Brain,
I'm able to get only 10 results from the mutliFetch though I'm passing more
number of keys. I'm getting the keys from Riak Search and Building
locaitons from those keys, setting
List<Location> locationList = new ArrayList<Location>();
for(String key : keys){
locationList.add(new Location(bucketNamespace, key));
}
MultiFetch multiFetch = new MultiFetch.Builder().addLocations(locationList
).build();
MultiFetch.Response multi_responses = client.execute(multiFetch);
for (RiakFuture<FetchValue.Response, Location> f :
accessEntriesResponses.getResponses())
{
FetchValue.Response response = f.get();
.....
Adding to list
}
At the end when I see I'm getting only 10 objects.
On Tue, Jan 20, 2015 at 10:43 PM, Brian Roach <ro...@basho.com> wrote:
> Santi -
>
> The core of the 2.0 Java client uses Netty. There's a fixed number of
> worker threads that process sockets using non-blocking IO ( w/
> polling/select).
>
> Due to the synchronous nature of the Riak API, each fetch operation
> requires its own socket connection and we can't pipeline.
>
> As noted in the Javadoc for MultiFetch
> (
> http://basho.github.io/riak-java-client/2.0.0/com/basho/riak/client/api/commands/kv/MultiFetch.html
> )
> there is a default of 10 simultaneous fetches in flight at once. This
> is pretty conservative but ... that's what defaults are for :)
>
> How does it scale? Depends on the hardware, network latency, etc. In
> the end ... you've only got so many threads and you prob don't want
> 1000 sockets being created at once. Best suggestion is to adjust up
> the number of simultaneous inflight and measure it.
>
> Thanks,
> - Roach
>
>
>
> On Mon, Jan 19, 2015 at 6:43 PM, Santi Kumar <sa...@veradocs.com> wrote:
> > Brain
> > I"m using Riak Client 2.0.0 and Riak 2.0.2.
> >
> >
> >
> > On Tue, Jan 20, 2015 at 1:54 AM, Brian Roach <ro...@basho.com> wrote:
> >>
> >> Santi -
> >>
> >> Which version of the Java client?
> >>
> >> Thanks,
> >> - Roach
> >>
> >> On Mon, Jan 19, 2015 at 7:36 AM, Santi Kumar <sa...@veradocs.com>
> wrote:
> >> > Hi
> >> >
> >> > We are using java client for accessing Riak KV/Search. For some use
> >> > cases,
> >> > we go to search, get the keys and access the data from Riak. There
> might
> >> > be
> >> > a case where we might get 1000's of keys. So want to understand what
> is
> >> > the
> >> > impact of that on multiget and how does it scale.
> >> >
> >> >
> >> > We were using RDBMS and Elastic search earlier. Now we moved all the
> >> > data to
> >> > Riak KV and Search. We used to query all the audit entries from ES as
> we
> >> > used to store the complete data there. Applicaitons flows used to get
> >> > the
> >> > data from RDBMS. Now as we replaced everything with Riak KV and
> search,
> >> > We
> >> > need to go to Riak Search for auditing reports. If we store all the
> >> > objects
> >> > in Riak Search, how does it impact ?
> >> >
> >> > Thanks
> >> > Santi
> >> >
> >> > _______________________________________________
> >> > riak-users mailing list
> >> > riak-users@lists.basho.com
> >> > http://lists.basho.com/mailman/listinfo/riak-users_lists.basho.com
> >> >
> >
> >
>
_______________________________________________
riak-users mailing list
riak-users@lists.basho.com
http://lists.basho.com/mailman/listinfo/riak-users_lists.basho.com
