Parallel Compute¶

RAS Commander provides multiple strategies for scaling HEC-RAS computations, from single-machine parallelization to distributed execution across multiple remote workers.

Scaling Strategies Overview¶

Strategy	Use Case	Throughput	Complexity
Sequential	Debugging, simple workflows	1x	Low
Local Parallel	Multi-core workstation	2-8x	Low
Remote Parallel	Multiple machines	10-100x+	Medium
Hybrid	Mixed local + remote	Maximum	High

Key Concepts¶

Worker Folders¶

All parallel execution methods use worker folders - isolated copies of the HEC-RAS project where computations run independently. This approach:

Prevents file locking conflicts
Enables concurrent plan execution
Preserves the original project files
Allows results collection from multiple runs

Plan Independence¶

HEC-RAS plans can run independently when they:

Don't share geometry preprocessor files
Don't write to the same output locations
Have no interdependencies (e.g., initial conditions from another run)

Core Allocation¶

HEC-RAS 2D computations can use multiple CPU cores. The optimal allocation depends on:

Model size: Larger meshes benefit from more cores
Available cores: Diminishing returns beyond 8-16 cores per plan
Concurrent plans: Balance cores per plan vs. simultaneous plans

Quick Comparison¶

Python

from ras_commander import RasCmdr, init_ras_project

init_ras_project("/path/to/project", "6.5")

# Sequential - one plan at a time
result = RasCmdr.compute_plan("01")

# Local Parallel - multiple plans on same machine
results = RasCmdr.compute_parallel(
    plan_number=["01", "02", "03", "04"],
    num_workers=4,
    num_cores=4
)

# Remote Parallel - distributed across machines
from ras_commander.remote import init_ras_worker, compute_parallel_remote

workers = [
    init_ras_worker("local", ras_version="6.5", num_cores=8),
    init_ras_worker("psexec", host="192.168.1.100", ...),
]
results = compute_parallel_remote(
    plan_number=["01", "02", "03", "04"],
    workers=workers
)

Documentation Sections¶

Local Parallel Execution - Scale on a single machine with compute_parallel
Remote Parallel Execution - Distribute across multiple machines with compute_parallel_remote
Scaling Strategies - Best practices for maximizing throughput
Worker Types - Available worker backends and configuration