Arithmetic overflow

[Racerx]

Hi All
I m trying to use a query that would calculate the table size in MB s
rather than pages

The query is as follows
db2 "select rtrim(t.TABSCHEMA) || '.' || t.tabname ,npages,pagesize,
dec((float(( npages*pagesize )) / float((1024*1024))) ,8,3)
"Table_size_in_MBs" from syscat.tables t,syscat.tablespaces ts where
ts.TBSPACEID=t.TBSPACEID and t.TYPE ='T' and npages not in ( -1, 0)
order by Table_size_in_MBs DESC"

But it gives me a arithmetic overflow error when i run it on a AIX
box ...it seems to run properly on Windows machines...
Can some advice me where am i going wrong????
Rgards
Racerx....

[Tonkuma]

On Jun 8, 3:03 pm, Racerx <[e-mail]:

Hi All
I m trying to use a query that would calculate the table size in MB s
rather than pages

The query is as follows

db2 "select rtrim(t.TABSCHEMA) || '.' || t.tabname ,npages,pagesize,
dec((float(( npages*pagesize )) / float((1024*1024))) ,8,3)
"Table_size_in_MBs" from syscat.tables t,syscat.tablespaces ts where
ts.TBSPACEID=t.TBSPACEID and t.TYPE ='T' and npages not in ( -1, 0)
order by Table_size_in_MBs DESC"

But it gives me a arithmetic overflow error when i run it on a AIX
box ...it seems to run properly on Windows machines...

Can some advice me where am i going wrong????

Rgards
Racerx....

Although I don't know the reason, moving position of float function
may be a workaround.
db2 "select rtrim(t.TABSCHEMA) || '.' || t.tabname ,npages,pagesize,
dec((( float(npages)*pagesize ) / (float(1024)*1024)) ,8,3)
"Table_size_in_MBs" from syscat.tables t,syscat.tablespaces ts where
ts.TBSPACEID=t.TBSPACEID and t.TYPE ='T' and npages not in ( -1, 0)
order by Table_size_in_MBs DESC"