Nightly parity: regressions detected

Nightly parity run https://github.com/lanl/bngsim/actions/runs/40712357505 on 9a26f5b5bd61.

job	result
amici-sens	success
bng-ode	success
bng-stoch	success
corpus	success
differential	failure
fd-sens	success
plan	success
publish	skipped
rr-ode	success
rr-ssa	success
sbml	success

A job that did not succeed either found a regression (below) or could not run; a job that could not run means that suite was not checked tonight.

differential — vs committed baseline

1 alert(s) · 13530 cases tonight, 13530 in baseline

outcome	baseline	tonight
DIFF	8	8
EXCEPTION	7	8
PASS	11225	11226
SKIP	2217	2217
TIMEOUT	20	18
UNSUPPORTED	53	53

Regressed (good before, not now): 1

case	before	after	note
BIOMD0000000338\|klu_vs_dense	PASS	EXCEPTION	SimulationError: Simulation failed: CVODE integration failed at t=0.720000 with flag=-4 (CV_CONV_FAILURE).

Improved (failing before, good now) — re-baseline to lock these in: 2

case	before	after	note
BIOMD0000000385\|ajac_vs_fdjac	TIMEOUT	PASS
BIOMD0000000385\|klu_vs_dense	TIMEOUT	PASS

differential — vs previous night (run 40521487203)

1 alert(s) · 13530 cases tonight, 13530 in baseline

outcome	baseline	tonight
DIFF	8	8
EXCEPTION	7	8
PASS	11225	11226
SKIP	2217	2217
TIMEOUT	20	18
UNSUPPORTED	53	53

Regressed (good before, not now): 1

case	before	after	note
BIOMD0000000338\|klu_vs_dense	PASS	EXCEPTION	SimulationError: Simulation failed: CVODE integration failed at t=0.720000 with flag=-4 (CV_CONV_FAILURE).

Changed between failing states: 2

case	before	after	note
BIOMD0000000385\|ajac_vs_fdjac	TIMEOUT	PASS	[timeout flip between two nights (runner speed)]
BIOMD0000000385\|klu_vs_dense	TIMEOUT	PASS	[timeout flip between two nights (runner speed)]

The one regression here, BIOMD0000000338|klu_vs_dense, is #1302's. It is not this night's; the same row has failed every PR's Path-agreement check since #1302 merged.

Summary

Since #1302 merged (aedadd8, 2026-10-27 17:31 UTC), every pull request's Path-agreement check fails on one row and nothing else:

BIOMD0000000338|klu_vs_dense   PASS -> EXCEPTION
SimulationError: Simulation failed: CVODE integration failed at t=0.720000 with flag=-4 (CV_CONV_FAILURE).

It isn't any one PR's: the five runs since #1302, on unrelated branches, each read exactly this single alert, against nightly/baselines/differential.json:

run	branch	alerts
40586213480	`fix/1290-ssa-record-past-cap` (#1325)	1, this row
40565116208	`fix/differential-model-cap` (#1321)	1, this row
40591591923	`fix/1313-typed-refusals` (#1336)	1, this row
40577384024	`fix/1309-attribution` (#1319)	1, this row
40577376012	`fix/1307-state-leaks` (#1318)	1, this row

Every PR run before #1302 merged had the row PASS.

Confirmed: #1302 causes it

Reproduced off CI, in a Linux x86-64 container (python:3.12-slim, pip install . from a clean clone), running the harness's own check_klu on BIOMD0000000338 at its nightly job (t 0–2, 101 points, rtol 1e-6, atol 1e-9):

build	`klu_vs_dense`
`b543f7b` (main just before #1302)	PASS: `jacobian="fd", KLU against dense: 1.12e-04 of the trajectory's peak (bound 1.0e-03)`
`aedadd8` (#1302's merge)	EXCEPTION: `CVODE integration failed at t=0.720000 with flag=-4 (CV_CONV_FAILURE)`

klu_vs_dense runs KLU against the dense solver on jacobian="fd". #1302 changed exactly that Jacobian: difference quotients across a jump in a rate law, #1002 (fd_jacobian.hpp, sparse_jacobian.hpp, cvode_simulator.cpp). With it, CVODE's corrector no longer converges at t = 0.72 on the KLU path.

On macOS arm64 the row reads #1304's DIFF (3.47e-2) both before and after #1302, so the change is platform-dependent. That fits #1304's model, whose KLU and dense runs already take different branches past a step-size collapse. The baseline's PASS was recorded on CI's x86-64 runner.

What's needed

A decision on this row:

Acceptable for a model already contested (BIOMD0000000338 and 339: KLU and the dense solver take different branches past a step-size collapse (Fg 311 on KLU, 0.73 on dense and libRoadRunner) #1304): re-record the row in differential.json as EXCEPTION, so other PRs stop carrying it;
or a regression from Difference quotients that step across a jump in a rate law (#1002) #1302 on the FD Jacobian's KLU path, to fix in a The differenced state Jacobian steps every species by 1.5e-8 of the largest concentration: dY_ss/dp with jacobian="fd" is 7% off beside an unrelated species at 1e7 and 44% at 1e8, with no error #1002 follow-up. The x86-64 Docker setup above reproduces it in a few minutes.

