On Thu, 16 Oct 2008, Achilleas Mantzios wrote:
> i am experiencing the following situation with 8.3.3 backend and
> postgresql-8.3-603.jdbc3.jar driver. In a simple program such as shown
> below i try to execute a rather large SQL worth of 2.5 MBytes:
> i noticed that when i run the above program against postgresql 7.4.2
> (with the contrib stock jdbc driver) the above program executes and ends
> successfully. However when i try to run it against 8.3.3 server and
> postgresql-8.3-603.jdbc3.jar driver, it seems to stall at a statement
> which is located somewhere around 1.5 MByte offset. If i split the SQL
> file into two parts 1.5Mbytes and 1Mbyte respectively then the above
> program runs fine. Experimenting i found out that a file of 1623377
> bytes would rather be executed (i think it is not deterministic when it
> comes to file sizes), while a file of 1635008 would stall. Can somebody
> think of anything about this (change of) behaviour? Any way to by pass
> this in 8.3 without resorting to drastically change one's architecture?
See this report on -bugs which describes the problem in detail.
I haven't looked into a fix, but this is what's going on.
In response to
pgsql-jdbc by date
|Next:||From: Achilleas Mantzios||Date: 2008-10-17 07:04:09|
|Subject: Re: issues with SQL size st.execute(SQL) in 8.3.3|
|Previous:||From: Achilleas Mantzios||Date: 2008-10-16 14:17:16|
|Subject: issues with SQL size st.execute(SQL) in 8.3.3|