Lines of Code
[root]/Packages/Uintah/CCA/Components/LoadBalancers
| Author | Changes | Lines of Code | Lines per Change |
|---|---|---|---|
| Totals | 73 (100.0%) | 5384 (100.0%) | 73.7 |
| luitjens | 56 (76.7%) | 4912 (91.2%) | 87.7 |
| worthen | 15 (20.5%) | 472 (8.8%) | 31.4 |
| dav | 2 (2.7%) | 0 (0.0%) | 0.0 |
Readded code that was removed because after further analysis we realized
it is still needed..... Justin & Bryan
0 lines of code changed in 4 files:
Missed a function in my code cleanup of the DLB
0 lines of code changed in 2 files:
Removed some more dangling code that is not used anymore
9 lines of code changed in 2 files:
Removed some code that is not used and was not maintained
1 lines of code changed in 2 files:
move SCISHARE to UINTAHSHARE. Also disable the share flags if building static
18 lines of code changed in 6 files:
Fix OSX linkng... depends on Core/DataArchive. Reorder list of libs to be alphabetical to make it easier to find what is being linked against and to remove duplicate libs.
0 lines of code changed in 2 files:
Leverage the new proc-per-patch info saved in timestep.xml to assign procs in the LB faster and with less memory
58 lines of code changed in 2 files:
Merge from 38823
3 lines of code changed in 1 file:
Merge from 38467
34 lines of code changed in 1 file:
Changed the load balancing assignment algorithm to a top down divide and conqure approach. This method should produce better assignments than the previous method. This method is O(P) in parallel.
125 lines of code changed in 1 file:
Undid a tweak to load balancer logic. It causes massive imbalance. I'm looking into why.
1 lines of code changed in 1 file:
Minor tweak to the load balancer logic
1 lines of code changed in 1 file:
Modified the knobs for controlling the load balancer weights.
There are now 3 knobs and they are listed as follows:
patchCost: cost per patch
cellCost: cost per cell
particleCost: cost per particle
The formula for determining a patches cost is as follows:
cost=patchCost+numCells*cellCost+numParticles*particleCost
The defaults are as follows:
patchCost: 0
cellCost: 2
particleCost: 1
These defaults seem to work well for an MPMICE exploding container.
And should also work well for ICE problems.
12 lines of code changed in 2 files:
Actually use the cellFactor
2 lines of code changed in 1 file:
Bug fixes in hybrid load balancer.
34 lines of code changed in 2 files:
removed accidently commited spew
1 lines of code changed in 1 file:
Fixed a couple bugs that would cause the simulation to run with only a
fraction of the processors until a regrid when restarting when the uda was output using
nperproc.
9 lines of code changed in 1 file:
Fixed a bug in the hybrid load balancer/regridder.
6 lines of code changed in 1 file:
Fix a few DLB bugs and add some SCI_DEBUG statements
44 lines of code changed in 2 files:
Implement getCosts, with particles, for Regions as well as the current grid (if load balancing independent of regridding).
346 lines of code changed in 2 files:
(19 more)