How should I practice Concurrent Buffered File Logger?

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

Is Concurrent Buffered File Logger associated with Databricks?

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

Problem · Design

Concurrent Buffered File Logger

Learn this problem

● Medium

DatabricksFULLTIMEONSITE INTERVIEW

Problem statement

Model a lock-protected buffered file logger receiving writes from multiple logical threads. The input is the deterministic order in which operations acquire the logger lock, so process it from left to right.

["WRITE", threadId, message] appends threadId:message to the in-memory buffer.
["FLUSH"] appends every buffered line to the file in arrival order and empties the buffer.
["CLOSE"] flushes the remaining buffer and closes the logger.

When a WRITE makes the buffer size equal to bufferCapacity, flush automatically. Each line must reach the file exactly once. Return the complete file contents as an array of lines.

Function

runConcurrentFileLogger(bufferCapacity: int, operations: String[][]) → String[]

Examples

Example 1

bufferCapacity = 2operations = [["WRITE","t1","alpha"],["WRITE","t2","beta"],["WRITE","t1","gamma"],["FLUSH"],["WRITE","t3","delta"],["CLOSE"]]return = ["t1:alpha","t2:beta","t1:gamma","t3:delta"]

The first two writes trigger an automatic flush. FLUSH writes gamma, and CLOSE writes the remaining delta line.

Constraints

1 <= bufferCapacity <= 200000
1 <= operations.length <= 200000
The final operation is CLOSE, there is exactly one CLOSE, and no operation follows it.
Thread identifiers and messages are nonempty strings of at most 100 characters and contain neither a colon nor a newline.