[Ipopt] Segmentation fault due to MUMPS memory allocation

Sean Hardesty sean.hardesty at gmail.com
Wed Jan 15 23:58:09 EST 2020


Do you know what version of MUMPS you are using? According to my notes from
experiments a couple years ago, MUMPS 4 does not use internal 64-bit
integer indexing, so if you have more than 2^31 non-zeros (or maybe 2^32),
it will crash. I think this is supported as of MUMPS 5.1.0, but I don't
know whether Ipopt works with MUMPS 5. In order to solve problems requiring
64-bit indexing, I had to use MA86 or MA97. MA27 and MA57 don't use 64-bit
integers either. I tried to use Pardiso, but had difficulty getting it to
link against MKL, so I don't know how it behaves...

    Sean

On Mon, Jan 13, 2020 at 1:34 AM Elias Trommer <elias.trommer at rl-institut.de>
wrote:

> Hi everyone,
>
> I'm trying to use ipopt as a solver for a non-linear optimization problem
> in Julia. This works fine for smaller instances of the problem, however for
> larger instances MUMPS seems to fail because of a failing memory
> allocation. The full error output is at the end of this eMail.
>
>
> Because I previously had problems with this, I've decreased the option
> `mumps_mem_percent` to a value of 100 so that the restoration would not
> fail during the first iteration as described in the error message.
>
> However, for this particular problem, I'm having a hard time pinning down
> the source. It's running on an Ubuntu 18.04 machine with 64Gb of RAM and
> 16Gb of Swap, so I presume it's not a lack of actual memory, but rather an
> allocation issue. I've also checked the ulimit on the machine I'm working
> on and it's set to unlimited.
>
>
> I was hoping someone could help me pin down the cause of this issue (and
> ideally how to fix it)!
>
>
> Thank you in advance,
>
> Elias Trommer
>
>
> ```
>
> This is Ipopt version 3.12.10, running with linear solver mumps.
> NOTE: Other linear solvers might be more efficient (see Ipopt
> documentation).
>
> Number of nonzeros in equality constraint Jacobian...:   863116
> Number of nonzeros in inequality constraint Jacobian.:   108177
> Number of nonzeros in Lagrangian Hessian.............:   358996
>
>
> Total number of variables............................:   145445
>                      variables with only lower bounds:        0
>                 variables with lower and upper bounds:   144771
>                      variables with only upper bounds:        0
> Total number of equality constraints.................:   145338
> Total number of inequality constraints...............:    36059
>         inequality constraints with only lower bounds:        0
>    inequality constraints with lower and upper bounds:        0
>         inequality constraints with only upper bounds:    36059
>
> iter    objective    inf_pr   inf_du lg(mu)  ||d||  lg(rg) alpha_du
> alpha_pr  ls
>    0  2.3869702e+03 4.60e+01 1.00e+00  -1.0 0.00e+00    -  0.00e+00
> 0.00e+00   0
>    1  2.3914541e+03 4.48e+01 2.58e+05  -1.0 6.23e+01    -  2.96e-05
> 2.61e-02F  1
>    2  2.3936754e+03 3.08e+02 7.11e+07  -1.0 6.07e+01    -  4.74e-01
> 9.61e-01f  1
>    3  2.3924281e+03 3.08e+02 7.10e+07  -1.0 4.05e+03  -2.0 1.11e-03
> 6.56e-04f  1
