Waiting for other threads to finish takes a long time
Awaiting user inputI'm using the Gurobi optimizer via Julia in Visual Studio code.
The terminal tells me that the barrier solved the model in 20 iterations and 38.50 seconds. The terminal however is waiting for other threads to finish, which seems to not come to an end. How do I ensure that the model can finish and return my results?
Hope to hear back from you soon.
Cheers,
julia> optimize!(m)
Set parameter TimeLimit to value 120
Gurobi Optimizer version 11.0.2 build v11.0.2rc0 (win64 - Windows 11.0 (22621.2))
CPU model: Intel(R) Core(TM) i7-1065G7 CPU @ 1.30GHz, instruction set [SSE2|AVX|AVX2|AVX512]
Thread count: 4 physical cores, 8 logical processors, using up to 8 threads
Optimize a model with 1226389 rows, 1051210 columns and 2627879 nonzeros
Model fingerprint: 0x41d28332
Model has 175200 quadratic constraints
Variable types: 876010 continuous, 175200 integer (175200 binary)
Coefficient statistics:
Matrix range [1e-02, 2e+02]
QMatrix range [5e-03, 2e+01]
QLMatrix range [1e+00, 1e+00]
Objective range [1e+00, 1e+00]
Bounds range [0e+00, 0e+00]
RHS range [2e-02, 6e+06]
Presolve removed 989896 rows and 543147 columns (presolve time = 11s) ...
Presolve removed 989896 rows and 578187 columns
Presolve time: 11.99s
Presolved: 788373 rows, 788383 columns, 2049723 nonzeros
Presolved model has 157680 SOS constraint(s)
Variable types: 630703 continuous, 157680 integer (157680 binary)
Deterministic concurrent LP optimizer: primal simplex, dual simplex, and barrier
Showing barrier log only...
Root barrier log...
Ordering time: 0.26s
Barrier statistics:
AA' NZ : 1.729e+06
Factor NZ : 5.206e+06 (roughly 400 MB of memory)
Factor Ops : 7.194e+07 (less than 1 second per iteration)
Threads : 1
Objective Residual
Iter Primal Dual Primal Dual Compl Time
0 4.98926720e+10 -3.42574595e+11 9.86e+07 3.92e+01 2.06e+06 32s
1 1.30158983e+10 -8.69547291e+10 1.96e+07 1.71e-13 3.73e+05 32s
2 5.18175259e+08 -6.87464636e+09 1.91e+05 2.56e-13 8.87e+03 32s
3 3.73649737e+08 -4.45244779e+08 1.03e-08 1.63e-13 6.93e+02 33s
4 3.06998921e+08 7.97268736e+07 1.60e-08 6.39e-14 1.92e+02 33s
5 2.88362836e+08 2.19859372e+08 1.20e-08 3.20e-14 5.80e+01 33s
6 2.73817407e+08 2.47055844e+08 8.63e-09 3.20e-14 2.27e+01 33s
7 2.67168183e+08 2.57282034e+08 6.11e-09 3.55e-14 8.37e+00 34s
8 2.64554210e+08 2.60179181e+08 1.72e-08 3.20e-14 3.70e+00 34s
9 2.63495190e+08 2.61062753e+08 2.49e-08 3.55e-14 2.06e+00 34s
10 2.62903583e+08 2.61689914e+08 7.95e-09 3.02e-14 1.03e+00 35s
11 2.62436316e+08 2.62096165e+08 2.69e-08 3.20e-14 2.88e-01 35s
12 2.62324057e+08 2.62256559e+08 3.70e-08 4.26e-14 5.71e-02 35s
13 2.62304668e+08 2.62285007e+08 5.37e-08 2.84e-14 1.66e-02 35s
14 2.62301602e+08 2.62293929e+08 1.51e-08 3.00e-14 6.50e-03 36s
15 2.62300380e+08 2.62298832e+08 3.78e-09 3.55e-14 1.31e-03 36s
16 2.62300063e+08 2.62299415e+08 2.34e-09 2.84e-14 5.48e-04 37s
17 2.62299997e+08 2.62299997e+08 2.40e-09 4.26e-14 6.29e-07 37s
18 2.62299997e+08 2.62299997e+08 3.71e-07 4.62e-14 6.27e-07 38s
19 2.62299997e+08 2.62299997e+08 3.71e-07 5.33e-14 6.27e-07 38s
20 2.62299997e+08 2.62299997e+08 4.95e-07 3.37e-14 6.27e-07 39s
Barrier solved model in 20 iterations and 38.50 seconds (35.77 work units)
Optimal objective 2.62299997e+08
Root crossover log...
251310 DPushes remaining with DInf 0.0000000e+00 39s
0 DPushes remaining with DInf 0.0000000e+00 40s
85954 PPushes remaining with PInf 1.7983840e-04 40s
1853 PPushes remaining with PInf 1.5059971e-04 40s
0 PPushes remaining with PInf 0.0000000e+00 40s
Push phase complete: Pinf 0.0000000e+00, Dinf 4.6309987e-01 40s
Root simplex log...
Iteration Objective Primal Inf. Dual Inf. Time
224905 2.6230000e+08 0.000000e+00 4.630999e-01 40s
224914 2.6230000e+08 0.000000e+00 0.000000e+00 40s
Waiting for other threads to finish... 164s
-
Hi Peter,
This behavior looks suspicious. Could you please write the model to an MPS file, see How do I write an MPS model file from a third party API?, and then share the model such that we could have a closer look? Note that uploading files in the Community Forum is not possible but we discuss an alternative in Posting to the Community Forum.
Best regards,
Jaromił0
Please sign in to leave a comment.
Comments
1 comment