Until then every PR's Path-agreement check carries this one alert.

Nightly parity run https://github.com/lanl/bngsim/actions/runs/40899940910 on e42554658763.

job	result
amici-sens	success
bng-ode	success
bng-stoch	success
corpus	success
differential	success
fd-sens	success
plan	success
publish	failure
rr-ode	success
rr-ssa	success
sbml	success

A job that did not succeed either found a regression (below) or could not run; a job that could not run means that suite was not checked tonight.
No regressions.

Nightly parity run https://github.com/lanl/bngsim/actions/runs/40899940910 on e42554658763.

job	result
amici-sens	success
bng-ode	success
bng-stoch	success
corpus	success
differential	success
fd-sens	success
plan	success
publish	failure
rr-ode	success
rr-ssa	success
sbml	success

A job that did not succeed either found a regression (below) or could not run; a job that could not run means that suite was not checked tonight.
No regressions.

Nightly parity run https://github.com/lanl/bngsim/actions/runs/41210509438 on 7d38d6ffb29c.

job	result
amici-sens	success
bng-ode	success
bng-stoch	success
corpus	success
differential	success
differential-nf	failure
fd-sens	success
plan	success
publish	skipped
rr-ode	success
rr-ssa	success
sbml	success

A job that did not succeed either found a regression (below) or could not run; a job that could not run means that suite was not checked tonight.

differential_nf — vs committed baseline

2 alert(s) · 234 cases tonight, 234 in baseline

outcome	baseline	tonight
DIFF	5	6
PASS	225	223
TIMEOUT	3	4
UNSUPPORTED	1	1

Regressed (good before, not now): 2

case	before	after	note
original/bngl_models/my_models/nf/IL7_Carmen.bngl\|nf_reject_vs_exact	PASS	DIFF	10 replicates each: 0.983 of cells within 3.00 SE, max z 4.69; furthest apart: STAT5_tetramers \| escalation to N=30 aborted: _OverBudget
original/bngl_models/my_models/nf/local_func_bug.bngl\|nf_reject_vs_exact	PASS	TIMEOUT	cap of 900 s

differential_nf — vs previous night (run 41076909610)

1 alert(s) · 234 cases tonight, 234 in baseline

outcome	baseline	tonight
DIFF	5	6
PASS	225	223
TIMEOUT	3	4
UNSUPPORTED	1	1

Regressed (good before, not now): 1

case	before	after	note
original/bngl_models/my_models/nf/IL7_Carmen.bngl\|nf_reject_vs_exact	PASS	DIFF	10 replicates each: 0.983 of cells within 3.00 SE, max z 4.69; furthest apart: STAT5_tetramers \| escalation to N=30 aborted: _OverBudget

Changed between failing states: 1

case	before	after	note
original/bngl_models/my_models/nf/local_func_bug.bngl\|nf_reject_vs_exact	PASS	TIMEOUT	[timeout flip between two nights (runner speed)] cap of 900 s

Nightly parity run https://github.com/lanl/bngsim/actions/runs/41290236261 on 0456a9288f77.

job	result
amici-sens	success
bng-ode	success
bng-stoch	success
corpus	success
differential	success
differential-nf	success
fd-sens	success
plan	success
publish	success
rr-ode	failure
rr-ssa	failure
sbml	success

A job that did not succeed either found a regression (below) or could not run; a job that could not run means that suite was not checked tonight.

rr_ode — vs committed baseline

4 alert(s) · 1323 cases tonight, 1323 in baseline

outcome	baseline	tonight
BAD_TEST	53	53
PASS	1243	1249
REFERENCE_FAILED	27	16
TIMEOUT	0	5

Regressed (good before, not now): 4

case	before	after	note
BIOMD0000000338\|ode	PASS	REFERENCE_FAILED	roadrunner: RuntimeError: CVODE Error: CV_TOO_MUCH_WORK: The solver took mxstep (20000) internal steps but could not reach tout.; In virtual double rr::CVODEIntegrator::integrate(double, double)
MODEL1006230049\|ode	PASS	TIMEOUT	killed at 600.0s wall cap
MODEL1006230053\|ode	PASS	TIMEOUT	killed at 600.0s wall cap
MODEL1006230077\|ode	PASS	TIMEOUT	killed at 600.0s wall cap

Total bngsim work over cases good in both runs:

counter	baseline	tonight	ratio
n_steps	5363883	5364806	1.0002
n_rhs_evals	6702944	6703648	1.0001
n_jac_evals	881852	882037	1.0002
n_jac_rhs_evals	0	0	None

bngsim wall over 1239 cases: 404.537 s -> 398.351 s (x0.985; same CPU (AMD EPYC 7763 64-Core Processor); alerts above x2.0 and +30.0 s)

