Ractor for true parallelism in Ruby - Step-by-Step Execution

Concept Flow - Ractor for true parallelism

Create Ractor

→Ractor runs code in parallel

↓

Main thread continues

↓

Send message to Ractor

↓

Ractor processes message

↓

Ractor sends result back

↓

Main thread receives result

↓

Ractor ends or waits for more messages

Ractor creates a parallel thread that runs code independently, communicates via messages, and allows true parallelism.

Execution Sample

Ruby

r = Ractor.new do
  msg = Ractor.receive
  msg * 2
end

r.send(10)
result = r.take

This code creates a Ractor that doubles a received number and sends back the result.

Execution Table

Step	Action	Ractor State	Main Thread State	Message Passing	Output
1	Create Ractor	Waiting for message	Running main code	No message yet	None
2	Main thread sends 10	Received message 10	Waiting for result	Message 10 sent to Ractor	None
3	Ractor processes message	Computing 10 * 2 = 20	Waiting for result	Processing message	None
4	Ractor sends result	Result 20 ready	Waiting to receive	Result 20 sent to main thread	None
5	Main thread receives result	Idle or waiting	Received 20	Result 20 received	20
6	Program ends or Ractor waits	Idle or terminated	Program ends	No messages	20

💡 Execution stops after main thread receives the doubled result and program ends.

Variable Tracker

Variable	Start	After Step 2	After Step 3	After Step 5	Final
r	Ractor object created	Ractor received 10	Ractor computed 20	Ractor idle	Ractor idle or terminated
msg	None	10	10	None	None
result	None	None	None	20	20

Key Moments - 3 Insights

Why does the main thread not wait immediately after creating the Ractor?

Because the Ractor runs in parallel, the main thread continues until it explicitly waits for a result (see step 2 and 5 in execution_table).

How does the main thread get the result from the Ractor?

The main thread uses r.take to receive the message sent back by the Ractor (step 5 in execution_table).

What happens if the Ractor does not receive a message?

The Ractor waits (blocks) at Ractor.receive until a message arrives (step 1 in execution_table).

Visual Quiz - 3 Questions

Look at the execution_table, what is the Ractor state right after the main thread sends 10?

A Waiting for message

B Computing 10 * 2 = 20

C Received message 10

D Idle or terminated

💡 Hint

Check step 2 in the execution_table under 'Ractor State'.

At which step does the main thread receive the result from the Ractor?

A Step 5

B Step 4

C Step 3

D Step 6

💡 Hint

Look for 'Main thread receives result' in the execution_table.

If the main thread sends 20 instead of 10, what would change in the execution_table?

A The Ractor would not receive any message

B The Ractor state at step 3 would show computing 20 * 2 = 40

C The main thread would receive 10 at step 5

D The program would end earlier

💡 Hint

Check how the message value affects the computation in step 3.

Concept Snapshot

Ractor.new { code } creates a parallel Ractor.
Use r.send(msg) to send messages.
Use Ractor.receive inside Ractor to get messages.
Use r.take to get messages from Ractor.
Ractors run truly in parallel, no shared memory.
Communication only via messages.

Full Transcript

This example shows how Ruby's Ractor enables true parallelism by running code in a separate thread. The main thread creates a Ractor that waits for a message, doubles it, and sends back the result. The main thread sends the number 10, the Ractor receives it, computes 20, and sends it back. The main thread then receives 20 and continues. Variables like 'r', 'msg', and 'result' change states as messages pass. Key points include that the main thread does not block on Ractor creation, communication is via messages, and Ractor waits for messages if none are sent. The quiz tests understanding of Ractor states and message passing.