How should I practice Design Hit Counter?

Start by reading the full statement, solve the visible examples, then compare the pattern against related Nutanix onsite interview problems in FastPrep.

Is Design Hit Counter associated with Nutanix?

FastPrep tags this as a Nutanix onsite interview coding practice problem when the problem catalog has that company and stage context.

Design Hit Counter

Nutanix● MediumFULLTIMEONSITE INTERVIEW

Problem statement

Source note: The interviewer asked LeetCode 362, then three follow-ups: "If there are many hits at the same time, how do you optimize?", "What if I want you to optimize time and space?" and "In the multithreaded case, how do you keep it thread safe?" The author answered the first with a count per hit and the second with two fixed-length arrays, and said the third was talk only, with no code.

The report does not restate the problem or give bounds. Here, the window is the usual 300 seconds, so a query at time t counts the hits from t - 299 through t. Timestamps go up to 2 * 10^9 and never decrease. You get the calls as a list of up to 10000 operations instead of a class, and you return one number for each getHits. The tests check answers only, not your memory use or speed. The thread-safety follow-up is in the editorial and is not judged.

The judged task matches the report at about 90%.

Build a counter that records website hits and reports how many happened in the last five minutes. Process the calls in order. The two calls are:

"hit": one hit happens at second timestamps[i]. Several hits can happen in the same second.
"getHits": report how many hits so far happened in the last 300 seconds. For a call at time t, count the hits from time t - 299 through time t, both included.

A call only sees the calls before it, so a hit that comes later in the list does not count, even if it has the same timestamp.

Return an array with one number for each "getHits" call, in order. A "hit" call adds nothing to the array.

Function

hitCounter(operations: String[], timestamps: int[]) → int[]

Examples

Example 1

operations = ["hit","hit","hit","getHits","hit","getHits","getHits"]timestamps = [1,2,3,4,300,300,301]return = [3,4,3]

At time 4, the window reaches back before the first second, so it holds the hits at 1, 2 and 3, and the count is 3.
At time 300, the window is 1 through 300 and also holds the new hit at 300, so the count is 4.
At time 301, the window is 2 through 301. The hit at 1 has left, so the count is 3.

Example 2

operations = ["hit","hit","hit","getHits","getHits","getHits"]timestamps = [5,5,5,5,304,305]return = [3,3,0]

Three hits happen in second 5. At time 5 the count is 3.
At time 304, the window is 5 through 304, so all three still count.
At time 305, the window is 6 through 305. All three leave together, so the count is 0.

Constraints

1 <= operations.length <= 10000
timestamps.length == operations.length
Each operations[i] is "hit" or "getHits".
1 <= timestamps[i] <= 2 * 10^9
timestamps never decreases: timestamps[i] <= timestamps[i + 1].