Out of Memory

[Fran Simon <fs...@sciquest.com>]

Hi All,

I have been digging through the archives looking at all of the out of memory
problems and
suggestions and have seen many solutions, but I am not sure how to do some
of them,
there are no real examples.

I think the best approach is to breakup your page sequences but am not sure
how to go
about it.  I was hoping someone could show me a example.  For instance how
do I
determine the number of rows per page when I have to deal with hyphenation
or wrapping?
I assume I have to take font size into account as well.

Basically what I am doing is generating a FOP document from a XML template.
We use
this for generating reports.  I am not using stylesheets to transform the
xml because the
xml is just a template of where the data comes from and what the report
should look like.

Generally the report is one table with some header information.  The table
could be 2 to
50000 rows depending on the data source.  It made sens to me to make a
single page
sequence which contained a table.  The table will have summary rows which I
write out
at the bottom of the table.

If I generate multiple pages sequences I assume I will have to break it into
many tables
which I can do.  But how do I determine when and where to break it and still
make it
look like a single table???

Any examples or information would be really helpful!!!!!!!
Thanks

Fran Simon
Sciquest, Inc.

RE: Out of Memory

[Victor Mote <vi...@outfitr.com>]

Out of MemoryFran Simon wrote
  I think the best approach is to breakup your page sequences but am not
sure how to go
  about it.  I was hoping someone could show me a example.  For instance how
do I
  determine the number of rows per page when I have to deal with hyphenation
or wrapping?
  I assume I have to take font size into account as well.

  Basically what I am doing is generating a FOP document from a XML
template.  We use
  this for generating reports.  I am not using stylesheets to transform the
xml because the
  xml is just a template of where the data comes from and what the report
should look like.

  Generally the report is one table with some header information.  The table
could be 2 to
  50000 rows depending on the data source.  It made sens to me to make a
single page
  sequence which contained a table.  The table will have summary rows which
I write out
  at the bottom of the table.

  If I generate multiple pages sequences I assume I will have to break it
into many tables
  which I can do.  But how do I determine when and where to break it and
still make it
  look like a single table???

  Any examples or information would be really helpful!!!!!!!
  Thanks

  Fran Simon
  Sciquest, Inc.

RE: Out of Memory

[Victor Mote <vi...@outfitr.com>]

Fran Simon wrote:

<--- Start --->
I think the best approach is to breakup your page sequences but am not sure
how to go about it.  I was hoping someone could show me a example.  For
instance how do I
determine the number of rows per page when I have to deal with hyphenation
or wrapping?

I assume I have to take font size into account as well.
Basically what I am doing is generating a FOP document from a XML template.
We use
this for generating reports.  I am not using stylesheets to transform the
xml because the
xml is just a template of where the data comes from and what the report
should look like.
Generally the report is one table with some header information.  The table
could be 2 to
50000 rows depending on the data source.  It made sens to me to make a
single page
sequence which contained a table.  The table will have summary rows which I
write out
at the bottom of the table.
If I generate multiple pages sequences I assume I will have to break it into
many tables
which I can do.  But how do I determine when and where to break it and still
make it
look like a single table???

<--- End --->

Breaking into smaller page sequences is often the best way to avoid memory
problems. However, in your case, that does not sound possible without
arbitrary page breaks in your data, which would seem, from your description,
to be unwanted. In your case, I would recommend using one of the other
approaches, such as making more memory available to the JVM.

Victor Mote