Monthly issue metrics report

[julesghub]

Issue Metrics

Metric	Average	Median	90th percentile
Time to first response	4 days, 6:48:34	4 days, 4:16:18	8 days, 3:16:44
Time to first review	None	None	None
Time to close	5 days, 15:29:29	2 days, 16:21:29	13 days, 11:21:38
Time to answer	None	None	None

Metric	Count
Number of items that remain open	50
Number of items closed	36
Total number of items created	86

Title	URL	Assignee	Author	Time to first response	Time to first review	Time to close	Time to answer
test_0053_hang_watchdog crashes its xdist worker intermittently (1 in ~15 CI runs)	underworldcode/underworld3#661	None	lmoresi	None	None	7 days, 11:53:44	None
SUPG and Levelset	underworldcode/underworld3#657	None	NengLu	6 days, 18:33:38	None	None	None
Stokes_Constrained: consolidated state of the open work (supersedes #244, #248, #254, #495)	underworldcode/underworld3#651	None	lmoresi	None	None	None	None
geoid harmonic projection re-derives the surface mesh with ConvexHull instead of integrating over the boundary facets we already have	underworldcode/underworld3#647	None	lmoresi	None	None	None	None
Monthly issue metrics report	#767	julesghub	julesghub	None	None	None	None
Document long HDF5 path failures in parallel PETSc output	underworldcode/underworld3#645	None	gthyagi	None	None	None	None
FMG geometric-MG coarse solve uses PCREDUNDANT: coarse LU replicated on every rank — OOM at scale when the base mesh is fine	underworldcode/underworld3#644	None	bknight1	1 day, 13:58:59	None	1 day, 13:58:59	None
FMG + redundant coarse solver: replicated memory scales with base mesh size, causing OOM at HPC scale	underworldcode/underworld3#643	None	bknight1	None	None	0:41:16	None
Problem with setting VelocityBC using LecodeIsostasy	#766	None	jaime-almeida	8 days, 17:18:04	None	0:08:17	None
uw.Params silently ignores negative CLI values (-uw_x -1 falls back to the default)	underworldcode/underworld3#642	None	lmoresi	None	None	None	None
evaluate() returns large negative values for a non-negative expression (sqrt invariant) at in-domain points near a boundary singularity	underworldcode/underworld3#641	None	lmoresi	None	None	None	None
Checkpoint round-trip corrupts fault-adjacent DOFs on split meshes	underworldcode/underworld3#640	None	lmoresi	None	None	None	None
test_0054_hang_report leaks a spinning child process whenever pytest is killed (7 orphans, 7 cores)	underworldcode/underworld3#639	None	lmoresi	None	None	None	None
3-D boundary-flux recovery supports only P1/P2 triangular traces, so dynamic topography cannot be cross-validated	underworldcode/underworld3#637	None	lmoresi	None	None	None	None
Document Gadi MPI rank-placement performance for MatMult-dominated Stokes solves	underworldcode/underworld3#635	None	gthyagi	None	None	None	None
Grad-div penalty breaks POINTWISE traction recovery (28% on dynamic topography); facet-integrated is fine	underworldcode/underworld3#633	None	lmoresi	None	None	None	None
Geometric aggregation for placed-hierarchy FMG: coarse levels as decimated point clouds, not band-only meshes	underworldcode/underworld3#629	None	lmoresi	None	None	None	None
adapt: snap_level_boundaries guards its allreduce on a rank-local mask (np>1 hang)	underworldcode/underworld3#627	None	lmoresi	None	None	1 day, 2:09:15	None
3-D cut-in-band: author the slit into the thin-volume assembly (the carve-based mid-plane cut cannot work)	underworldcode/underworld3#626	None	lmoresi	None	None	None	None
Stokes pressure sub-solve runs at its 200-iteration cap — the inner Schur solve is silently unconverged and dominates solve time	underworldcode/underworld3#625	None	lmoresi	None	None	None	None
Stokes outer Krylov is PETSc-default gmres around fgmres inner solves — the flexible-outer rule applied to the sub-blocks but the outer was never set	underworldcode/underworld3#624	None	lmoresi	None	None	None	None
