You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@ignite.apache.org by Saikat Maitra <sa...@gmail.com> on 2016/06/19 15:42:23 UTC
IGNITE-3337
Hello,
I have raised the PR[1] for the following Jira ticket[2].
Please review and let me know any feedback
Regards
Saikat
[1] https://github.com/apache/ignite/pull/816
[2] https://issues.apache.org/jira/browse/IGNITE-3337
Re: IGNITE-3337
Posted by Saikat Maitra <sa...@gmail.com>.
Hi Alexey , Sergey
Thank you for your time and reviewing the PR. I agree we can deprecate
cacheName for metadata command. I have updated the files and request you to
review the changes.
Also can you please share where the docs are hosted and I can then update
the rest api doc.
Regards
Saikat
On Mon, Jun 20, 2016 at 3:08 PM, Sergey Kozlov <sk...@gridgain.com> wrote:
> Hi Saikat and Alexey.
>
> I think that the solution suggested by Alexey is simple one. Actually
> null-named cache is widely used for now and we can't remove it (or at least
> we should re-think the approach for such caches). On the other hand the
> size of metadata returned by REST command even for dozen caches is not
> large and can be filtered (iterated) on the client side. So I suppose the
> ticket should be updated according suggested approach.
>
>
>
>
> On Mon, Jun 20, 2016 at 12:28 PM, Alexey Kuznetsov <
> akuznetsov@gridgain.com>
> wrote:
>
> > Hi, Saikat
> >
> > I reviewed you PR and I think we should deprecate "cacheName" parameter
> > in metadata command.
> >
> > In current implementation when "cacheName" was not specified that means
> to
> > get "default" cache (with name=null)
> > And now it will be impossible to take metadata for such cache.
> >
> > But I think for this command there is a little sense to extract metadata
> > for single cache,
> > because any way on server side they will be extracted for all caches.
> > See:
> >
> org.apache.ignite.internal.processors.cache.query.GridCacheQueryManager#sqlMetadata
> > line 999.
> >
> > Also for performance reason it is much faster to get all metadata at
> once.
> >
> > So, I suggest - to deprecate "cacheName" parameter and fix documentation.
> >
> > Thoughts?
> >
> >
> > On Sun, Jun 19, 2016 at 10:42 PM, Saikat Maitra <saikat.maitra@gmail.com
> >
> > wrote:
> >
> > > Hello,
> > >
> > > I have raised the PR[1] for the following Jira ticket[2].
> > >
> > > Please review and let me know any feedback
> > >
> > > Regards
> > > Saikat
> > >
> > > [1] https://github.com/apache/ignite/pull/816
> > > [2] https://issues.apache.org/jira/browse/IGNITE-3337
> > >
> >
> >
> >
> > --
> > Alexey Kuznetsov
> > GridGain Systems
> > www.gridgain.com
> >
>
>
>
> --
> Sergey Kozlov
> GridGain Systems
> www.gridgain.com
>
Re: IGNITE-3337
Posted by Sergey Kozlov <sk...@gridgain.com>.
Hi Saikat and Alexey.
I think that the solution suggested by Alexey is simple one. Actually
null-named cache is widely used for now and we can't remove it (or at least
we should re-think the approach for such caches). On the other hand the
size of metadata returned by REST command even for dozen caches is not
large and can be filtered (iterated) on the client side. So I suppose the
ticket should be updated according suggested approach.
On Mon, Jun 20, 2016 at 12:28 PM, Alexey Kuznetsov <ak...@gridgain.com>
wrote:
> Hi, Saikat
>
> I reviewed you PR and I think we should deprecate "cacheName" parameter
> in metadata command.
>
> In current implementation when "cacheName" was not specified that means to
> get "default" cache (with name=null)
> And now it will be impossible to take metadata for such cache.
>
> But I think for this command there is a little sense to extract metadata
> for single cache,
> because any way on server side they will be extracted for all caches.
> See:
> org.apache.ignite.internal.processors.cache.query.GridCacheQueryManager#sqlMetadata
> line 999.
>
> Also for performance reason it is much faster to get all metadata at once.
>
> So, I suggest - to deprecate "cacheName" parameter and fix documentation.
>
> Thoughts?
>
>
> On Sun, Jun 19, 2016 at 10:42 PM, Saikat Maitra <sa...@gmail.com>
> wrote:
>
> > Hello,
> >
> > I have raised the PR[1] for the following Jira ticket[2].
> >
> > Please review and let me know any feedback
> >
> > Regards
> > Saikat
> >
> > [1] https://github.com/apache/ignite/pull/816
> > [2] https://issues.apache.org/jira/browse/IGNITE-3337
> >
>
>
>
> --
> Alexey Kuznetsov
> GridGain Systems
> www.gridgain.com
>
--
Sergey Kozlov
GridGain Systems
www.gridgain.com
Re: IGNITE-3337
Posted by Alexey Kuznetsov <ak...@gridgain.com>.
Hi, Saikat
I reviewed you PR and I think we should deprecate "cacheName" parameter
in metadata command.
In current implementation when "cacheName" was not specified that means to
get "default" cache (with name=null)
And now it will be impossible to take metadata for such cache.
But I think for this command there is a little sense to extract metadata
for single cache,
because any way on server side they will be extracted for all caches.
See: org.apache.ignite.internal.processors.cache.query.GridCacheQueryManager#sqlMetadata
line 999.
Also for performance reason it is much faster to get all metadata at once.
So, I suggest - to deprecate "cacheName" parameter and fix documentation.
Thoughts?
On Sun, Jun 19, 2016 at 10:42 PM, Saikat Maitra <sa...@gmail.com>
wrote:
> Hello,
>
> I have raised the PR[1] for the following Jira ticket[2].
>
> Please review and let me know any feedback
>
> Regards
> Saikat
>
> [1] https://github.com/apache/ignite/pull/816
> [2] https://issues.apache.org/jira/browse/IGNITE-3337
>
--
Alexey Kuznetsov
GridGain Systems
www.gridgain.com