Undici failures

[targos]

With nodejs/node#52002:

https://ci.nodejs.org/view/Node.js-citgm/job/citgm-smoker/3398/nodes=alpine-latest-x64/testReport/junit/(root)/citgm/undici_v6_7_0/
https://ci.nodejs.org/view/Node.js-citgm/job/citgm-smoker/3398/nodes=rhel8-x64/testReport/junit/(root)/citgm/undici_v6_7_0/

@nodejs/undici

[Uzlopak]

Connected with nodejs/node#51997

[targos]

Is it? The run was done after the revert.

[Uzlopak]

Does the citgm connect via debug port?

[targos]

Not sure I understand the question but probably not.
I basically does npm install followed by npm test

[KhafraDev]

 test at test/node-test/debug.js:54:1
  debug#fetch (346.571414ms)
   AssertionError [ERR_ASSERTION]: Expected values to be strictly equal:
   
   4 !== 6
   
       at res.<computed> [as strictEqual] (/home/iojs/tmp/citgm_tmp/19886eeb-1a0b-44d6-951c-284ce1a45418/undici/node_modules/@matteo.collina/tspl/tspl.js:52:35)
       at Socket.<anonymous> (/home/iojs/tmp/citgm_tmp/19886eeb-1a0b-44d6-951c-284ce1a45418/undici/test/node-test/debug.js:79:12)
       at Socket.emit (node:events:531:35)
       at endReadableNT (node:internal/streams/readable:1696:12)
       at process.processTicksAndRejections (node:internal/process/task_queues:82:21) {
     generatedMessage: true,
     code: 'ERR_ASSERTION',
     actual: 4,
     expected: 6,
     operator: 'strictEqual'
   }
 test at test/node-test/debug.js:89:1
  debug#undici (186.572836ms)
   AssertionError [ERR_ASSERTION]: Expected values to be strictly equal:
   
   21 !== 6
   
       at res.<computed> [as strictEqual] (/home/iojs/tmp/citgm_tmp/19886eeb-1a0b-44d6-951c-284ce1a45418/undici/node_modules/@matteo.collina/tspl/tspl.js:52:35)
       at Socket.<anonymous> (/home/iojs/tmp/citgm_tmp/19886eeb-1a0b-44d6-951c-284ce1a45418/undici/test/node-test/debug.js:117:12)
       at Socket.emit (node:events:531:35)
       at endReadableNT (node:internal/streams/readable:1696:12)
       at process.processTicksAndRejections (node:internal/process/task_queues:82:21) {
     generatedMessage: true,
     code: 'ERR_ASSERTION',
     actual: 21,
     expected: 6,
     operator: 'strictEqual'
   }
 test at test/node-test/autoselectfamily.js:1:1
  test/node-test/autoselectfamily.js (30000.124732ms)
   'test timed out after 30000ms'

related to nodejs/undici#2600

[targos]

Ok, so we have to wait for the next libuv update.

[KhafraDev]

it looked like a different error to me, this test has had quite a few failures cc @metcoder95

[Uzlopak]

yeah, it is an annoying file

[metcoder95]

Ok... that's new, the mismatch on the assertions is weird; I'll try to revisit this and possibly rewrite the testing scenarios, haven't had the time yet.

If this is a blocker of some sort, feel free to skip them for now

[Uzlopak]

Should now work as we patched the failing test.

[targos]

https://ci.nodejs.org/view/Node.js-citgm/job/citgm-smoker-pipeline/240/

[targos]

On OSX:

09:22:16 error:                     | ✖ failing tests:                                                                          
09:22:16 error:                     |                                                                                           
09:22:16 error:                     | test at test/issue-803.js:1:1                                                             
09:22:16 error:                     | ✖ test/issue-803.js (30147.778213ms)                                                      
09:22:16 error:                     | 'test timed out after 30000ms'                                                            
09:22:16 error:                     |                               

[Uzlopak]

create a PR to add macos-latest to our ci

nodejs/undici#2952

[mcollina]

https://ci.nodejs.org/view/Node.js-citgm/job/citgm-smoker-pipeline/241/

[KhafraDev]

 test at test/issue-803.js:1:1
  test/issue-803.js (30101.051783ms)
   'test timed out after 30000ms'

@Uzlopak

[Uzlopak]

What should I do? The test has afaik already a timeout of 60000ms. So failing after 30000 ms, makes no sense.

[metcoder95]

Maybe refactor to AsyncIterable to see if it’s more deterministic?
Just throwing ideas

[mcollina]

Just skip the test if citgm.

[Uzlopak]

Is there a env variable set in case of citgm?

[targos]

I don't think you can specifically detect citgm by default, but you can add custom env variables using the envVar key in the lookup.

Example:

citgm/lib/lookup.json

Lines 245 to 252 in 12e6902

	"jose": {
	"maintainers": ["panva"],
	"prefix": "v",
	"scripts": ["test", "tap:node"],
	"envVar": { "CITGM": true },
	"skip": ["aix"],
	"comment": "esbuild doesn't support aix"
	},

[mcollina]

I've released v6.19.3 that should address some of this