Improved (failing before, good now) — re-baseline to lock these in: 10

case	before	after
BIOMD0000000356\|ode	REFERENCE_FAILED	PASS
BIOMD0000000474\|ode	REFERENCE_FAILED	PASS
BIOMD0000000603\|ode	REFERENCE_FAILED	PASS
BIOMD0000000604\|ode	REFERENCE_FAILED	PASS
BIOMD0000000605\|ode	REFERENCE_FAILED	PASS
BIOMD0000000606\|ode	REFERENCE_FAILED	PASS
BIOMD0000000735\|ode	REFERENCE_FAILED	PASS
MODEL1912090001\|ode	REFERENCE_FAILED	PASS
MODEL2403070001\|ode	REFERENCE_FAILED	PASS
MODEL9811206584\|ode	REFERENCE_FAILED	PASS

Changed between failing states: 2

case	before	after	note
MODEL1504130000\|ode	REFERENCE_FAILED	TIMEOUT	killed at 600.0s wall cap
MODEL1603150001\|ode	REFERENCE_FAILED	TIMEOUT	killed at 600.0s wall cap

rr_ode — vs previous night (run 41210509438)

1 alert(s) · 1323 cases tonight, 1323 in baseline

outcome	baseline	tonight
BAD_TEST	53	53
PASS	1243	1249
REFERENCE_FAILED	27	16
TIMEOUT	0	5

Regressed (good before, not now): 1

case	before	after	note
BIOMD0000000338\|ode	PASS	REFERENCE_FAILED	roadrunner: RuntimeError: CVODE Error: CV_TOO_MUCH_WORK: The solver took mxstep (20000) internal steps but could not reach tout.; In virtual double rr::CVODEIntegrator::integrate(double, double)

Work on 186 case(s) the committed baseline re-recorded since then, which tonight does not exceed, is not compared.

Total bngsim work over cases good in both runs:

counter	baseline	tonight	ratio
n_steps	2655567	2655555	1.0
n_rhs_evals	3153250	3153205	1.0
n_jac_evals	320804	320796	1.0
n_jac_rhs_evals	0	0	None

bngsim wall over 1239 cases: 284.14 s -> 398.351 s (x1.402; CPU AMD EPYC 9V74 80-Core Processor -> AMD EPYC 7763 64-Core Processor; alerts above x4.0 and +30.0 s)

Improved (failing before, good now) — re-baseline to lock these in: 10

case	before	after
BIOMD0000000356\|ode	REFERENCE_FAILED	PASS
BIOMD0000000474\|ode	REFERENCE_FAILED	PASS
BIOMD0000000603\|ode	REFERENCE_FAILED	PASS
BIOMD0000000604\|ode	REFERENCE_FAILED	PASS
BIOMD0000000605\|ode	REFERENCE_FAILED	PASS
BIOMD0000000606\|ode	REFERENCE_FAILED	PASS
BIOMD0000000735\|ode	REFERENCE_FAILED	PASS
MODEL1912090001\|ode	REFERENCE_FAILED	PASS
MODEL2403070001\|ode	REFERENCE_FAILED	PASS
MODEL9811206584\|ode	REFERENCE_FAILED	PASS

Changed between failing states: 5

case	before	after	note
MODEL1006230049\|ode	PASS	TIMEOUT	[timeout flip between two nights (runner speed)] killed at 600.0s wall cap
MODEL1006230053\|ode	PASS	TIMEOUT	[timeout flip between two nights (runner speed)] killed at 600.0s wall cap
MODEL1006230077\|ode	PASS	TIMEOUT	[timeout flip between two nights (runner speed)] killed at 600.0s wall cap
MODEL1504130000\|ode	REFERENCE_FAILED	TIMEOUT	[timeout flip between two nights (runner speed)] killed at 600.0s wall cap
MODEL1603150001\|ode	REFERENCE_FAILED	TIMEOUT	[timeout flip between two nights (runner speed)] killed at 600.0s wall cap

rr_ssa — vs committed baseline

1 alert(s) · 269 cases tonight, 269 in baseline

outcome	baseline	tonight
BAD_TEST	7	7
DIFF	98	94
PASS	138	141
REFERENCE_FAILED	3	3
TIMEOUT	23	24

Regressed (good before, not now): 1

case	before	after	note
BIOMD0000000312\|ssa	PASS	TIMEOUT	exceeded 120.0s wall cap

bngsim wall over 137 cases: 277.861 s -> 221.775 s (x0.798; CPU None -> None; alerts above x4.0 and +30.0 s)

Improved (failing before, good now) — re-baseline to lock these in: 4

case	before	after
BIOMD0000000530\|ssa	TIMEOUT	PASS (vacuous_lowcount)
BIOMD0000000863\|ssa	DIFF (diff_not_bngsim)	PASS
BIOMD0000001026\|ssa	DIFF (rr_known)	PASS
BIOMD0000001040\|ssa	DIFF (rr_known)	PASS

