openg2g.coordinator¶
openg2g.coordinator
¶
Central coordinator: multi-rate simulation loop.
SimulationLog
dataclass
¶
Bases: Generic[DCStateT, GridStateT]
Accumulated simulation data from a coordinator run.
Attributes:
| Name | Type | Description |
|---|---|---|
dc_states |
list[DCStateT]
|
Every datacenter state (flat list, all sites). |
dc_states_by_site |
dict[str, list[DCStateT]]
|
Per-site datacenter states keyed by DC name. |
grid_states |
list[GridStateT]
|
Every grid state produced by the grid. |
commands |
list[DatacenterCommand | GridCommand]
|
All commands emitted by controllers. |
time_s |
list[float]
|
Simulation time at each grid step (seconds). |
events |
list[SimEvent]
|
Clock-stamped simulation events from all components. |
Source code in openg2g/coordinator.py
TickOutput
dataclass
¶
Bases: Generic[DCStateT, GridStateT]
Outputs produced by one base tick of Coordinator.step.
Empty-or-None fields mean the corresponding component did not step on
this tick. The base tick is the GCD of all component periods, so on any
given tick only the components whose dt_s is currently due will appear.
Attributes:
| Name | Type | Description |
|---|---|---|
t_s |
float
|
Simulation time at the start of this tick (seconds). |
dc_states |
dict[str, DCStateT]
|
Datacenter states produced this tick, keyed by DC name. Empty if no DC was due. |
grid_state |
GridStateT | None
|
Grid state produced this tick, or |
commands |
list[DatacenterCommand | GridCommand]
|
Commands emitted by controllers and dispatched during this tick (in the order they fired). |
sim_events |
list[SimEvent]
|
|
Source code in openg2g/coordinator.py
Coordinator
¶
Bases: Generic[DCStateT, GridStateT]
Multi-rate simulation coordinator.
Orchestrates datacenter, grid, and controller components at their respective rates. The base tick is the GCD of all component periods.
Parameters:
| Name | Type | Description | Default |
|---|---|---|---|
datacenters
|
Sequence[DatacenterBackend[DCStateT]]
|
List of datacenter backends. |
required |
grid
|
GridBackend[GridStateT]
|
Grid simulator backend. |
required |
controllers
|
Sequence[Controller[Any, Any]] | None
|
List of controllers, applied in order each tick. |
None
|
total_duration_s
|
int
|
Total simulation duration (integer seconds). |
0
|
live
|
bool
|
If True, synchronize with wall-clock time. |
False
|
Source code in openg2g/coordinator.py
120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 | |
reset()
¶
Reset coordinator and all sub-components for a fresh run.
start()
¶
Acquire resources on all sub-components and prepare event routing.
Must be called before [step][openg2g.coordinator.Coordinator.Coordinator.step] or
[dispatch_commands][openg2g.coordinator.Coordinator.Coordinator.dispatch_commands].
Source code in openg2g/coordinator.py
stop()
¶
Release resources on all sub-components (LIFO order).
Source code in openg2g/coordinator.py
step()
¶
Advance the simulation by one base tick and return its outputs.
The caller is responsible for the surrounding lifecycle: invoke
[reset][openg2g.coordinator.Coordinator.Coordinator.reset] and [start][openg2g.coordinator.Coordinator.Coordinator.start]
before the first call, and [stop][openg2g.coordinator.Coordinator.Coordinator.stop] after the
last. Iterating with run_iter() (for tick in coord.run_iter()) handles this
automatically; [run][openg2g.coordinator.Coordinator.Coordinator.run] is a thin wrapper that
accumulates every TickOutput into a SimulationLog.
Each tick:
- Step every datacenter whose
dt_sis due. - Step the grid if its
dt_sis due, passing the per-DC power sub-trace accumulated since the previous grid step. - Step every controller whose
dt_sis due, in registration order, and dispatch its commands immediately via [dispatch_commands][openg2g.coordinator.Coordinator.Coordinator.dispatch_commands]. - Drain
SimEvents emitted during the tick into the returnedTickOutput, then advance the clock.
Source code in openg2g/coordinator.py
dispatch_commands(commands)
¶
Route commands to the grid or to their target datacenter.
Uses the same dispatch rules as the in-loop controller block: a
DatacenterCommand is delivered to command.target (which must be
set), and a GridCommand goes to the coordinator's grid. External
drivers (e.g. an RL training environment) call this between
[step][openg2g.coordinator.Coordinator.Coordinator.step] invocations to inject control actions.
Source code in openg2g/coordinator.py
run_iter()
¶
Yield one TickOutput per base tick until total_duration_s.
Manages the [reset][openg2g.coordinator.Coordinator.Coordinator.reset] /
[start][openg2g.coordinator.Coordinator.Coordinator.start] / [stop][openg2g.coordinator.Coordinator.Coordinator.stop]
lifecycle automatically, including on early termination (e.g. when
the consumer breaks out of the loop). For fully external drivers
that need per-tick control (e.g. an RL Gym env, an interactive
viewer), use [step][openg2g.coordinator.Coordinator.Coordinator.step] directly and manage the
lifecycle by hand.
Source code in openg2g/coordinator.py
run()
¶
Run the full simulation end-to-end and return the accumulated log.