are doing remains the same but ‘HOW’ we are doing is different now

 
Sai: Um..
Me : See, we are reading books, watching movies but the way we are doing is different.

S: But the technical landscape behind that is different now.

M: Both Yes & NO. There also, same rule applies…’WHAT’ remains same but ‘HOW’ is different.

 S: ?
M: Take big data. How do you think, any data system should work.

S: It should read the input, store the input, process the input &
provide output.
M: Same is happening in BIG Data.

We call this Ingestion/ pulling in raw data( by Pig/Hive) , storage (in HDFS) , analysis ( by Spark/MapReduce), output presentation ( by
 Tableau/Plotly )
 
Only thing is that now each component is doing say only 1 job but
doing that job very effectively.
 
So, while earlier, all that work was getting done by a big program, now different components of that program are available & overall
combination of those components is the final output. 

Each component is really expert in doing its job. So, any type of data (word, image,) can be read or analyzed or presented...
 
S: You mean to say that we are only eliminating the pain point of earlier process? Nothing new?

M: No, if you can connect the right dots, you can offer better experience to customer.
 
For example, on social media, when you like any picture, that is an input data, which is read, stored analyzed & given to another program so that next time, more pictures like that are shown to you.

Just connection of right dots.
 
S: Don’t trivialize this.

So many people are using social media.
How that volume is getting handled. CLOUD surely is doing magic.
 
M: I remember, you used to write big programs!!
And whenever many customers were trying to use your program; we were almost boiling the ocean by adding more hardware power.

S: Um..Leg pulling but ok, carry on.

M: But now, we have broken up those parts of programs in logical units.

Whichever unit get more customer traffic, only that part of unit is scaled up.
For simplicity's sake, you can think of each independent logical s/w unit is running as a CONTAINER.

And Kubernetes is a container orchestration platform.
Based on the traffic load, it brings up/down more relevant containers.
So again, it automates the load balancing concept, which was getting used earlier.
 
S: Ok, but what is an API?
M: Good software programs, which are able to exchange data based

In 90’s when we were talking on banking through mobile, many of my friends were quite skeptical of that.
I mean, who will open a banking website, punch in all the data for payments.
 
S: Um
 M: But that time also, bar code readers were there.
Now, we have QR codes, which our mobile APP sees & does the payments.
And APIs does the necessary data exchange.
 
In the process of making ‘HOW’ part different, you need to identify logical units (DOTS) of a monolith program.
 
Now connect the right dots to create new experiences. 

#cloud #data #digital #expereince