Having docs is great, but if we can get away with zero-config - that should be done for sure. While on the 0-config point: when dealing with an normal RDBMS one can browse the schema and list the tables in it. In Ignite, the cache name (ie table name) should be explicitly listed. While I think I understand the reason behind this, the approach limits the JDBC client functionality a lot.
As an example: in Zeppelin, the interpreter will have to configured to work with just a single cache, meaning that it is impossible to have two different notebooks for two separate caches. Is there something that could be done, or we are limited by some architectural or historical properties? Thanks, Cos On Thu, Dec 10, 2015 at 11:29AM, Dmitriy Setrakyan wrote: > Andrey, > > I agree on both points. However, I still think that we should provide > sample step-by-step guide with a sample configuration. This configuration > should include a couple of data types, like Person and Organization, define > indexes for them, etc. > > This guide should include: > > 1. Start Ignite Servers > 2. Populate data - should have example with configuration. > 3. Query data using JDBC driver - example code and, if needed, > configuration to support it. > > Do you mind adding it? > > D. > > On Thu, Dec 10, 2015 at 7:50 AM, Andrey Gura <ag...@gridgain.com> wrote: > > > I didn't think that we should force client mode at moment of > > implementation. But I can't provide any reason not to do it. > > > > So we need two small improvements to JDBC driver: > > > > 1. Always force client mode. > > 2. Use default configuration in cases when user did't specify any URL. This > > configuration should just use client mode and default values for rest > > properties. > > > > Right? > > > > On Thu, Dec 10, 2015 at 9:05 AM, Dmitriy Setrakyan <dsetrak...@apache.org> > > wrote: > > > > > On Wed, Dec 9, 2015 at 10:04 PM, Valentin Kulichenko < > > > valentin.kuliche...@gmail.com> wrote: > > > > > > > Actually, why don't we set it automatically? It doesn't make much sense > > > to > > > > have a server node in a JDBC driver. > > > > > > > > > > Completely agree. > > > > > > > > > > > > > > -Val > > > > > > > > On Wed, Dec 9, 2015 at 10:02 PM, Dmitriy Setrakyan < > > > dsetrak...@apache.org> > > > > wrote: > > > > > > > > > On Wed, Dec 9, 2015 at 6:46 PM, Konstantin Boudnik <c...@apache.org> > > > > wrote: > > > > > > > > > > > On Wed, Dec 09, 2015 at 04:01PM, Dmitriy Setrakyan wrote: > > > > > > > Thanks Andrey! > > > > > > > > > > > > > > Why are we enabling peer class loading for the JDBC driver in > > your > > > > > > example? > > > > > > > To my knowledge, we are not deploying any classes, are we? Also, > > > the > > > > > TCP > > > > > > > discovery with multicast is the default, so no reason to specify > > it > > > > > > either. > > > > > > > > > > > > > > It seems that the only properly required is clientMode=true, no? > > > > > > > > > > > > If this is the case, then _no_ config is really needed, perhaps? > > > > > > > > > > > > > > > > Well, we still need to specify clientMode, no? > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > D. > > > > > > > > > > > > > > On Wed, Dec 9, 2015 at 8:54 AM, Andrey Gura <ag...@gridgain.com> > > > > > wrote: > > > > > > > > > > > > > > > Configuration sample added. > > > > > > > > > > > > > > > > On Tue, Dec 8, 2015 at 10:20 PM, Dmitriy Setrakyan < > > > > > > dsetrak...@apache.org> > > > > > > > > wrote: > > > > > > > > > > > > > > > > > Thanks Andrey! > > > > > > > > > > > > > > > > > > In my view, we should also provide a sample XML configuration > > > > file, > > > > > > > > > especially given that we refer to it in the sample URL. > > > > > > > > > > > > > > > > > > D. > > > > > > > > > > > > > > > > > > On Tue, Dec 8, 2015 at 10:17 AM, Andrey Gura < > > > ag...@gridgain.com > > > > > > > > > > > wrote: > > > > > > > > > > > > > > > > > > > Done. I hope that is more clear now. > > > > > > > > > > > > > > > > > > > > On Tue, Dec 8, 2015 at 1:14 AM, Dmitriy Setrakyan < > > > > > > > > dsetrak...@apache.org > > > > > > > > > > > > > > > > > > > > wrote: > > > > > > > > > > > > > > > > > > > > > On Mon, Dec 7, 2015 at 10:17 AM, Andrey Gura < > > > > > ag...@gridgain.com > > > > > > > > > > > > > > > > wrote: > > > > > > > > > > > > > > > > > > > > > > > Cos, > > > > > > > > > > > > > > > > > > > > > > > > <config_url> sepcifies configuration file for Ignite > > > client > > > > > > node > > > > > > > > that > > > > > > > > > > > will > > > > > > > > > > > > be started during connection establishing by JDBC > > driver. > > > > So > > > > > > this > > > > > > > > > file > > > > > > > > > > > > should be available for JDBC driver client. > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > Andrey, should this be documented here? > > > > > > > > > > > https://apacheignite.readme.io/docs/jdbc-driver > > > > > > > > > > > > > > > > > > > > > > D. > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > On Mon, Dec 7, 2015 at 8:59 PM, Konstantin Boudnik < > > > > > > c...