Changed between failing states: 8

case	before	after	note
BIOMD0000000762\|ssa	DIFF (partial_horizon)	TIMEOUT	exceeded 120.0s wall cap
BIOMD0000000767\|ssa	DIFF (partial_horizon)	TIMEOUT	exceeded 120.0s wall cap
BIOMD0000000800\|ssa	TIMEOUT	DIFF (partial_horizon)	partial-horizon adjudication over t=0→0.1: ensembles agreed (exact SSA intractable beyond this window — no full-horizon verdict)
BIOMD0000001013\|ssa	TIMEOUT	DIFF (partial_horizon)	partial-horizon adjudication over t=0→0.001: ensembles agreed (exact SSA intractable beyond this window — no full-horizon verdict)
BIOMD0000001053\|ssa	PASS	PASS (vacuous_lowcount)
MODEL1608100001\|ssa	DIFF (diff_not_bngsim)	TIMEOUT	exceeded 120.0s wall cap
MODEL2001160001\|ssa	TIMEOUT	DIFF (partial_horizon)	partial-horizon adjudication over t=0→0.001: ensembles agreed (exact SSA intractable beyond this window — no full-horizon verdict)
MODEL8459127548\|ssa	DIFF (partial_horizon)	TIMEOUT	exceeded 120.0s wall cap

Nightly parity run https://github.com/lanl/bngsim/actions/runs/41320113549 on 0c1732dc0655.

job	result
amici-sens	success
bng-ode	success
bng-stoch	success
corpus	success
differential	failure
differential-nf	success
fd-sens	success
plan	success
publish	skipped
rr-ode	failure
rr-ssa	failure
sbml	success

A job that did not succeed either found a regression (below) or could not run; a job that could not run means that suite was not checked tonight.

differential — vs committed baseline

3 alert(s) · 28949 cases tonight, 28949 in baseline

outcome	baseline	tonight
DIFF	22	22
EXCEPTION	6	6
PASS	23090	23087
SKIP	5005	4984
TIMEOUT	56	80
UNSUPPORTED	770	770

Regressed (good before, not now): 3

case	before	after	note
MODEL1006230049\|locale	PASS	TIMEOUT	not finished within the model's cap: Command '['/home/runner/work/bngsim/bngsim/.venv/bin/python', '-c', '\nimport json, locale, logging, sys, warnings\nwarnings.filterwarnings("ignore"); logging.d...
MODEL1006230053\|locale	PASS	TIMEOUT	not finished within the model's cap: Command '['/home/runner/work/bngsim/bngsim/.venv/bin/python', '-c', '\nimport json, locale, logging, sys, warnings\nwarnings.filterwarnings("ignore"); logging.d...
MODEL1006230077\|locale	PASS	TIMEOUT	not finished within the model's cap: Command '['/home/runner/work/bngsim/bngsim/.venv/bin/python', '-c', '\nimport json, locale, logging, sys, warnings\nwarnings.filterwarnings("ignore"); logging.d...

Changed between failing states: 21

case	before	after	note
MODEL1006230090\|ajac_vs_fdjac	SKIP	TIMEOUT	cap of 1500 s
MODEL1006230090\|backend_as_asked	SKIP	TIMEOUT	cap of 1500 s
MODEL1006230090\|batch	SKIP	TIMEOUT	cap of 1500 s
MODEL1006230090\|cc_vs_interp	SKIP	TIMEOUT	cap of 1500 s
MODEL1006230090\|cc_vs_mir	SKIP	TIMEOUT	cap of 1500 s
MODEL1006230090\|grid	SKIP	TIMEOUT	cap of 1500 s
MODEL1006230090\|jacobian_values	SKIP	TIMEOUT	cap of 1500 s
MODEL1006230090\|klu_vs_dense	SKIP	TIMEOUT	cap of 1500 s
MODEL1006230090\|locale	SKIP	TIMEOUT	cap of 1500 s
MODEL1006230090\|newton_vs_run	SKIP	TIMEOUT	cap of 1500 s
MODEL1006230090\|rename	SKIP	TIMEOUT	cap of 1500 s
MODEL1006230090\|reorder	SKIP	TIMEOUT	cap of 1500 s
MODEL1006230090\|rescale_time	SKIP	TIMEOUT	cap of 1500 s
MODEL1006230090\|restart	SKIP	TIMEOUT	cap of 1500 s
MODEL1006230090\|reuse	SKIP	TIMEOUT	cap of 1500 s
MODEL1006230090\|sens_runs	SKIP	TIMEOUT	cap of 1500 s
MODEL1006230090\|sens_vs_plain	SKIP	TIMEOUT	cap of 1500 s
MODEL1006230090\|sparsity_covers_jacobian	SKIP	TIMEOUT	cap of 1500 s
MODEL1006230090\|split	SKIP	TIMEOUT	cap of 1500 s
MODEL1006230090\|stable	SKIP	TIMEOUT	cap of 1500 s
MODEL1006230090\|steady_vs_run	SKIP	TIMEOUT	cap of 1500 s

rr_ode — vs committed baseline

4 alert(s) · 1323 cases tonight, 1323 in baseline

outcome	baseline	tonight
BAD_TEST	53	52
PASS	1243	1249
REFERENCE_FAILED	27	16
TIMEOUT	0	6

Regressed (good before, not now): 4

case	before	after	note
BIOMD0000000338\|ode	PASS	REFERENCE_FAILED	roadrunner: RuntimeError: CVODE Error: CV_TOO_MUCH_WORK: The solver took mxstep (20000) internal steps but could not reach tout.; In virtual double rr::CVODEIntegrator::integrate(double, double)
MODEL1006230049\|ode	PASS	TIMEOUT	killed at 300.0s wall cap
MODEL1006230053\|ode	PASS	TIMEOUT	killed at 300.0s wall cap
MODEL1006230077\|ode	PASS	TIMEOUT	killed at 300.0s wall cap

