On Tue, Aug 6, 2019 at 12:11 PM Matt Sicker <[email protected]> wrote:
> If the batch size is set to 1 (or 0), shouldn't that just disable > batching entirely? I don't see the point of creating and executing > batched statements if you only use it once. > Good point, so there is also a bug in the JDBC Appender where bufferSize="0" still uses a JDBC batch. I'll fix that... Gary > > On Tue, 6 Aug 2019 at 09:40, Gary Gregory <[email protected]> wrote: > > > > On Tue, Aug 6, 2019 at 10:39 AM Gary Gregory <[email protected]> > wrote: > > > > > On Mon, Aug 5, 2019 at 11:00 AM Gary Gregory <[email protected]> > > > wrote: > > > > > >> FYI: I plan on looking at > > >> https://issues.apache.org/jira/browse/LOG4J2-2667 today/tonight. > > >> > > > > > > Please see my comments in the Jira above WRT batching. The upshot is > that > > > to workaround buggy JDBC drivers (My tests show no issues using H2, > MySQL, > > > and MS-SQL), is that we should provide a setting to NOT use JDBC SQL > > > batches even if the JDBC driver supports them such that the user can > say > > > batchPolicy="always|never|driver" in a JDBC Appender. > > > > > > > where the default is "driver" to match what we do today. > > > > Gary > > > > > > > > > > If that works for you all, I can implement this before we release > 2.12.1. > > > > > > Gary > > > > > > > > >> Gary > > >> > > >> On Mon, Aug 5, 2019 at 10:48 AM Ralph Goers < > [email protected]> > > >> wrote: > > >> > > >>> I fixed the performance issue but I keep getting a test failure in > > >>> SocketReconnectTest on my build machine. I don’t see it on my newer > MacBook > > >>> Pro. > > >>> > > >>> Ralph > > >>> > > >>> > On Aug 4, 2019, at 3:14 PM, Ralph Goers < > [email protected]> > > >>> wrote: > > >>> > > > >>> > I am running a compile before performing the release and am seeing > > >>> significant negative performance changes in the core integration > tests > > >>> module. I am going to have to investigate those before performing a > release. > > >>> > > > >>> > Ralph > > >>> > > > >>> >> On Aug 4, 2019, at 8:26 AM, Matt Sicker <[email protected]> wrote: > > >>> >> > > >>> >> Sounds like a plan! > > >>> >> > > >>> >> On Sun, Aug 4, 2019 at 10:13, Carter Kozak <[email protected]> > wrote: > > >>> >> > > >>> >>> Thanks Ralph! > > >>> >>> > > >>> >>> -ck > > >>> >>> > > >>> >>> On Sun, Aug 4, 2019, at 06:56, Gary Gregory wrote: > > >>> >>>> Sounds good. > > >>> >>>> > > >>> >>>> Gary > > >>> >>>> > > >>> >>>> On Sun, Aug 4, 2019, 03:11 Ralph Goers < > [email protected]> > > >>> >>> wrote: > > >>> >>>> > > >>> >>>>> We have a few commits on the release-2.x branch that are > important > > >>> that > > >>> >>>>> user’s who still use Java 7 would want. So I am planning on > > >>> performing > > >>> >>> a > > >>> >>>>> 2.12.1 release later today. I will then make the changes to the > > >>> pom to > > >>> >>>>> target it to Java 8 and start bringing in some Java 8 specific > code > > >>> >>> from > > >>> >>>>> master. > > >>> >>>>> > > >>> >>>>> Ralph > > >>> >>>>> > > >>> >>>> > > >>> >>> > > >>> >> -- > > >>> >> Matt Sicker <[email protected]> > > >>> > > > >>> > > > >>> > > > >>> > > >>> > > >>> > > > > -- > Matt Sicker <[email protected]> >