place_sheet fill drifts the domain volume on steeply graded edge_split children	underworldcode/underworld3#623	None	lmoresi	None	None	None	None
Custom-P V-cycles across a cut level cost more than they save (7x slower than GAMG at 11x fewer iterations)	underworldcode/underworld3#622	None	lmoresi	None	None	None	None
Fault-network place route: h_far builds an 8x-denser far field than embed	underworldcode/underworld3#621	None	lmoresi	None	None	None	None
cells_supporting returns an empty fault zone on a 3-D mesh	underworldcode/underworld3#620	None	lmoresi	None	None	None	None
StructuredQuadBox mesh cache filename omits the element resolution: concurrent runs at different resolutions collide	underworldcode/underworld3#618	None	lmoresi	None	None	13:42:31	None
boundary_flux() returns ~1e12 on Stokes_Constrained	underworldcode/underworld3#614	None	lmoresi	None	None	None	None
place_surface: _sheet_distance's per-triangle loop is the placement bottleneck once sheets are resampled fine	underworldcode/underworld3#613	None	lmoresi	None	None	None	None
test_global_evaluate_after_migration hangs at np=4, and its file is in neither parallel-test glob	underworldcode/underworld3#611	None	lmoresi	None	None	None	None
adapt-on-top: ~90% of wall clock is repeated whole-mesh Python bookkeeping, not the transform	underworldcode/underworld3#610	None	lmoresi	None	None	None	None
Rotated free slip meets a component Dirichlet wall: a local pressure error at the shared corner	underworldcode/underworld3#608	None	lmoresi	None	None	None	None
Stokes_Constrained: multiplier() is not the whole boundary traction — the augmented-Lagrangian share is missing	underworldcode/underworld3#607	None	lmoresi	None	None	None	None
evaluate() silently returns wrong values in parallel for points the rank does not own — including mesh nodes	underworldcode/underworld3#606	None	lmoresi	None	None	None	None
Points on a convex curved boundary are always extrapolated, silently, with O(h) error	underworldcode/underworld3#605	None	lmoresi	None	None	1:40:15	None
A located point can interpolate to NaN, and #602 makes that invisible	underworldcode/underworld3#604	None	lmoresi	None	None	None	None
Two test files share the number 1029 (faulted medium and Zhong 2008)	underworldcode/underworld3#600	None	lmoresi	None	None	1 day, 21:49:17	None
Cutting a thin labelled band along its centre-line can snap a band-boundary vertex onto the cut	underworldcode/underworld3#595	None	lmoresi	None	None	None	None
Eighteen test modules build state at import: retire the collection-guard ratchet list	underworldcode/underworld3#587	None	lmoresi	None	None	None	None
Stokes velocity KSP truncates silently at its default 200-iteration cap	underworldcode/underworld3#585	None	lmoresi	None	None	None	None
GAMG defaults deviate from PETSc's and cripple it: pc_mg_type=additive, and block size 1 on vector fields	underworldcode/underworld3#579	None	lmoresi	None	None	2 days, 9:51:55	None
CylindricalStokes(boundary='zero') leaves the pressure nullspace in place	underworldcode/underworld3#577	None	lmoresi	None	None	4 days, 1:13:20	None
Stokes outer Krylov defaults to gmres while the fieldsplit sub-blocks are Krylov solves — must be fgmres	underworldcode/underworld3#576	None	lmoresi	None	None	2 days, 22:51:03	None
Nothing prevents a test module from mutating global state at import — add a collection-time guard	underworldcode/underworld3#575	None	lmoresi	None	None	1 day, 16:41:58	None
CI wall-clock tracks total test time: split the batches across parallel jobs (the 60-minute cap was reached, 120 is a stop-gap)	underworldcode/underworld3#573	None	lmoresi	None	None	None	None
Audit scripts/test.sh against the test tree: four coverage gaps in two days, each of which hid a real defect	underworldcode/underworld3#570	None	lmoresi	None	None	None	None