Total bngsim work over cases good in both runs:

counter	baseline	tonight	ratio
n_steps	5363883	5364806	1.0002
n_rhs_evals	6702944	6703648	1.0001
n_jac_evals	881852	882037	1.0002
n_jac_rhs_evals	0	0	None

bngsim wall over 1239 cases: 404.537 s -> 280.416 s (x0.693; CPU AMD EPYC 7763 64-Core Processor -> AMD EPYC 9V45 96-Core Processor; alerts above x4.0 and +30.0 s)

Improved (failing before, good now) — re-baseline to lock these in: 10

case	before	after
BIOMD0000000356\|ode	REFERENCE_FAILED	PASS
BIOMD0000000474\|ode	REFERENCE_FAILED	PASS
BIOMD0000000603\|ode	REFERENCE_FAILED	PASS
BIOMD0000000604\|ode	REFERENCE_FAILED	PASS
BIOMD0000000605\|ode	REFERENCE_FAILED	PASS
BIOMD0000000606\|ode	REFERENCE_FAILED	PASS
BIOMD0000000735\|ode	REFERENCE_FAILED	PASS
MODEL1912090001\|ode	REFERENCE_FAILED	PASS
MODEL2403070001\|ode	REFERENCE_FAILED	PASS
MODEL9811206584\|ode	REFERENCE_FAILED	PASS

Changed between failing states: 3

case	before	after	note
MODEL1006230090\|ode	BAD_TEST	TIMEOUT	killed at 300.0s wall cap
MODEL1504130000\|ode	REFERENCE_FAILED	TIMEOUT	killed at 600.0s wall cap
MODEL1603150001\|ode	REFERENCE_FAILED	TIMEOUT	killed at 600.0s wall cap

rr_ssa — vs committed baseline

1 alert(s) · 269 cases tonight, 269 in baseline

outcome	baseline	tonight
BAD_TEST	7	7
DIFF	98	96
PASS	138	138
REFERENCE_FAILED	3	3
TIMEOUT	23	25

Regressed (good before, not now): 1

case	before	after	note
BIOMD0000000312\|ssa	PASS	TIMEOUT	exceeded 120.0s wall cap

bngsim wall over 137 cases: 277.861 s -> 276.799 s (x0.996; CPU None -> None; alerts above x4.0 and +30.0 s)

Improved (failing before, good now) — re-baseline to lock these in: 1

case	before	after	note
BIOMD0000000863\|ssa	DIFF (diff_not_bngsim)	PASS

Changed between failing states: 6

case	before	after	note
BIOMD0000000800\|ssa	TIMEOUT	DIFF (partial_horizon)	partial-horizon adjudication over t=0→0.01: ensembles agreed (exact SSA intractable beyond this window — no full-horizon verdict)
BIOMD0000001026\|ssa	DIFF (rr_known)	TIMEOUT	exceeded 120.0s wall cap
BIOMD0000001040\|ssa	DIFF (rr_known)	TIMEOUT	exceeded 120.0s wall cap
BIOMD0000001053\|ssa	PASS	PASS (vacuous_lowcount)
MODEL1608100001\|ssa	DIFF (diff_not_bngsim)	TIMEOUT	exceeded 120.0s wall cap
MODEL2002030001\|ssa	TIMEOUT	DIFF (partial_horizon)	partial-horizon adjudication over t=0→0.001: ensembles agreed (exact SSA intractable beyond this window — no full-horizon verdict)

Nightly parity run https://github.com/lanl/bngsim/actions/runs/41499451003 on c77304da2a1b.

job	result
amici-sens	success
bng-ode	success
bng-stoch	success
corpus	success
differential	failure
differential-nf	success
fd-sens	success
plan	success
publish	skipped
rr-ode	success
rr-ssa	success
sbml	success

A job that did not succeed either found a regression (below) or could not run; a job that could not run means that suite was not checked tonight.

differential — vs committed baseline

1 alert(s) · 28949 cases tonight, 28949 in baseline

outcome	baseline	tonight
DIFF	22	22
EXCEPTION	6	6
PASS	23090	23117
SKIP	5005	5003
TIMEOUT	56	22
UNSUPPORTED	770	779

New crashes: 1

case	before	after	note
MODEL9811206584\|klu_vs_dense	SKIP	EXCEPTION	SimulationError: Simulation failed: CVODE integration failed at t=100.000000 with flag=-10 (CV_REPTD_RHSFUNC_ERR). The RHS returned a non-finite value at t=99.3034. Non-finite there: phi_b_s() = (1...

