In a system like Cassandra you don't know what you have until you have
it. The compacted thresholds will give you an idea of whether there
are some rows that are significantly larger than others, but finding
an individual row (and thus it's columns) will be difficult.
What sort of query is causing the heap size issues?
On Fri, Sep 28, 2012 at 4:04 PM, Vinay Kumar Chella