M001:Chapter 4 Lab issues/sample dataset reimport

Got stuck on Chapter 4 Lab 1. It asked
How many zips in the sample_training.zips dataset are neither over-populated nor under-populated?
In this case, we consider population of more than 1,000,000 to be over- populated and less than 5,000 to be under-populated.
I ran:
the query returned 18277, which the lab marked incorrect. On the chance that the dataset has been changed during earlier exercises, I dropped sample db’s and attempted to reimport sample data. However, I got an error: " An error occurred loading sample data: Target cluster does not have enough free space to import dataset". Please assist with reloading the sample dataset, so I could complete the exercises.

It means you exceeded 512mb quota limit

Please delete unnecessary DBs and try again
Did you import sample_data twice?
Check this link

Over Space Quota while importing restaurants.json

I have deleted all the db’s that were imported through Atlas…

MongoDB Enterprise atlas-tvcjyo-shard-0:PRIMARY> show dbs
admin 0.000GB
local 3.987GB
MongoDB Enterprise atlas-tvcjyo-shard-0:PRIMARY>

Is there some sort of recycle bin that I’d have to empty?

With a standard atlas admin you don’t have permissions to drop collections in the local database. You can create a power user probably, or just use Atlas UI to remove all data (easiest).

Now, if you run locally, you could remove oplog.rs which is the source of 99% of the storage you see.

Example from my cluster

sandbox-shard-00-01(mongod-4.4.3)[PRIMARY:atlas-n7cidd-shard-0] local>
show collections
clustermanager                  →    0.000MB /    0.035MB
oplog.rs                        → 7860.029MB / 5625.266MB
replset.election                →    0.000MB /    0.035MB
replset.initialSyncId           →    0.000MB /    0.031MB
replset.minvalid                →    0.000MB /    0.035MB
replset.oplogTruncateAfterPoint →    0.000MB /    0.035MB
startup_log                     →    0.148MB /    0.070MB

The nice output is generated by MongoDB hacker, a plugin.