Improved (failing before, good now) — re-baseline to lock these in: 27

case	before	after
BIOMD0000000385\|grid	TIMEOUT	PASS
BIOMD0000000385\|locale	TIMEOUT	PASS
BIOMD0000000385\|sparsity_covers_jacobian	TIMEOUT	PASS
MODEL1006230049\|batch	TIMEOUT	PASS
MODEL1006230049\|rename	TIMEOUT	PASS
MODEL1006230049\|restart	TIMEOUT	PASS
MODEL1006230049\|reuse	TIMEOUT	PASS
MODEL1006230053\|batch	TIMEOUT	PASS
MODEL1006230053\|rename	TIMEOUT	PASS
MODEL1006230053\|restart	TIMEOUT	PASS
MODEL1006230053\|reuse	TIMEOUT	PASS
MODEL1006230077\|batch	TIMEOUT	PASS
MODEL1006230077\|rename	TIMEOUT	PASS
MODEL1006230077\|restart	TIMEOUT	PASS
MODEL1006230077\|reuse	TIMEOUT	PASS
MODEL1812040005\|stable	DIFF	PASS
MODEL9811206584\|ajac_vs_fdjac	SKIP	PASS
MODEL9811206584\|backend_as_asked	SKIP	PASS
MODEL9811206584\|batch	SKIP	PASS
MODEL9811206584\|cc_vs_interp	SKIP	PASS
MODEL9811206584\|grid	SKIP	PASS
MODEL9811206584\|locale	SKIP	PASS
MODEL9811206584\|rename	SKIP	PASS
MODEL9811206584\|restart	SKIP	PASS
MODEL9811206584\|reuse	SKIP	PASS
MODEL9811206584\|sparsity_covers_jacobian	SKIP	PASS
MODEL9811206584\|stable	SKIP	PASS

Changed between failing states: 23

case	before	after	note
BIOMD0000000385\|cc_vs_mir	TIMEOUT	SKIP	a build without MIR
MODEL1006230049\|jacobian_values	TIMEOUT	SKIP	no analytic Jacobian: Model.jacobian differences the RHS
MODEL1006230049\|newton_vs_run	TIMEOUT	UNSUPPORTED	SimulationError: steady_state() is not supported for this model: it reads the time, through time() or a table function indexed by time, in a rate law, a rule or a reported function. The steady-stat...
MODEL1006230049\|reorder	TIMEOUT	SKIP	reorder does not apply: fewer than two species and reactions to reorder
MODEL1006230049\|rescale_time	TIMEOUT	SKIP	rescale_time does not apply: reads the clock or a delay
MODEL1006230049\|split	TIMEOUT	SKIP	split does not apply: no reaction whose rate is not read by name elsewhere
MODEL1006230049\|steady_vs_run	TIMEOUT	UNSUPPORTED	SimulationError: steady_state() is not supported for this model: it reads the time, through time() or a table function indexed by time, in a rate law, a rule or a reported function. The steady-stat...
MODEL1006230053\|jacobian_values	TIMEOUT	SKIP	no analytic Jacobian: Model.jacobian differences the RHS
MODEL1006230053\|newton_vs_run	TIMEOUT	UNSUPPORTED	SimulationError: steady_state() is not supported for this model: it reads the time, through time() or a table function indexed by time, in a rate law, a rule or a reported function. The steady-stat...
MODEL1006230053\|reorder	TIMEOUT	SKIP	reorder does not apply: fewer than two species and reactions to reorder
MODEL1006230053\|rescale_time	TIMEOUT	SKIP	rescale_time does not apply: reads the clock or a delay
MODEL1006230053\|split	TIMEOUT	SKIP	split does not apply: no reaction whose rate is not read by name elsewhere
MODEL1006230053\|steady_vs_run	TIMEOUT	UNSUPPORTED	SimulationError: steady_state() is not supported for this model: it reads the time, through time() or a table function indexed by time, in a rate law, a rule or a reported function. The steady-stat...
MODEL1006230077\|jacobian_values	TIMEOUT	SKIP	no analytic Jacobian: Model.jacobian differences the RHS
MODEL1006230077\|newton_vs_run	TIMEOUT	UNSUPPORTED	SimulationError: steady_state() is not supported for this model: it reads the time, through time() or a table function indexed by time, in a rate law, a rule or a reported function. The steady-stat...
MODEL1006230077\|reorder	TIMEOUT	SKIP	reorder does not apply: fewer than two species and reactions to reorder
MODEL1006230077\|rescale_time	TIMEOUT	SKIP	rescale_time does not apply: reads the clock or a delay
MODEL1006230077\|split	TIMEOUT	SKIP	split does not apply: no reaction whose rate is not read by name elsewhere
MODEL1006230077\|steady_vs_run	TIMEOUT	UNSUPPORTED	SimulationError: steady_state() is not supported for this model: it reads the time, through time() or a table function indexed by time, in a rate law, a rule or a reported function. The steady-stat...
MODEL1812040005\|reorder	EXCEPTION	DIFF	reorder (species, reactions and parameters reversed): 2.56e+01 of the trajectory's peak (bound 1.0e-03, own drift 1.1e-06)
MODEL9811206584\|newton_vs_run	SKIP	UNSUPPORTED	SimulationError: steady_state() is not supported for this model: it reads the time, through time() or a table function indexed by time, in a rate law, a rule or a reported function. The steady-stat...
MODEL9811206584\|sens_runs	SKIP	UNSUPPORTED	SimulationError: Simulation failed: CVODE integration failed at t=1.000000 with flag=-42 (CV_FIRST_SRHSFUNC_ERR). The compiled sensitivity RHS returned a non-finite value at t=0. Every rate law and...
MODEL9811206584\|steady_vs_run	SKIP	UNSUPPORTED	SimulationError: steady_state() is not supported for this model: it reads the time, through time() or a table function indexed by time, in a rate law, a rule or a reported function. The steady-stat...

