test: fix test-tls-ticket-cluster.js

[huseyinacacak-janea]

Problem

I've investigated the related issue in depth and found out that this test sometimes fails because of the handshake process. When the error happens and I look at the Wireshark logs, I see that the client doesn't send a Change Cipher Spec packet, but instead sends a FIN packet after a while. The FIN packet is sent in approximately 10ms. This may be some timeout, but I wasn't able to find proof for that.

I've added 2 screenshots from the Wireshark logs (which I can provide too if requested). The first one shows the expected packet flow when the test passes, and the second one shows the packet flow when the test fails.

Pass:

Failure:

As can be seen, the Change Cipher Spec packet was missing, and the FIN packet had already been sent before the handshake was complete.

I've seen discussions about the issues in the handshake process of Node.js in the following links: openssl/openssl#8069, openssl/openssl#7199, #18770.

As I don't have in-depth knowledge about OpenSSL, I'm not sure if this issue is caused by the way OpenSSL is used by Node. If there is a need to investigate it, a new issue can be opened.

Fix

In this PR, I’ve made the Primary (client) wait for the FIN packet from the Worker (server) to close the connection. This approach ensures that the Primary waits until it receives a FIN packet from the Worker and doesn’t time out during the handshake. I've run the test with the fix roughly 11k times and it hasn't failed once. Without the fix, I would usually get a failure with 1k runs easily.

Fixes: #2510

[lpinca]

Suggested change

	c.on('end', () => {
	c.end();
	});

Removing the whole 'secureConnect' listener should have the same effect.

[huseyinacacak-janea]

Thank you for your review. I applied your suggestion and tested it locally. I ran it over 4000 times and got 2 timeouts. The following is the test log:

E:\github\node>python tools\test.py test\parallel\test-tls-ticket-cluster
=== release test-tls-ticket-cluster ===
Path: parallel/test-tls-ticket-cluster
spawn file: E:\github\node\out\Release\node.exe
spawn file: E:\github\node\out\Release\node.exe
spawn file: E:\github\node\out\Release\node.exe
spawn file: E:\github\node\out\Release\node.exe
[primary] got "listening"
[primary] connecting 49193 session? false
[primary] got "listening"
[primary] got "listening"
[primary] got "listening"
[worker] connection reused? false
[primary] got "not-reused"
[primary] connecting 49193 session? true
[worker] connection reused? true
[primary] got "reused"
[primary] connecting 49193 session? true
[worker] connection reused? true
[primary] got "reused"
[primary] connecting 49193 session? true
[worker] connection reused? true
[primary] got "reused"
[primary] connecting 49193 session? true
[worker] connection reused? true
[primary] got "reused"
[primary] connecting 49193 session? true
[worker] connection reused? true
[primary] got "reused"
[primary] connecting 49193 session? true
[worker] connection reused? true
[primary] got "reused"
[primary] connecting 49193 session? true
[worker] connection reused? true
[primary] got "reused"
[primary] connecting 49193 session? true
[worker] connection reused? true
[primary] got "reused"
[primary] connecting 49193 session? true
[worker] connection reused? true
[primary] got "reused"
[primary] connecting 49193 session? true
[worker] connection reused? true
[primary] got "reused"
[primary] connecting 49193 session? true
[worker] connection reused? true
[primary] got "reused"
[primary] connecting 49193 session? true
[worker] connection reused? true
[primary] got "reused"
[primary] connecting 49193 session? true
[worker] connection reused? true
[primary] got "reused"
[primary] connecting 49193 session? true
[worker] connection reused? true
[primary] got "reused"
[primary] connecting 49193 session? true
[worker] connection reused? true
[primary] got "reused"
[worker] got "die"
[worker] server close
[worker] exit
[worker] got "die"
[worker] server close
[worker] exit
[worker] got "die"
[worker] server close
[worker] exit
[primary] worker died
[worker] got "die"
[worker] server close
[worker] exit
[primary] worker died
[primary] worker died
[primary] worker died
Command: E:\github\node\out\Release\node.exe E:\github\node\test\parallel\test-tls-ticket-cluster.js
--- TIMEOUT ---


[02:32|% 100|+   0|-   1]: Done

Failed tests:
E:\github\node\out\Release\node.exe E:\github\node\test\parallel\test-tls-ticket-cluster.js

Since the initial solution doesn't have that issue, I'd go with it to avoid timeouts.

[lpinca]

I think you can reproduce the timeouts also with the initial version then as the only difference is that c.end() is called in the next tick. See

node/lib/internal/streams/readable.js

Line 1699 in 9ef03f1

process.nextTick(endWritableNT, stream);

[lpinca]

I'm fine either way, as is or with the 'secureListener' event listener removed. The result is the same.

[huseyinacacak-janea]

I left the test running overnight and saw the error, which is related to the handshaking. It appears that removing secureConnect doesn't have the same effect as adding an end callback. Do you have any other suggestion I could try here?
P.S. I applied your other suggestion, the code looks better now.

[lpinca]

I left the test running overnight and saw the error, which is related to the handshaking. It appears that removing secureConnect doesn't have the same effect as adding an end callback. Do you have any other suggestion I could try here?

I have no suggestions. What is currently done in the 'end' callback, is already done automatically even if the callback is not added. The only difference is that c.end() is called in the next tick. This makes it explicit and I'm fine with it.

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/58238/

[lpinca]

# https://github.com/nodejs/node/pull/52563
$ python tools/test.py -J --repeat=1000 parallel/test-tls-ticket-cluster
=== release test-tls-ticket-cluster ===

...

node:assert:126
  throw new AssertionError(obj);
  ^

AssertionError [ERR_ASSERTION]: Expected values to be strictly equal:

15 !== 16

    at process.<anonymous> (C:\Users\lpinca\node\test\parallel\test-tls-ticket-cluster.js:92:12)
    at process.emit (node:events:532:35) {
  generatedMessage: true,
  code: 'ERR_ASSERTION',
  actual: 15,
  expected: 16,
  operator: 'strictEqual'
}

Node.js v22.0.0-pre
Command: C:\Users\lpinca\node\out\Release\node.exe C:\Users\lpinca\node\test\parallel\test-tls-ticket-cluster.js

...

[09:13|% 100|+ 986|-  14]: Done

# This rebased on top of https://github.com/nodejs/node/pull/52563
$ python tools/test.py -J --repeat=1000 parallel/test-tls-ticket-cluster
[08:43|% 100|+ 1000|-   0]: Done

All tests passed.

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/58459/

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/58465/

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/58467/

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/58468/

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/58469/

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/58475/

[nodejs-github-bot]

Landed in cfca9e3