OilfieldToolsNet/opm-simulators
4
0
Fork 0
mirror of https://github.com/OPM/opm-simulators.git synced 2024-11-29 20:43:49 -06:00
Code Issues Packages Projects Releases Wiki Activity

Remove parallel try-catch from prepareTimeStep because of nestedness.

Browse Source 
Although not declared as such, prepareTimeStep seems to be an internal
function (despite usage in a test) and hence error control can be done
in code calling it.

There was the following problem with the try-catch approach taken:
The calling site `BlackoilWellModel::assemble` looked like this:

```
OPM_BEGIN_PARALLEL_TRY_CATCH();
{
    if (iterationIdx == 0) {
        calculateExplicitQuantities(local_deferredLogger); // no parallel try-catch
        prepareTimeStep(local_deferredLogger); //includes parallel try-catch
    }
    updateWellControls(local_deferredLogger, /* check group controls */ true);

    // Set the well primary variables based on the value of well solutions
    initPrimaryVariablesEvaluation();

    maybeDoGasLiftOptimize(local_deferredLogger);
    assembleWellEq(dt, local_deferredLogger);
}
OPM_END_PARALLEL_TRY_CATCH_LOG(local_deferredLogger, "assemble() failed: ",
                               terminal_output_);
```

calculateExplicitQuantities had no parallel-try-catch clause inside,
but prepareTimeStep had one.
Unfortunately, calculateExplicitQuantities might throw (on some
processors). In that case non-throwing processors will try to trigger a
collective communication (to check for errors) in
prepareTimeStep. While the one throwing will move to the
OPM_END_PARALLEL_TRY_CATCH_LOG macro at the end and also trigger a different
collective communication. Booom, we have a deadlock.

With this patch there is no (nested parallel)-try-catch clause in the
functions called. (And if an exception is thrown in prepareTimeStep, it
will be logged as being an assemble failure).

The other option would have been to add parallel-try-catch clauses
to all functions called. That would have created a lot more
synchronization points limiting scalability even further.
This commit is contained in:
Markus Blatt 2021-09-23 15:22:48 +02:00
parent 3cda8a2fdb
commit f64230e462
1 changed files with 30 additions and 35 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

65
opm/simulators/wells/BlackoilWellModel_impl.hpp
View File

@ -1405,48 +1405,43 @@ namespace Opm {
BlackoilWellModel<TypeTag>::
prepareTimeStep(DeferredLogger& deferred_logger)
{
OPM_BEGIN_PARALLEL_TRY_CATCH();
{
for (const auto& well : well_container_) {
const bool old_well_operable = well->isOperable();
well->checkWellOperability(ebosSimulator_, this->wellState(), deferred_logger);
for (const auto& well : well_container_) {
const bool old_well_operable = well->isOperable();
well->checkWellOperability(ebosSimulator_, this->wellState(), deferred_logger);
if (!well->isOperable() ) continue;
if (!well->isOperable() ) continue;
auto& events = this->wellState().well(well->indexOfWell()).events;
if (events.hasEvent(WellState::event_mask)) {
well->updateWellStateWithTarget(ebosSimulator_, this->groupState(), this->wellState(), deferred_logger);
// There is no new well control change input within a report step,
// so next time step, the well does not consider to have effective events anymore.
events.clearEvent(WellState::event_mask);
}
auto& events = this->wellState().well(well->indexOfWell()).events;
if (events.hasEvent(WellState::event_mask)) {
well->updateWellStateWithTarget(ebosSimulator_, this->groupState(), this->wellState(), deferred_logger);
// There is no new well control change input within a report step,
// so next time step, the well does not consider to have effective events anymore.
events.clearEvent(WellState::event_mask);
}
// solve the well equation initially to improve the initial solution of the well model
if (param_.solve_welleq_initially_) {
well->solveWellEquation(ebosSimulator_, this->wellState(), this->groupState(), deferred_logger);
}
// solve the well equation initially to improve the initial solution of the well model
if (param_.solve_welleq_initially_) {
well->solveWellEquation(ebosSimulator_, this->wellState(), this->groupState(), deferred_logger);
}
const bool well_operable = well->isOperable();
if (!well_operable && old_well_operable) {
const Well& well_ecl = getWellEcl(well->name());
if (well_ecl.getAutomaticShutIn()) {
deferred_logger.info(" well " + well->name() + " gets SHUT at the beginning of the time step ");
} else {
if (!well->wellIsStopped()) {
deferred_logger.info(" well " + well->name() + " gets STOPPED at the beginning of the time step ");
well->stopWell();
}
const bool well_operable = well->isOperable();
if (!well_operable && old_well_operable) {
const Well& well_ecl = getWellEcl(well->name());
if (well_ecl.getAutomaticShutIn()) {
deferred_logger.info(" well " + well->name() + " gets SHUT at the beginning of the time step ");
} else {
if (!well->wellIsStopped()) {
deferred_logger.info(" well " + well->name() + " gets STOPPED at the beginning of the time step ");
well->stopWell();
}
} else if (well_operable && !old_well_operable) {
deferred_logger.info(" well " + well->name() + " gets REVIVED at the beginning of the time step ");
well->openWell();
}
} else if (well_operable && !old_well_operable) {
deferred_logger.info(" well " + well->name() + " gets REVIVED at the beginning of the time step ");
well->openWell();
}
} // end of for (const auto& well : well_container_)
updatePrimaryVariables(deferred_logger);
}
OPM_END_PARALLEL_TRY_CATCH_LOG(deferred_logger, "prepareTimestep() failed: ",
terminal_output_);
} // end of for (const auto& well : well_container_)
updatePrimaryVariables(deferred_logger);
}