<div class="gmail_quote">On Fri, Nov 25, 2011 at 15:11, Dominik Szczerba <span dir="ltr"><<a href="mailto:dominik@itis.ethz.ch">dominik@itis.ethz.ch</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<div id=":8a3">Yes, the system is badly conditioned. But am I right with my<br>
(experimental) finding that the ultimate way to go is -ksp_type<br>
preonly? Will it do what I want: just solve my system exactly once<br>
using MUMPS and not Krylov?</div></blockquote></div><br><div>Direct solvers are not immune to numerical stability issues. -ksp_type preonly does not check whether the system has been solved, it just uses whatever the preconditioner (direct solve in this case) returned. Putting it inside a Krylov method generally makes it more robust. If the residual is not actually small, you should try different MUMPS flags (consult their user's manual) or try a different solver.</div>