mesh.cell_size() is partition-dependent (kd-tree of this rank's centroids) — and add_nitsche_bc(local_h=True) is the DEFAULT, so Nitsche answers move 6.6e-03 between rank counts	underworldcode/underworld3#569	None	lmoresi	None	None	None	None
Three point-locator tests fail when xdist worker count changes the file grouping	underworldcode/underworld3#567	None	lmoresi	None	None	11:55:13	None
Constrained free-slip is partition-dependent (3.4%): boundary normals accumulated rank-locally — and the rotated tests compare against a golden recorded on another host	underworldcode/underworld3#564	None	lmoresi	None	None	6:56:17	None
The mesh cache races: '.meshes' is hardcoded in 42 places, so two concurrent UW3 processes in one directory corrupt each other	underworldcode/underworld3#563	None	lmoresi	None	None	17:43:16	None
FreeSurface records hscale but not tscale, so a published benchmark checkpoint cannot be evaluated at matched physical time	underworldcode/underworld3#562	None	lmoresi	None	None	None	None
Rotated free-slip: the vertex normal is a bisector but assembly is facet-wise, so a kinked boundary with unequal facets leaves the pressure gauge unpinned	underworldcode/underworld3#560	None	lmoresi	None	None	19:54:17	None
pull_vertex_onto: two control points can claim the same vertex, silently unplacing the first	underworldcode/underworld3#559	None	lmoresi	None	None	None	None
evaluate returns NaN in parallel for interior points the 50-neighbour walk cannot find (18 at np2, 120 at np4) — pre-existing	underworldcode/underworld3#558	None	lmoresi	None	None	4:26:00	None
Outcrop on curved boundaries: the wall descriptor is (axis, value) everywhere	underworldcode/underworld3#555	None	lmoresi	None	None	None	None
Thin viscous sheet (plane stress) solver — with a published analytic benchmark already available	underworldcode/underworld3#554	None	lmoresi	None	None	None	None
fault_split: allow a slit that crosses the boundary transversally (the daylighting contact)	underworldcode/underworld3#553	None	lmoresi	None	None	None	None
evaluate / global_evaluate: one pass over the locator, not five patches (anti-scaling to np=8, a smoothed substitution, an uncontained cell hint)	underworldcode/underworld3#551	None	lmoresi	None	None	None	None
2-D fault zones cannot reach a boundary — the outcrop machinery exists but is never wired into the ribbon path	underworldcode/underworld3#549	None	lmoresi	None	None	14:36:23	None
Stokes pointwise-function JIT: cache=False bypasses the rank-0 compile gate (every MPI rank compiles), and -O3 OOM-kills gcc on large expressions	underworldcode/underworld3#547	None	bknight1	None	None	None	None
Rotated workspace cache follow-ups from the #543 review: config latch into the cached KSP (m1), enumeration reuse by coincidence (m3)	underworldcode/underworld3#546	None	lmoresi	None	None	None	None
What should the rheology be where two fault zones overlap at a high angle?	underworldcode/underworld3#544	None	lmoresi	None	None	None	None
pull_vertex_onto can take the vertex from a neighbouring cell, so the tip cell has no on-line corner	underworldcode/underworld3#542	None	lmoresi	None	None	3 days, 6:31:57	None
SurfaceCollection nearest-fault director is non-functional in 2-D (three separate faults)	underworldcode/underworld3#540	None	lmoresi	None	None	None	None
TI: expose the compliance (inverse) tensor — same structure, reciprocal scale	underworldcode/underworld3#539	None	lmoresi	None	None	None	None
Gamma_P1 returns wrong boundary normals on a box: left/bottom rotated 90 deg, right/top position-dependent	underworldcode/underworld3#538	None	lmoresi	None	None	None	None
MMPDE mover: accel="cg" stalls after ~9 outer iterations and reports convergence	underworldcode/underworld3#531	None	lmoresi	None	None	None	None
