Skip to content

Commit 69f086e

Browse files
sjoelundsjoelund
committed
Give ThermoSysPro LoadVariation 360s to simulate
Both `CombinedCyclePowerPlant.LoadVariation` models need more than the 240s default, so they have been flapping between phase 6 and phase 5 for months: over the recorded wasm-jit runs they finish in 185.6-271.0s and 186.0-279.3s, and 20 of 33 runs of the first and 22 of 33 of the second were killed by the timeout instead. `update-ulimit-exe.py --branch "master wasm-jit"` asks for 360s for both, which is its usual 1.25 factor over the longest run that finished rounded up to the next 30s. Only those two entries are taken from that run; the rest of what it proposes is a separate question, and it does not offer these at all on master alone, where the model is reported as "ran into the timeout in force, so there is nothing to measure". Nothing here is wasm-jit specific, but it is where the effect shows: master's C executable never simulates either model -- it fails in NLS 10596 at t=283.315 -- while wasm-jit gets past that system and finishes the run, 4.2x faster than C on the same machine (158s vs 662s to reach t=250). So the timeout was scoring as a failure the one target that can do the model at all. Assisted-by: Claude Opus 5
1 parent 72e17dd commit 69f086e

1 file changed

Lines changed: 2 additions & 0 deletions

File tree

‎configs/conf.json‎

Lines changed: 2 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -641,7 +641,9 @@
641641
"libraryVersion":"master",
642642
"libraryVersionNameForTests":"",
643643
"ulimitExeModels":{
644+
"ThermoSysPro.Examples.Book.PowerPlants.CombinedCyclePowerPlant.LoadVariation":360,
644645
"ThermoSysPro.Examples.SimpleExamples.TestCentrifugalPump7":570,
646+
"ThermoSysPro.Fluid.Examples.Book.PowerPlants.CombinedCyclePowerPlant.LoadVariation":360,
645647
"ThermoSysPro.Fluid.Examples.SimpleExamples.TestDynamicWaterHeating":540
646648
}
647649
},

0 commit comments

Comments
 (0)