differential — vs previous night (run 41320113549)

1 alert(s) · 28949 cases tonight, 28949 in baseline

outcome	baseline	tonight
DIFF	22	22
EXCEPTION	6	6
PASS	23087	23117
SKIP	4984	5003
TIMEOUT	80	22
UNSUPPORTED	770	779

New crashes: 1

case	before	after	note
MODEL9811206584\|klu_vs_dense	SKIP	EXCEPTION	SimulationError: Simulation failed: CVODE integration failed at t=100.000000 with flag=-10 (CV_REPTD_RHSFUNC_ERR). The RHS returned a non-finite value at t=99.3034. Non-finite there: phi_b_s() = (1...

Improved (failing before, good now) — re-baseline to lock these in: 12

case	before	after
MODEL1812040005\|stable	DIFF	PASS
MODEL9811206584\|ajac_vs_fdjac	SKIP	PASS
MODEL9811206584\|backend_as_asked	SKIP	PASS
MODEL9811206584\|batch	SKIP	PASS
MODEL9811206584\|cc_vs_interp	SKIP	PASS
MODEL9811206584\|grid	SKIP	PASS
MODEL9811206584\|locale	SKIP	PASS
MODEL9811206584\|rename	SKIP	PASS
MODEL9811206584\|restart	SKIP	PASS
MODEL9811206584\|reuse	SKIP	PASS
MODEL9811206584\|sparsity_covers_jacobian	SKIP	PASS
MODEL9811206584\|stable	SKIP	PASS

