[infinispan-dev] Performance of Infinispan-Query Module

[Sanne Grinovero]

On 22 October 2013 17:37, Emmanuel Bernard <emmanuel at hibernate.org> wrote:
> It's more how random your dataset is. If you're indexing the sex
> property of a user, your index is non discriminant.
> If you're indexing by (real) names, it is very efficient, unless youa re
> in Smurf land.
>
> In all cases, generating data out of say 10 fake names, leads to
> inefficient indexes because the dataset is non discriminant.
>
> Emmanuel
>
> On Tue 2013-10-22 12:29, Shane Johnson wrote:
>> Perhaps map / reduce is a better option for smaller data sets?

That's probably the simplest solution for experimentation on a small
data set, especially if you don't need any of the more advanced
free-text functions like term vectors and frequencies, but if this is
a small test aimed at designing a larger scale system you would
probably need Query or a combination of Map/Reduce with Query.

Sanne

>>
>> Shane
>>
>> ----- Original Message -----
>> From: "Faseela K" <faseela.k at ericsson.com>
>> To: "infinispan -Dev List" <infinispan-dev at lists.jboss.org>
>> Sent: Tuesday, October 22, 2013 9:32:26 AM
>> Subject: Re: [infinispan-dev] Performance of Infinispan-Query Module
>>
>> Hi,
>>
>>   I just tried with 10MB data in the cache.
>>   Could you please let me know if there is any lower limit for the cache size, for which Query will show some better performance.
>>
>>   For iteration, I did cache.keySet(), and then a get for each key.
>>   I hope that will work in clustered case also.
>>
>> Thanks,
>> Faseela
>>
>> -----Original Message-----
>> From: infinispan-dev-bounces at lists.jboss.org [mailto:infinispan-dev-bounces at lists.jboss.org] On Behalf Of Sanne Grinovero
>> Sent: Tuesday, October 22, 2013 7:25 PM
>> To: infinispan -Dev List
>> Subject: Re: [infinispan-dev] Performance of Infinispan-Query Module
>>
>> Hi Faseela,
>> did you run a large scale test? there is a constant cost for the Query creation and index engine initialization; if the number of entries to iterate through is very small then there is no advantage in running a Query.
>>
>> Also, to keep in mind that you can't use iterate() on a distributed or replicated cache as it won't return results which are not stored locally.
>>
>> Sanne
>>
>>
>> On 22 October 2013 14:36, Faseela K <faseela.k at ericsson.com> wrote:
>> > Hi,
>> >
>> >    I tried a simple performance test for infinispan-query module using
>> > the example mentioned in :
>> >
>> >
>> > https://docs.jboss.org/author/display/ISPN/Querying+Infinispan
>> >
>> >    What I see is, the querying is much slower than iterating through
>> > the cache entries and checking for a match.
>> >    Is this behaviour correct, or am I missing some configuration?
>> >    My infinispan configuration is :
>> >
>> >   <infinispan
>> >         xsi:schemaLocation="urn:infinispan:config:5.2
>> > http://www.infinispan.org/schemas/infinispan-config-5.2.xsd"
>> >         xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
>> > xmlns="urn:infinispan:config:5.2">
>> >         <default>
>> >                 <indexing enabled="true" indexLocalOnly="true">
>> >                         <properties>
>> >                                 <property name="default.directory_provider"
>> > value="ram" />
>> >                         </properties>
>> >                 </indexing>
>> >         </default>
>> > </infinispan>
>> >
>> > Thanks,
>> > Faseela
>> >
>> >
>> >
>> > _______________________________________________
>> > infinispan-dev mailing list
>> > infinispan-dev at lists.jboss.org
>> > https://lists.jboss.org/mailman/listinfo/infinispan-dev
>> _______________________________________________
>> infinispan-dev mailing list
>> infinispan-dev at lists.jboss.org
>> https://lists.jboss.org/mailman/listinfo/infinispan-dev
>>
>> _______________________________________________
>> infinispan-dev mailing list
>> infinispan-dev at lists.jboss.org
>> https://lists.jboss.org/mailman/listinfo/infinispan-dev
>> _______________________________________________
>> infinispan-dev mailing list
>> infinispan-dev at lists.jboss.org
>> https://lists.jboss.org/mailman/listinfo/infinispan-dev
> _______________________________________________
> infinispan-dev mailing list
> infinispan-dev at lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/infinispan-dev