>    4  2.3924033e+03 3.08e+02 7.10e+07  -1.0 1.40e+03  -1.6 5.44e-03
> 3.68e-05f  1
>    5  2.3920925e+03 3.08e+02 7.10e+07  -1.0 3.06e+03  -2.1 2.68e-03
> 3.03e-04f  1
>    6  2.3918892e+03 3.08e+02 7.10e+07  -1.0 1.07e+04    -  6.52e-04
> 1.30e-04f  1
>    7  2.3915687e+03 3.07e+02 7.09e+07  -1.0 1.11e+03    -  5.12e-03
> 7.38e-04f  1
>    8  2.3908132e+03 3.07e+02 7.08e+07  -1.0 1.13e+03    -  1.08e-02
> 2.08e-03f  1
>    9  2.3907401e+03 3.07e+02 9.31e+09  -1.0 7.76e+02    -  2.79e-02
> 1.84e-04f  1
> iter    objective    inf_pr   inf_du lg(mu)  ||d||  lg(rg) alpha_du
> alpha_pr  ls
>   10  2.3892870e+03 3.05e+02 1.62e+10  -1.0 5.84e+02    -  2.56e-02
> 6.94e-03f  1
>   11  2.3894049e+03 3.04e+02 3.09e+10  -1.0 4.25e+02    -  4.15e-02
> 1.21e-03f  1
>   12  2.3960908e+03 2.88e+02 3.92e+10  -1.0 3.76e+02    -  8.17e-02
> 5.33e-02f  1
>   13  2.4009546e+03 2.78e+02 2.85e+10  -1.0 3.05e+02    -  3.87e-03
> 3.45e-02f  1
>   14  2.4012483e+03 2.77e+02 3.19e+10  -1.0 2.93e+02    -  1.40e-02
> 2.36e-03f  1
>   15  2.4038152e+03 2.72e+02 3.44e+10  -1.0 2.93e+02    -  3.19e-02
> 2.10e-02f  1
>   16  2.4038403e+03 2.72e+02 4.23e+10  -1.0 2.86e+02    -  2.55e-02
> 2.15e-04h  1
>   17  2.4036648e+03 2.71e+02 4.17e+10  -1.0 4.57e+03    -  3.36e-06
> 3.15e-04f  1
>   18  2.4069581e+03 2.67e+02 3.57e+10  -1.0 2.89e+02    -  2.86e-04
> 1.71e-02f  1
>   19  2.4069888e+03 2.67e+02 6.39e+10  -1.0 2.83e+02    -  2.26e-02
> 1.75e-04h  1
> iter    objective    inf_pr   inf_du lg(mu)  ||d||  lg(rg) alpha_du
> alpha_pr  ls
>   20  2.4068113e+03 2.67e+02 6.38e+10  -1.0 3.88e+04    -  1.85e-06
> 2.86e-05h  1
>   21  2.4067828e+03 2.67e+02 4.51e+10  -1.0 4.20e+03    -  3.96e-06
> 7.14e-05h  1
> MUMPS returned INFO(1) = -9 and requires more memory, reallocating.
> Attempt 1
>   Increasing icntl[13] from 80 to 160.
> MUMPS returned INFO(1) = -9 and requires more memory, reallocating.
> Attempt 1
>   Increasing icntl[13] from 160 to 320.
> MUMPS returned INFO(1) = -9 and requires more memory, reallocating.
> Attempt 2
>   Increasing icntl[13] from 320 to 640.
> MUMPS returned INFO(1) = -9 and requires more memory, reallocating.
> Attempt 3
>   Increasing icntl[13] from 640 to 1280.
> MUMPS returned INFO(1) = -9 and requires more memory, reallocating.
> Attempt 4
>   Increasing icntl[13] from 1280 to 2560.
>   22  2.4067829e+03 2.67e+02 4.51e+10  -1.0 5.70e+00   5.8 2.39e-05
> 1.79e-05h  1
> MUMPS returned INFO(1) = -9 and requires more memory, reallocating.
> Attempt 1
>   Increasing icntl[13] from 2560 to 5120.
> MUMPS returned INFO(1) = -9 and requires more memory, reallocating.
> Attempt 2
>   Increasing icntl[13] from 5120 to 10240.
> MUMPS returned INFO(1) = -9 and requires more memory, reallocating.
> Attempt 3
>   Increasing icntl[13] from 10240 to 20480.
>  Problem with integer stack size           1           1           6
>
> signal (11): Segmentation Fault
> ```
>
>
> _______________________________________________
> Ipopt mailing list
> Ipopt at list.coin-or.org
> https://list.coin-or.org/mailman/listinfo/ipopt
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://list.coin-or.org/pipermail/ipopt/attachments/20200115/1267a9f4/attachment.html>


More information about the Ipopt mailing list