jwkvam / datastructs-algorithms-interview-problems

On-site interview problem for Full Stack position

Query Batching

At Verdigris, we have database that stores data about our customers' electrical circuits.

Problem

Included in the repository is a Python API for querying the data stored on our database. Currently, the client is written to perform each query individually. Since this data query must be made over a network, the amount of time to establish a connection and perform the query is the bottleneck.

Write a batched version of the client such that we can minimize the network overhead.

To query for data, run:

python batched_queries.py

IMPORTANT: You are forbidden from just dispatching requests asynchronously without batching first.

Output

It is very important that the batched query generates the output in the EXACT same order as the non-batched version.

If the input looks like this:

queries = [
    DataQuery("Boutique Panel A15", 150),
    DataQuery("Lab Cooling Towers", 500.25),
    DataQuery("Verdigris HQ Elevator", 157),
    DataQuery("Lab Cooling Towers", 460.25),
    DataQuery("Boutique Panel A15", 140),
    DataQuery("Verdigris HQ Elevator", 7),
    DataQuery("Verdigris HQ Elevator", 1003.5),
]

Running non-batched method produces the following result:

>>> run_all_queries(queries)
[CIRCUIT] Boutique Panel A15: 0.7 Watts @ t=150
[CIRCUIT] Lab Cooling Towers: 2.7 Watts @ t=500.25
[CIRCUIT] Verdigris HQ Elevator: 1.6 Watts @ t=157
[CIRCUIT] Lab Cooling Towers: 3.7 Watts @ t=460.25
[CIRCUIT] Boutique Panel A15: 2.2 Watts @ t=140
[CIRCUIT] Verdigris HQ Elevator: 1.6 Watts @ t=7
[CIRCUIT] Verdigris HQ Elevator: 1.3 Watts @ t=1003.5

Running the batched version should also produce the same result:

>>> run_all_queries_batched(queries)
[CIRCUIT] Boutique Panel A15: 0.7 Watts @ t=150
[CIRCUIT] Lab Cooling Towers: 2.7 Watts @ t=500.25
[CIRCUIT] Verdigris HQ Elevator: 1.6 Watts @ t=157
[CIRCUIT] Lab Cooling Towers: 3.7 Watts @ t=460.25
[CIRCUIT] Boutique Panel A15: 2.2 Watts @ t=140
[CIRCUIT] Verdigris HQ Elevator: 1.6 Watts @ t=7
[CIRCUIT] Verdigris HQ Elevator: 1.3 Watts @ t=1003.5