You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@marmotta.apache.org by Jakob Frank <ja...@salzburgresearch.at> on 2013/10/08 11:53:22 UTC
Re: [lmf-users] LDCache
[ cross-posting to users@marmotta.incubator.apache.org as LDCache has
moved to Apache Marmotta ]
Hi Ivan,
did you try explicitly requesting a resource from LDCache, e.g. by
querying to "live" webservice?
e.g.
<http://localhost:8080/LMF/cache/live?uri=http://dbpedia.org/resource/Berlin>
Plus: which version are you currently using?
Best,
Jakob
On Tue, 2013-10-08 at 02:26 -0700, Ivan Ruiz Rube wrote:
> Dear LMF Users,
>
> I have a problem using the cache capabilities of LMF. I can not retrieve
> resources from the external endpoints configurated in my LMF installation,
> neither with the DBPedia SPARQL sample endpoint.
>
> The property ldcache.enabled is activated and I have found interesting
> nothing about it in the logs marmotta-main, catalina and access-log.
>
> Please, any idea?
>
> Thanks in advance.
>
> Regards.
>
>
>
--
DI Jakob Frank
Knowledge and Media Technologies
Salzburg Research Forschungsgesellschaft mbH
Jakob Haringer-Strasse 5/3 | 5020 Salzburg, Austria
T: +43.662.2288-419
jakob.frank@salzburgresearch.at
http://www.salzburgresearch.at
Re: [lmf-users] LDCache
Posted by Ivan Ruiz Rube <ru...@gmail.com>.
ok.
I will trigger the caching using a LDPath such that, *owl:sameAs / * / * /
* / * :: xsd:string*
:-)
Thank you for your attention.
Regards
Iván
On Mon, Oct 21, 2013 at 11:08 AM, Sergio Fernández <
sergio.fernandez@salzburgresearch.at> wrote:
> Hi Ivan,
>
>
> On 21/10/13 09:37, Ivan Ruiz Rube wrote:
>
>> Then, I requested this new query:
>>
>> SELECT DISTINCT ?externalResource ?name
>> WHERE
>> { <http://localhost:8080/LMF/**resource/380014300<http://localhost:8080/LMF/resource/380014300>>
>> owl:sameAs
>> ?externalResource .
>> ?externalResource rdfs:label ?name
>> }
>>
>> obtaining no results.
>>
>
> OK. Please, read bellow further explanation.
>
> There are no error messages in the log files. In a next step, I asked for
>> the following LDPath query
>>
>> ...
>>
>>
>> Meanwhile marmotta-main.log throws the following proper message:
>>
>> 09:06:04.306 INFO o.a.m.l.s.p.**AbstractHttpProvider - retrieving
>> resource
>> data for http://spi-fm.uca.es/d2r4ea/**resource/projects/1377<http://spi-fm.uca.es/d2r4ea/resource/projects/1377>from 'SPARQL'
>> endpoint, request URI is <
>> http://spi-fm.uca.es/d2r4ea/**sparql?query=SELECT+%3Fp+%3Fo+**
>> WHERE+%7B+%3Chttp%3A%2F%2Fspi-**fm.uca.es%2Fd2r4ea%2Fresource%**
>> 2Fprojects%2F1377%3E+%3Fp+%**3Fo+%7D<http://spi-fm.uca.es/d2r4ea/sparql?query=SELECT+%3Fp+%3Fo+WHERE+%7B+%3Chttp%3A%2F%2Fspi-fm.uca.es%2Fd2r4ea%2Fresource%2Fprojects%2F1377%3E+%3Fp+%3Fo+%7D>
>>
>>>
>>> 09:06:04.478 INFO o.a.m.l.s.p.**AbstractHttpProvider - retrieved 10
>> triples
>> for resource http://spi-fm.uca.es/d2r4ea/**resource/projects/1377<http://spi-fm.uca.es/d2r4ea/resource/projects/1377>;
>> expiry
>> date: Mon Oct 21 09:06:14 CEST 2013
>>
>>
>> Finally, I requested the same query
>>
>> SELECT DISTINCT ?externalResource ?name
>> WHERE
>> { <http://localhost:8080/LMF/**resource/380014300<http://localhost:8080/LMF/resource/380014300>>
>> owl:sameAs
>> ?externalResource .
>> ?externalResource rdfs:label ?name
>> }
>>
>> obtaining in this time, the expected results:
>>
>> externalResource name
>> ----------------------
>> -------
>> <http://spi-fm.uca.es/d2r4ea/**resource/projects/1377<http://spi-fm.uca.es/d2r4ea/resource/projects/1377>>
>> "Proyecto MPS-G3"
>>
>
> That's expected, because the cache previously retrieved the resource, and
> store it in a specific named graph, so it's completely transparent to the
> SPARQL query evaluation
>
> You can see the cache resource, for instance, by browsing it:
>
> <http://localhost:8080/LMF/**resource?uri=http://spi-fm.**
> uca.es/d2r4ea/resource/**projects/1377<http://localhost:8080/LMF/resource?uri=http://spi-fm.uca.es/d2r4ea/resource/projects/1377>
>
>
> I am using SESAME strategy, but it seems that the remote resource
>> retrieval
>> is only triggered with LDPath queries, but not with SPARQL.