Changed between failing states: 62

case	before	after	note
BIOMD0000000385\|cc_vs_mir	TIMEOUT	SKIP	[timeout flip between two nights (runner speed)] a build without MIR
BIOMD0000000385\|grid	TIMEOUT	PASS	[timeout flip between two nights (runner speed)]
BIOMD0000000385\|locale	TIMEOUT	PASS	[timeout flip between two nights (runner speed)]
BIOMD0000000385\|sparsity_covers_jacobian	TIMEOUT	PASS	[timeout flip between two nights (runner speed)]
MODEL1006230049\|batch	TIMEOUT	PASS	[timeout flip between two nights (runner speed)]
MODEL1006230049\|jacobian_values	TIMEOUT	SKIP	[timeout flip between two nights (runner speed)] no analytic Jacobian: Model.jacobian differences the RHS
MODEL1006230049\|locale	TIMEOUT	PASS	[timeout flip between two nights (runner speed)]
MODEL1006230049\|newton_vs_run	TIMEOUT	UNSUPPORTED	[timeout flip between two nights (runner speed)] SimulationError: steady_state() is not supported for this model: it reads the time, through time() or a table function indexed by time, in a rate law, a rule or a reported function. The steady-stat...
MODEL1006230049\|rename	TIMEOUT	PASS	[timeout flip between two nights (runner speed)]
MODEL1006230049\|reorder	TIMEOUT	SKIP	[timeout flip between two nights (runner speed)] reorder does not apply: fewer than two species and reactions to reorder
MODEL1006230049\|rescale_time	TIMEOUT	SKIP	[timeout flip between two nights (runner speed)] rescale_time does not apply: reads the clock or a delay
MODEL1006230049\|restart	TIMEOUT	PASS	[timeout flip between two nights (runner speed)]
MODEL1006230049\|reuse	TIMEOUT	PASS	[timeout flip between two nights (runner speed)]
MODEL1006230049\|split	TIMEOUT	SKIP	[timeout flip between two nights (runner speed)] split does not apply: no reaction whose rate is not read by name elsewhere
MODEL1006230049\|steady_vs_run	TIMEOUT	UNSUPPORTED	[timeout flip between two nights (runner speed)] SimulationError: steady_state() is not supported for this model: it reads the time, through time() or a table function indexed by time, in a rate law, a rule or a reported function. The steady-stat...
MODEL1006230053\|batch	TIMEOUT	PASS	[timeout flip between two nights (runner speed)]
MODEL1006230053\|jacobian_values	TIMEOUT	SKIP	[timeout flip between two nights (runner speed)] no analytic Jacobian: Model.jacobian differences the RHS
MODEL1006230053\|locale	TIMEOUT	PASS	[timeout flip between two nights (runner speed)]
MODEL1006230053\|newton_vs_run	TIMEOUT	UNSUPPORTED	[timeout flip between two nights (runner speed)] SimulationError: steady_state() is not supported for this model: it reads the time, through time() or a table function indexed by time, in a rate law, a rule or a reported function. The steady-stat...
MODEL1006230053\|rename	TIMEOUT	PASS	[timeout flip between two nights (runner speed)]
MODEL1006230053\|reorder	TIMEOUT	SKIP	[timeout flip between two nights (runner speed)] reorder does not apply: fewer than two species and reactions to reorder
MODEL1006230053\|rescale_time	TIMEOUT	SKIP	[timeout flip between two nights (runner speed)] rescale_time does not apply: reads the clock or a delay
MODEL1006230053\|restart	TIMEOUT	PASS	[timeout flip between two nights (runner speed)]
MODEL1006230053\|reuse	TIMEOUT	PASS	[timeout flip between two nights (runner speed)]
MODEL1006230053\|split	TIMEOUT	SKIP	[timeout flip between two nights (runner speed)] split does not apply: no reaction whose rate is not read by name elsewhere
MODEL1006230053\|steady_vs_run	TIMEOUT	UNSUPPORTED	[timeout flip between two nights (runner speed)] SimulationError: steady_state() is not supported for this model: it reads the time, through time() or a table function indexed by time, in a rate law, a rule or a reported function. The steady-stat...
MODEL1006230077\|batch	TIMEOUT	PASS	[timeout flip between two nights (runner speed)]
MODEL1006230077\|jacobian_values	TIMEOUT	SKIP	[timeout flip between two nights (runner speed)] no analytic Jacobian: Model.jacobian differences the RHS
MODEL1006230077\|locale	TIMEOUT	PASS	[timeout flip between two nights (runner speed)]
MODEL1006230077\|newton_vs_run	TIMEOUT	UNSUPPORTED	[timeout flip between two nights (runner speed)] SimulationError: steady_state() is not supported for this model: it reads the time, through time() or a table function indexed by time, in a rate law, a rule or a reported function. The steady-stat...
MODEL1006230077\|rename	TIMEOUT	PASS	[timeout flip between two nights (runner speed)]
MODEL1006230077\|reorder	TIMEOUT	SKIP	[timeout flip between two nights (runner speed)] reorder does not apply: fewer than two species and reactions to reorder
MODEL1006230077\|rescale_time	TIMEOUT	SKIP	[timeout flip between two nights (runner speed)] rescale_time does not apply: reads the clock or a delay
MODEL1006230077\|restart	TIMEOUT	PASS	[timeout flip between two nights (runner speed)]
MODEL1006230077\|reuse	TIMEOUT	PASS	[timeout flip between two nights (runner speed)]
MODEL1006230077\|split	TIMEOUT	SKIP	[timeout flip between two nights (runner speed)] split does not apply: no reaction whose rate is not read by name elsewhere
MODEL1006230077\|steady_vs_run	TIMEOUT	UNSUPPORTED	[timeout flip between two nights (runner speed)] SimulationError: steady_state() is not supported for this model: it reads the time, through time() or a table function indexed by time, in a rate law, a rule or a reported function. The steady-stat...
MODEL1006230090\|ajac_vs_fdjac	TIMEOUT	SKIP	[timeout flip between two nights (runner speed)] the base run: SimulationError: Simulation failed: CVODE integration failed at t=1.000000 with flag=-4 (CV_CONV_FAILURE). The RHS returned a non-finite value at t=0.282204. Non-finite there: K_eq_F1...
MODEL1006230090\|backend_as_asked	TIMEOUT	SKIP	[timeout flip between two nights (runner speed)] the base run: SimulationError: Simulation failed: CVODE integration failed at t=1.000000 with flag=-4 (CV_CONV_FAILURE). The RHS returned a non-finite value at t=0.282204. Non-finite there: K_eq_F1...
MODEL1006230090\|batch	TIMEOUT	SKIP	[timeout flip between two nights (runner speed)] the base run: SimulationError: Simulation failed: CVODE integration failed at t=1.000000 with flag=-4 (CV_CONV_FAILURE). The RHS returned a non-finite value at t=0.282204. Non-finite there: K_eq_F1...
... 22 more (see the run's artifacts)

Navigation Menu

Repository navigation

Nightly parity: regressions detected #1355

differential — vs committed baseline

differential — vs previous night (run 40521487203)

Summary

Confirmed: #1302 causes it

What's needed

differential_nf — vs committed baseline

differential_nf — vs previous night (run 41076909610)

rr_ode — vs committed baseline

rr_ode — vs previous night (run 41210509438)

rr_ssa — vs committed baseline

differential — vs committed baseline

rr_ode — vs committed baseline

rr_ssa — vs committed baseline

differential — vs committed baseline

differential — vs previous night (run 41320113549)

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Nightly parity: regressions detected #1355

Description

differential — vs committed baseline

differential — vs previous night (run 40521487203)

Activity

Montana commented on Oct 28, 2026

Summary

Confirmed: #1302 causes it

What's needed

github-actions commented on Oct 29, 2026

github-actions commented on Oct 29, 2026

github-actions commented on Oct 31, 2026

differential_nf — vs committed baseline

differential_nf — vs previous night (run 41076909610)

github-actions commented on Nov 1, 2026

rr_ode — vs committed baseline

rr_ode — vs previous night (run 41210509438)

rr_ssa — vs committed baseline

github-actions commented on Nov 1, 2026

differential — vs committed baseline

rr_ode — vs committed baseline

rr_ssa — vs committed baseline

github-actions commented on Nov 2, 2026

differential — vs committed baseline

differential — vs previous night (run 41320113549)

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions