what is the storage space required to host all images on google street view?
How much storage space do yous need for Google maps?
A Production Estimation Question for Product Management Interviews
by Malena Mesarina , Co-founder, StellarPeers
StellarPeers is a community platform that helps professionals prepare for interviews. We call up the best way to prepare, is to work through questions and practice mock interviews every bit much as possible. We meet weekly to talk over product management interview questions on product design, product launch, strategy, marketing, pricing, and others. Final week, we worked on a production estimation question from Lewis Lin'south book PM Interview Questions.
Question
How much storage space exercise you need to store all the information from Google maps?
Answer Structure
In interpretation questions, interviewers are evaluating your problem-solving and quantitative skills. They are looking for a ballpark number, not an authentic number. This production direction interview question tests whether yous are logical, can explicate all your assumptions clearly, are organized in your work, and good with numbers.
The volume Bully the PM Interview by Gayle Laakmann McDowell and Jackie Bavaro provides a skilful proffer on how to structure an reply for this blazon of question. It can exist summarized equally:
- Enquire clarifying questions. This will eliminate any ambivalence of what should be included in your calculations.
- Make an equation. Consider border cases or alternate sources of information and write down any facts that yous know will help with calculations.
- Break down the equation into components. Write downwards whatsoever assumptions adjacent to the components and then you don't forget to explicate them clearly to the interviewer.
- Do the math. Calculate the event of each component and compute the result.
- Practise a sanity cheque. Do your results make sense? If not, recheck your equation, assumptions and arithmetic.
Respond Example
INTERVIEWEE: Is this estimation for the US alone or the world?
INTERVIEWER: For the world.
INTERVIEWEE: Okay. Based on my feel with Google maps, information technology provides information about:
- Street names
- Homes
- Public buildings
- Businesses
- Traffic
- Non-urban areas, like mountainous terrains
INTERVIEWER: Okay, good, but why are yous differentiating betwixt homes, public buildings and businesses?
INTERVIEWEE: I have seen that Google maps displays more photos for buildings and business locations than for homes. Therefore, public buildings and businesses will require more storage than homes and I need to care for them separately.
INTERVIEWER: ok, agreed.
INTERVIEWEE: Here is how I would guess storage for each of these categories.
Street names storage:
[# cities] Ten [# streets/metropolis] X [storage for street name]
Homes storage:
[# cities] X [# homes/city] X [ane photo storage + text storage]
Public buildings storage:
[# cities] X [# public buildings/city] X [x X one photo storage + text storage]
Businesses storage:
[# cities] X [# businesses/city] X [10 X ane photo storage + text storage]
Traffic storage:
[# streets] 10 [storage/minute] X [# minutes/year]
Non-urban areas:
[surface of the globe in square miles] Ten [25%] X [storage/100 sq miles]
Total storage
= [Street names storage]
+ [Homes storage]
+ [Public buildings storage]
+ [Businesses storage]
+ [Traffic storage]
+ [Not-urban areas]
To judge the storage required for street names, homes, public buildings and businesses, I will differentiate between large and pocket-sized cities, because big cities have more of these types of locations than small cities.
For traffic storage estimation, I volition assume that traffic data is kept simply for a year.
Okay, allow me starting time with the large urban center calculations:
Big metropolis streets calculation
I will use New York metropolis equally a proxy for a large metropolis. Information technology has nearly 1000 streets, and assuming information technology takes 100 KB to store the street name and other metadata, like its distance, so information technology takes 100 MB for a big city to shop street information. I am not including photos of the street, because I will account for the street photos in the homes, public buildings and businesses storage calculations.
Large city homes calculation
For the big metropolis homes adding, New York urban center has about 9M people, and assuming at that place are 3 people per home, that gives 1M homes. And then assuming 4 MB for a photo of the street view and 100 KB for the home location information, and multiplying past 1M homes results in nigh 12 TB.
Big city public buildings calculation
Bold there are about 10,000 public buildings in a large city, multiplying that by (40 MB + 100 KB) which is storage for photos and text, results in 400 GB. I am bold that there are about 4x as many photos for buildings than for homes.
Large city businesses adding
Assuming there are about 20,000 businesses in a big urban center, multiplying that by (twoscore MB + 100 KB) for photo and text storage results in 800 GB.
Permit's move to small city calculations:
Small-scale metropolis streets calculation
I will assume that the number of a streets in a small city is most 1/iv of the number of streets in a large city, so i,000 / four = 250. Multiplying this number of streets by 100 KB to store the street metadata results in 25 MB.
Pocket-size city homes calculation
I volition assume that an average of 300,000 people live in a small city. Carve up by 3 persons per domicile, gives 100,000 homes. And multiplying that past (four MB + 100 KB) for photograph and location storage, results in 410 GB.
Pocket-sized city public buildings calculation
I will assume the number of public buildings is 1/10 of the number of buildings in a large city, so 10,000 / x = 1,000 buildings. Multiplying that past (forty MB + 100 KB) of storage for photos and text, results in 40 GB.
Small-scale city businesses calculation
I volition assume the number of businesses is 1/iv of the number in a large city, so 20,000 / 4 = v,000. Multiplying that by (40 MB + 100 KB) of storage for photos and text results in 200 GB.
With these results, I can now calculate the full storage for one big city every bit 100 MB + 12 TB + 400 GB + 800 GB = 13 TB. And, the total storage for a pocket-size urban center is 25 MB + 410 GB + xl GB + 200 GB = 650 GB. Now nosotros need to multiply these numbers past the number of big cities and pocket-size cities. I know there are virtually 1,000 big cities in the world and I am going to guess that in that location are about 4,000 small cities, then the full storage for city information in Google maps is: 1,000 Ten 13 TB + four,000 Ten 650 GB, which is almost 15 PB.
Let'south calculate traffic data stored in Google maps. I will assume that traffic information is only kept for one day. Nether this assumption, I will gauge storage by multiplying the number of streets in the world times storage needed to tape traffic per minute times minutes in a day. To summate the number of streets, I will first gauge the number of streets in big cities and then modest cities. I know in that location are virtually 1000 big cities in the world and I volition assume that a large urban center, like New York City, has 1000 streets, so 2,000 10 1,000 results in 2M streets for large cities. Assuming there are 4,000 small-scale cities in the world, and that each minor city has nearly 250 streets, then there are iv,000 Ten 250 = 1M streets for all modest cities. Then multiplying the number of streets by the storage per minute and minutes in a day gives (1M + 1M ) X 10 MB / infinitesimal X one,440 minutes / day, resulting in 28 PB. I am assuming here that it takes 10 MB per minute to record traffic photos per street.
Now let'south calculate non-urban areas. I will use the area of the world to do this. The surface area of the earth is near 4 x π x radius², where the radius of the globe is about 4000 miles. Bold that 25% of the earth surface is solid, and to store metadata takes 400 MB per 100 sq miles, the equation becomes 4 x 4 10 (four,000 miles)^two X ¼ X 400 MB / 100 sq miles = 256 TB.
Finally, adding storage for cities, traffic, and non-urban areas results in 15 PB + 28 Atomic number 82 + 256 TB = 43 PB. Google holds about 15 EB per yr, and then I think that 43 PB for Google Maps is not unreasonable.
StellarPeers
To learn more than virtually product management interview questions and to find practise partners, visit us at StellarPeers.com.
Source: https://medium.com/stellarpeers/how-much-storage-space-do-you-need-for-google-maps-24e83a0d5436
0 Response to "what is the storage space required to host all images on google street view?"
Post a Comment