cumpiler

What are distributed systems in Job context?

Going through the job posting I've seen this sentence pop up in the requirements multiple times: "Have delivered, extended, and maintained large scale distributed systems." I'm having a hard time understanding what this means in a job context. Can anybody explain with like a simple example of delivering a distributed system that would make them eligible for this requirement? If I'm working on a webapp that uses tech like kaftka and zookeeper, can I consider myself as meeting this requirement?

11mo ago

Anon00

Walmart11mo

Assume a global website like Facebook which serve many nations-

At client - Do you know how to and what to cache at client for performance
At network - Do you know about CDN to serve content locally
At Api layer - Do you know horizontal scaling?. Do you know session stickiness?. Do you know how to route a request to nearest server?
At DB - Do you know CAP theorem. Do you know when to use No SQL vs SQL. How about sharding?

Of course these are very simplified above.

For a developer, you be expected to know atleast one of these in depth which you have worked on and an idea of atleast other 2.

For a lead role, I would expect to know 3 on 4 of these concepts.

For a architect, all of these indepth able to calculate traffic and server parameters also along with atleast one niche knowledge he might have implemented in his career.

Anon00

Walmart11mo

I probably missed to add event based systems in api backend that would include queueing, messaging, DLQ etc. You should be able to explain how to design async systems and it's advantages

Based1776

Remote11mo

Any books or courses to learn this stuff?

RealMichaelScott

Dunder Mifflin11mo

I guess large scale is a term to ponder.
Generally a db you would use in prod will be distributed but it doesn't mean that you have worked on large scale distributed systems until you have scaled the system to handle very large amount of data movement, made provisions to keep the system running even in inconsistent traffic, etc. Kafka is, well, distributed and if you have knowledge of partitioning, brokers, consumer groups, fault-tolerance, backpressure and ingested data from/in Kafka cluster in huge amount, yes you may have worked in large scale distributed systems. The main thing in a distributed system is how the nodes in the network coordinate to keep the system performing.

RealOldmonk

Stealth11mo

Sql dbs aren’t distributed btw

Discover more

Curated from across

Or get it on the stores.

Privacy Terms

Guidelines Help