On Tue, 10 Jan 2017, Mark Adams wrote:

> On Tue, Jan 10, 2017 at 4:28 PM, Satish Balay <[email protected]> wrote:
> 
> > Yup. Thats 4294967296*8bytes  34GB RAM - just for this one hash table?
> >

Got this wrong by a factor of 2.

  PetscInt *keytable;
  PetscInt *table;

So thats 4294967296 *2*8bytes (for 64bitindices) = 69GB

> 
> Humm, that is crazy large.
> 
> Treb: I would like to get the stack trace. Can you please send the whole
> output or grep on "[5]" (or any proc that fails) to get a stack trace. And
> can you tell me how many global equations you have here.

Must be the same place Fande encountered this error.

MatSetUpMultiply_MPIAIJ() does:

  ierr = 
PetscTableCreate(aij->B->rmap->n,mat->cmap->N+1,&gid1_lid1);CHKERRQ(ierr);

So the local matrix size (n) must be: 4294967296

[assuming atleast one nonzero per row - the local matrix itself must use
more memory than the hash-table..]

Satish

Reply via email to