Build Farmer Handoff 2018-05-23

[mjcarroll]

Incoming: @dirk-thomas
Outgoing: @mjcarroll

Expected failures:

• CMake Warnings from Cmake warning about unused variable SECURITY #79 have returned but are not ubiquitous pass --no-warn-unused-cli to CMake to avoid unwanted warnings ci#166 is a proposed solution.

New flakiness:

• rviz_common.selection_manager_test missing results on osx_nightly_repeated: Flaky rviz_common.selection_manager_test rviz#271 (seems to only affect macos)
• composition test flakiness increase (to be investigated) Composition test flakiness increase #118

Long-running flakiness:

• service/parameter tests with fastrtps (common, shows as test failures or timeouts)
• service/parameter tests with connext (less common, shows as a timeout) wait_for_service not being woken by graph events rmw_connext#280
• muiltithreaded tests hanging with connext Flaky: multithreaded comms tests hanging with connext on osx #103
• pendulum_teleop hanging (only runs on connext, linux) Pendulum teleop timeout [connext, linux] #94
  • should be less flaky after Pendulum teleop flakiness reduction demos#232
• hanging tests with fastrtps Various tests hanging sometimes on linux with fastrtps #98
  • TestFindWeakNodes, test_time_source, test_spin
• DomainParticipantFactory_impl::create_participant():!create failure creating participant failures on connext (seems can happen in any test): Connext: Automatic participant index failed to initialize #93
  • We should keep investigating/pushing on them for what is causing this now
• test_utilities access violation: Access violation in new test_utilities test rclcpp#458
• "performance" tests (action pending discussion in Test of node->now() flaky on OS X rclcpp#457)
  • e.g. rclcpp.TestNode.now, timer/waitset testing
• projectroot.test_dynamic_bridge__rmw_fastrtps_cpp / ros1_bridge.test_dynamic_bridge__rmw_fastrtps_cpp.xunit.missing_result (flaky due to fastrtps discovery issues)

Build farmer tasks

• Update Java to 1.8u162 on all hosts (including master): @nuclearsandwich
• Renew https certificate on ci.ros2.org: @nuclearsandwich
• Periodic upgrades on Mac and Windows: Build farmer
• Profile memory usage of bridge compilation job building the bridge requires a lot of memory #106 (failures currently resolved by increasing instance size)

[mjcarroll]

Rolling to you @dirk-thomas since @clalancette is busy with Melodic release and @mikaelarguedas is at a timezone disadvantage.

[dirk-thomas]

Expected failures:
CMake Warnings from #79 have returned but are not ubiquitous ros2/ci#166 is a proposed solution.

The patch has already been merged 8 days ago.

[mjcarroll]

Sorry, copy-paste error.
In that case, I would not expect it to fail.

[mikaelarguedas]

I saw one fail yesterday https://ci.ros2.org/view/nightly/job/nightly_linux_debug/841/ and opened ros2/rviz#273 to work around it

[dhood]

Timeouts are no longer expected for two-process client-service tests (e.g. test_client_scope_cpp, test_client_scope_consistency_cpp) with connext after ros2/rclcpp#476. Please point out if they show up again.

However, some other timeouts still happen e.g. test_tutorial_parameter_events__rmw_connext_cpp (see #118 (comment)) or test_remote_parameters[_async]

[mikaelarguedas]

MacOS nodes should be updated to install poco from brew and test that this allows to build workspace and overlays without needing to build poco from source.

• brew install poco
• run a test ci job
• run a test packaging job
• build the examples repo on top of the packaging artifact
• run some examples from the overlay
• propagate to the other Mac machines
• update node setup instructions and wiki installation instructions accordingly

[dirk-thomas]

What else happened in the paste two weeks:

• The Connext Debian package was imported to our apt repos.
• A new parameter to use that Debian package was added to the job (readd option to use Debian pkg of Connext ci#181).
• The new Visual Studio 2017 binaries of OpenSplice were installed on all three Windows executors (example build: https://ci.ros2.org/job/ci_windows/4642/).
• The packet machine went offline and only a restart of the master made it reconnect.