Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Scott Carey <scott@richrelevance.com> writes:
>> Gzip does have some quirky performance behavior depending on the
>> chunk size of data you stream into it.
>
> Can you enlarge on that comment? I'm not sure that pg_dump is aware
> that there's anything to worry about there.
If the library used here is anything like the native library used by
Java, it'd be worth putting a buffer layer ahead of the calls to gzip,
so it isn't dealing with each individual value as a separate call. I
seem to remember running into that issue in Java, where throwing a
BufferedOutputStream in there fixed the performance issue.
-Kevin