CI flake: MPI_Init dies in UCX on some Azure runners (mana_0 verbs transport)	underworldcode/underworld3#528	None	lmoresi	None	None	3:47:49	None
place_sheet parallel rebuild: face/edge cone orientations unreconciled across the point SF — parallel FE on placed meshes assembles a wrong operator	underworldcode/underworld3#520	None	lmoresi	None	None	2:06:04	None
Geometric-MG bundle pairs a gmres smoother with a non-flexible outer KSP: converges by recurrence while the true residual stalls	underworldcode/underworld3#514	None	lmoresi	None	None	3 days, 6:42:31	None
global_evaluate: standalone probe with rank-distinct off-domain points deadlocks at np=4 (predates #488)	underworldcode/underworld3#513	None	lmoresi	None	None	4 days, 23:30:31	None
adapt/relax follow-ups from the #488 re-review: empty-rank eval_metric skip (3 sites), reconnect invariant guard, vacuous partition test	underworldcode/underworld3#512	None	lmoresi	None	None	9 days, 3:30:08	None
UWGeodynamics equivalent	underworldcode/underworld3#508	None	juliusdeblaaij	1:23:35	None	None	None
ViscoPlasticFlowModel + split-node fault path + consistent tangent: velocity PC setup dies with 'Computed maximum singular value as zero' (error 77)	underworldcode/underworld3#507	None	lmoresi	None	None	None	None
Dead bc_is = bc_label.getStratumIS(value) beside the natural-BC DS registration invites a rank-local skip	underworldcode/underworld3#506	None	lmoresi	None	None	10 days, 17:53:45	None
test_0050_utils.py runs two Stokes solves at module scope — pytest COLLECTION executes them and whole-directory runs hang silently	underworldcode/underworld3#505	None	lmoresi	None	None	10 days, 18:07:08	None
CI: the rotated free-slip suite (test_1018_*) is in no scripts/test.sh batch — its regression tests never run	underworldcode/underworld3#504	None	lmoresi	None	None	5 days, 22:36:06	None
Analytic normal for add_rotated_freeslip_bc fails on any mesh after the first (unwrap canonicalises coordinates to the first mesh's frame)	underworldcode/underworld3#501	None	lmoresi	None	None	20 days, 17:46:00	None
timing.print_table dropped display_fraction/group_by/output_file; 18 examples abort on it	underworldcode/underworld3#499	None	lmoresi	None	None	7 days, 20:34:11	None
EnhancedMeshVariable.clone() drops its arguments, breaking six shipped examples	underworldcode/underworld3#498	None	lmoresi	None	None	7 days, 20:38:35	None
Rotated-path field copy-back drops inhomogeneous essential (Dirichlet) values	underworldcode/underworld3#497	None	lmoresi	None	None	2 days, 1:52:09	None
FreeSurface 3D: pin the Y20 spherical relaxation rate in CI (test_1072 is capability-only)	underworldcode/underworld3#496	None	lmoresi	None	None	29 days, 19:49:21	None
Stokes_Constrained free-slip: the answer is partition-dependent (np2 iso -3.4%, np4 ti +2.8%) — not an environment-sensitive golden	underworldcode/underworld3#495	None	lmoresi	None	None	23 days, 1:34:04	None
cut_along_lines with several lines in one call returns incomplete chains, silently	underworldcode/underworld3#494	None	lmoresi	None	None	None	None
Creating a MeshVariable destroys the previous mesh.dm; a held handle segfaults (use-after-free)	underworldcode/underworld3#492	None	lmoresi	None	None	9 days, 11:44:35	None
uw.function.evaluate returns negative values for a squared DERIVATIVE expression	underworldcode/underworld3#491	None	lmoresi	None	None	11 days, 1:45:31	None
yield_anchor follow-ups from the #475 review: snes_max_it latch resurrect, side-attribution docstring, small validation gaps	underworldcode/underworld3#490	None	lmoresi	None	None	15 days, 20:57:46	None

This report was generated with the Issue Metrics Action
Search query used to find these items: repo:underworldcode/underworld2 repo:underworldcode/underworld3 is:issue created:2026-08-01..2026-08-31 -reason:"not planned"