Use directed acyclic graphs in gitlab CI

[RudolfWeeber]

No description provided.

[codecov]

Codecov Report

Merging #3109 into python will increase coverage by <1%.
The diff coverage is n/a.

@@           Coverage Diff            @@
##           python   #3109     +/-   ##
========================================
+ Coverage      83%     83%    +<1%     
========================================
  Files         530     513     -17     
  Lines       26141   24604   -1537     
========================================
- Hits        21954   20664   -1290     
+ Misses       4187    3940    -247

Impacted Files	Coverage Δ
...al_sites/lb_inertialess_tracers_cuda_interface.cpp	0% <0%> (-100%)	⬇️
...rialization/ibm_cuda_particle_velocities_input.hpp	0% <0%> (-100%)	⬇️
src/core/actor/ActorList.cpp	0% <0%> (-80%)	⬇️
src/core/EspressoSystemInterface.cpp	40% <0%> (-52%)	⬇️
...tatics_magnetostatics/magnetic_non_p3m_methods.cpp	0% <0%> (-37%)	⬇️
...rc/core/grid_based_algorithms/lb_interpolation.cpp	78% <0%> (-12%)	⬇️
src/script_interface/shapes/Slitpore.hpp	80% <0%> (-11%)	⬇️
src/script_interface/shapes/HollowCone.hpp	75% <0%> (-9%)	⬇️
src/core/electrostatics_magnetostatics/dipole.cpp	58% <0%> (-6%)	⬇️
src/core/forces.cpp	96% <0%> (-4%)	⬇️
... and 168 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update e2e7c5d...553b01b. Read the comment docs.

[jngrad]

For reference: DAG in GitLab-CI, short explanation, documentation

[RudolfWeeber]

As far as I can tell from yesterday's build, itś working.

[jngrad]

This could also help in our docker repository, where stage N only starts after the slowest job at stage N-1 is complete. This would give us faster feedback during a PR and faster deployment after a merge. The only drawback I can think of is the possibility of a race condition during CI of a merge: a job at stage 1 might not start until a job at stage 3 is complete that requires a runner with the same tags, causing some images to be deployed later than expected. This concern is however irrelevant to the espresso repository.

[jngrad]

bors r+

[bors]

Build succeeded

• ICP GitLab CI