René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"Investigate flaky test-http-client-timeout-option-with-agent","articleBody":"* **Version**: v11.0.0-pre(master)\r\n* **Platform**: ubuntu1604_sharedlibs_debug_x64\r\n* **Subsystem**: test\r\n\r\nhttps://ci.nodejs.org/job/node-test-commit-linux-containered/5882/nodes=ubuntu1604_sharedlibs_debug_x64/console\r\n\r\n```\r\n11:15:20 not ok 833 parallel/test-http-client-timeout-option-with-agent\r\n11:15:20   ---\r\n11:15:20   duration_ms: 17.334\r\n11:15:20   severity: fail\r\n11:15:20   exitcode: 1\r\n11:15:20   stack: |-\r\n11:15:20     assert.js:84\r\n11:15:20       throw new AssertionError(obj);\r\n11:15:20       ^\r\n11:15:20     \r\n11:15:20     AssertionError [ERR_ASSERTION]: Input A expected to strictly equal input B:\r\n11:15:20     + expected - actual\r\n11:15:20     \r\n11:15:20     - 4\r\n11:15:20     + 3\r\n11:15:20         at ClientRequest.req.on.common.mustCall (/home/iojs/build/workspace/node-test-commit-linux-containered/nodes/ubuntu1604_sharedlibs_debug_x64/test/parallel/test-http-client-timeout-option-with-agent.js:46:12)\r\n11:15:20         at ClientRequest.\u003canonymous\u003e (/home/iojs/build/workspace/node-test-commit-linux-containered/nodes/ubuntu1604_sharedlibs_debug_x64/test/common/index.js:467:15)\r\n11:15:20         at ClientRequest.emit (events.js:182:13)\r\n11:15:20         at Socket.emitRequestTimeout (_http_client.js:669:40)\r\n11:15:20         at Object.onceWrapper (events.js:273:13)\r\n11:15:20         at Socket.emit (events.js:182:13)\r\n11:15:20         at Socket._onTimeout (net.js:448:8)\r\n11:15:20         at ontimeout (timers.js:454:11)\r\n11:15:20         at tryOnTimeout (timers.js:326:5)\r\n11:15:20         at listOnTimeout (timers.js:300:5)\r\n11:15:20   ...\r\n```\r\n\r\nThis test asserts the elapsed time of `timeout` in second units. But the actual `timeout` may be longer in some situations.\r\n\r\nIt seems that other related tests avoid measuring the elapsed time. I would like to modify this test to make it less likely fail and keep the consistency of these tests.","author":{"url":"https://github.com/oyyd","@type":"Person","name":"oyyd"},"datePublished":"2018-07-31T09:08:04.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":3},"url":"https://github.com/22041/node/issues/22041"}