@apache.org > > > > > > > > > > > > > > > > > > > > wrote: > > > > > > > > > > > > > > > > > > > > > > > > > Thanks Andrey! > > > > > > > > > > > > > > > > > > > > > > > > > > Still a bit unclear: the doc says > > > > > > > > > > > > > "<config_url> is required and represents any valid > > URL > > > > > which > > > > > > > > points > > > > > > > > > > to > > > > > > > > > > > > > Ignite configuration file" > > > > > > > > > > > > > > > > > > > > > > > > > > So, that's server side URL then, something like > > > > > > > > > > > > > > > > jdbc:ignite:cfg:///etc/ignite/conf/default-config.xml > > > > > > > > > > > > > > > > > > > > > > > > > > assuming that server nodes have said configuration > > > under > > > > > > > > > > > > > /etc/ignite/conf/default-config.xml ? Sorry, it's > > > either > > > > I > > > > > > am a > > > > > > > > bit > > > > > > > > > > > > dense, > > > > > > > > > > > > > or > > > > > > > > > > > > > the doc isn't written to a layman > > > > > > > > > > > > > > > > > > > > > > > > > > Thanks in advance! > > > > > > > > > > > > > Cos > > > > > > > > > > > > > > > > > > > > > > > > > > On Mon, Dec 07, 2015 at 01:03PM, Andrey Gura wrote: > > > > > > > > > > > > > > Cos, > > > > > > > > > > > > > > > > > > > > > > > > > > > > JDBC driver was reworked in order to use Ignite > > > client > > > > > node > > > > > > > > > instead > > > > > > > > > > > of > > > > > > > > > > > > > > Ignite Java client. It gives better performance. So > > > now > > > > > > JDBC > > > > > > > > > dirver > > > > > > > > > > > > uses > > > > > > > > > > > > > > Ignite xml configuration file using jdbc:ignite:cfg > > > > > > protocol. > > > > > > > > See > > > > > > > > > > > > > > documentation for version 1.4 > > > > > > > > > > > > > > > > https://apacheignite.readme.io/v1.4/docs/jdbc-driver > > > > > > > > > > > > > > > > > > > > > > > > > > > > At the same time we still support old version of > > JDBC > > > > > > driver > > > > > > > > that > > > > > > > > > > > uses > > > > > > > > > > > > > > jdbc:ignite protocol. See documentation for version > > > 1.3 > > > > > > > > > > > > > > > > https://apacheignite.readme.io/v1.3/docs/jdbc-driver > > > > > > > > > > > > > > > > > > > > > > > > > > > > Thus you can control used JDBC driver type (new or > > > old) > > > > > > just > > > > > > > > with > > > > > > > > > > > > choosen > > > > > > > > > > > > > > protocol. > > > > > > > > > > > > > > > > > > > > > > > > > > > > Please, note, that old JDBC driver is deprecated > > and > > > > can > > > > > be > > > > > > > > > removed > > > > > > > > > > > > from > > > > > > > > > > > > > > future releases. > > > > > > > > > > > > > > > > > > > > > > > > > > > > On Mon, Dec 7, 2015 at 6:41 AM, Konstantin Boudnik > > < > > > > > > > > > c...@apache.org > > > > > > > > > > > > > > > > > > > > > > > > wrote: > > > > > > > > > > > > > > > > > > > > > > > > > > > > > Guys, I am looking into [1] and see a reference > > to > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > ignite.jdbc.url = "jdbc:ignite:cfg://" > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > So, I am curious if there's a way to fetch node's > > > > > config > > > > > > via > > > > > > > > > JDBC > > > > > > > > > > > > > somehow? > > > > > > > > > > > > > > > Or > > > > > > > > > > > > > > > this is just a simple error in the doc, which > > needs > > > > to > > > > > be > > > > > > > > > fixed? > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > Would appreciate the insight from the team. > > Thanks! > > > > > > > > > > > > > > > Cos > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > [1] > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > https://zeppelin.incubator.apache.org/docs/0.5.5-incubating/interpreter/ignite.html > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > > > > > > > > > > > > Andrey Gura > > > > > > > > > > > > > > GridGain Systems, Inc. > > > > > > > > > > > > > > www.gridgain.com > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > > > > > > > > > > Andrey Gura > > > > > > > > > > > > GridGain Systems, Inc. > > > > > > > > > > > > www.gridgain.com > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > > > > > > > > Andrey Gura > > > > > > > > > > GridGain Systems, Inc. > > > > > > > > > > www.gridgain.com > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > > > > > > Andrey Gura > > > > > > > > GridGain Systems, Inc. > > > > > > > > www.gridgain.com > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > Andrey Gura > > GridGain Systems, Inc. > > www.gridgain.com > >
signature.asc
Description: Digital signature
