Repository navigation
test: parallel/test-stringbytes-external is too slow #2370
Description
Activity
- addedtestIssues and PRs related to Node.js core tests and test infrastructure.Issues and PRs related to Node.js core tests and test infrastructure.
on Aug 13, 2015 - addedarmIssues and PRs related to the ARM architecture.Issues and PRs related to the ARM architecture.
on Aug 13, 2015 Major issue identified. All the string concatenation in the
assert.*()calls in theforloops is killing things. e.g. change// make sure Buffers from externals are the same for (var i = 0; i < c_bin.length; i++) { assert.equal(c_bin[i], c_ucs[i], c_bin[i] + ' == ' + c_ucs[i] + ' : index ' + i); }
to:
// make sure Buffers from externals are the same for (var i = 0; i < c_bin.length; i++) { if (c_bin[i] !== c_ucs[i]) assert.equal(c_bin[i], c_ucs[i]); }
a good chunk of time is lost. Same with the other 4
forloops below.@bnoordhuis Easy fix is to just remove all the string concatenation. Thoughts?
Seems reasonable to me.
Yea, I got it down quite a bit by changing those asserts
[~/dev/code/forks/io.js] :] ➜ $ time ./iojs test/parallel/test-stringbytes-external.js (master) 0.36 real 0.33 user 0.03 sysRef: #2410
Should be fixed in 37ee43e
going to reopen this, it's failing still/again, e.g. https://ci.nodejs.org/job/node-test-binary-arm/RUN_SUBSET=2,nodes=pi1-raspbian-wheezy/34
not ok 98 test-stringbytes-external.js --- duration_ms: 208.602 ...Hopefully #3287 fixes this.
I'm pretty sure this is fixed at this point. Let's try a stress test: https://ci.nodejs.org/job/node-stress-single-test/398/nodes=pi1-raspbian-wheezy/console
The buildbot went offline halfway through but 323 successful and 0 failing runs seem like good indication this issue is fixed.
It fails sporadically on the rpi1 buildbot: https://jenkins-iojs.nodesource.com/job/node-test-commit-arm/153/nodes=pi1-raspbian-wheezy/tapTestReport/test.tap-667/
/cc @trevnorris