Run small Tor network as part of CI tests#1184
Conversation
5c8468d to
e6b75c4
Compare
Codecov Report
@@ Coverage Diff @@
## dev #1184 +/- ##
==========================================
+ Coverage 55.46% 55.50% +0.04%
==========================================
Files 140 140
Lines 20239 20267 +28
Branches 4895 4897 +2
==========================================
+ Hits 11225 11250 +25
- Misses 6007 6008 +1
- Partials 3007 3009 +2
Flags with carried forward coverage won't be shown. Click here to find out more.
Continue to review full report at Codecov.
|
f56450b to
306cd63
Compare
|
The Maybe we can just put the v3bw file in the 4uthority dir instead, and update this path? Or maybe we don't actually need the v3bw file after all? (I'm surprised that Tor does not fail since the above file no longer exists at the specified path.) In So removing the bridge dir seems ok. |
4e4da0d to
c151fa2
Compare
I added this file back.
Should I re-add the bridge and update the bridge line for torbridgeclient to use it? |
02d6d1b to
88f0e18
Compare
I don't think we need it because the authority already fills the bridge role. I don't think we need a bridge that does not also act as an authority, but if you are excited to get that working and want to add it, go ahead. |
I'm fine with leaving it out, just wanted to know if you had a preference. |
3c75b67 to
a6dec41
Compare
9360d5d to
31b2bc5
Compare
31b2bc5 to
afd3e14
Compare
|
@robgjansen Before merging, I'll disable the |
| add_shadow_tests(BASENAME tor-minimal | ||
| METHODS ptrace | ||
| LOGLEVEL info | ||
| ARGS | ||
| --pin-cpus | ||
| --workers 2 | ||
| --data-template "shadow.data.template" | ||
| PROPERTIES | ||
| TIMEOUT 180 | ||
| LABELS tor | ||
| CONFIGURATIONS tor | ||
| RUN_SERIAL TRUE) |
There was a problem hiding this comment.
I think we should check that Bootstrapped 100 appears 9 times in the Tor log files (once per Tor process) and stream-success 40 times in the tgen log files (10 downloads per tgen client) after the simulation finishes. Are we doing this somewhere that I am overlooking?
robgjansen
left a comment
There was a problem hiding this comment.
Thank you! Looks good (assuming the tests pass).
It turns out that setting `PASS_REGULAR_EXPRESSION` causes ctest to ignore the return value of the test, so now we add an explicit string in Shadow that signals a memory leak and check for it using `FAIL_REGULAR_EXPRESSION`.
b82bbd3 to
751ddf1
Compare
This PR adds a small tor experiment to the CMake tests that is not run by default. This test is run in the CI whenever the dev branch is updated, or if run manually using a
workflow_dispatchevent.