Be more precise about in-memory sorting limit

[drmirror]

I asked kernel engineering what exactly the size limit for in-memory sort was, and it turns out it's the size of all sorted documents combined.

[tychoish]

I think "the sort operation uses x megs" and "the size of all documents plus overhead" are equivalently precise.

[drmirror]

Beg to differ. My customer (and myself) wondered whether it was only the
sum of all sort keys that counted. That's quite a logical idea and would
make a vast difference. Talking to Hari, I believe that the overhead is
small, so we could make it a little more precise by saying "the size of all
documents plus a small overhead" :)

On Tue, Sep 23, 2014 at 10:09 AM, Sam Kleinman [email protected]
wrote:

I think "the sort operation uses x megs" and "the size of all documents
plus overhead" are equivalently precise.

—
Reply to this email directly or view it on GitHub
#2010 (comment).

[tychoish]

merged with an edit as discussed.