>>
>
> I've just have a discussion with Sebastian about that. And now I'd
> rephrase what we initially said: "by switching to the SPARQL
> Sesame-strategy, Marmotta could be able to lively cache missing resources
> referenced in query time". Besides the impact on performance we commented,
> the LDCache trigger is not warranted because it depends on how Sesame
> optimizes and evaluates the query.
>
> In short, I'd not recommend to depend on such strategy for querying
> Marmotta. If you need cache, query with LDPath, for the actual query or
> just to trigger the caching and then use SPARQL, as you prefer.
>
> Hope we have clarified the issue. I'll update the documentation at the web
> site to avoid confusions.
>
>
> Cheers,
>
> --
> Sergio Fernández
> Senior Researcher
> Knowledge and Media Technologies
> Salzburg Research Forschungsgesellschaft mbH
> Jakob-Haringer-Straße 5/3 | 5020 Salzburg, Austria
> T: +43 662 2288 318 | M: +43 660 2747 925
> sergio.fernandez@**salzburgresearch.at<se...@salzburgresearch.at>
> http://www.salzburgresearch.at
>
Re: [lmf-users] LDCache
Posted by Sergio Fernández <se...@salzburgresearch.at>.
Hi Ivan,
On 21/10/13 09:37, Ivan Ruiz Rube wrote:
> Then, I requested this new query:
>
> SELECT DISTINCT ?externalResource ?name
> WHERE
> { <http://localhost:8080/LMF/resource/380014300> owl:sameAs
> ?externalResource .
> ?externalResource rdfs:label ?name
> }
>
> obtaining no results.
OK. Please, read bellow further explanation.
> There are no error messages in the log files. In a next step, I asked for
> the following LDPath query
>
>...
>
> Meanwhile marmotta-main.log throws the following proper message:
>
> 09:06:04.306 INFO o.a.m.l.s.p.AbstractHttpProvider - retrieving resource
> data for http://spi-fm.uca.es/d2r4ea/resource/projects/1377 from 'SPARQL'
> endpoint, request URI is <
> http://spi-fm.uca.es/d2r4ea/sparql?query=SELECT+%3Fp+%3Fo+WHERE+%7B+%3Chttp%3A%2F%2Fspi-fm.uca.es%2Fd2r4ea%2Fresource%2Fprojects%2F1377%3E+%3Fp+%3Fo+%7D
>>
> 09:06:04.478 INFO o.a.m.l.s.p.AbstractHttpProvider - retrieved 10 triples
> for resource http://spi-fm.uca.es/d2r4ea/resource/projects/1377; expiry
> date: Mon Oct 21 09:06:14 CEST 2013
>
>
> Finally, I requested the same query
>
> SELECT DISTINCT ?externalResource ?name
> WHERE
> { <http://localhost:8080/LMF/resource/380014300> owl:sameAs
> ?externalResource .
> ?externalResource rdfs:label ?name
> }
>
> obtaining in this time, the expected results:
>
> externalResource name
> ----------------------
> -------
> <http://spi-fm.uca.es/d2r4ea/resource/projects/1377> "Proyecto MPS-G3"
That's expected, because the cache previously retrieved the resource,
and store it in a specific named graph, so it's completely transparent
to the SPARQL query evaluation
You can see the cache resource, for instance, by browsing it:
<http://localhost:8080/LMF/resource?uri=http://spi-fm.uca.es/d2r4ea/resource/projects/1377
> I am using SESAME strategy, but it seems that the remote resource retrieval
> is only triggered with LDPath queries, but not with SPARQL.
I've just have a discussion with Sebastian about that. And now I'd
rephrase what we initially said: "by switching to the SPARQL
Sesame-strategy, Marmotta could be able to lively cache missing
resources referenced in query time". Besides the impact on performance
we commented, the LDCache trigger is not warranted because it depends on
how Sesame optimizes and evaluates the query.
In short, I'd not recommend to depend on such strategy for querying
Marmotta. If you need cache, query with LDPath, for the actual query or
just to trigger the caching and then use SPARQL, as you prefer.
Hope we have clarified the issue. I'll update the documentation at the
web site to avoid confusions.
Cheers,
--
Sergio Fernández
Senior Researcher
Knowledge and Media Technologies
Salzburg Research Forschungsgesellschaft mbH
Jakob-Haringer-Straße 5/3 | 5020 Salzburg, Austria
T: +43 662 2288 318 | M: +43 660 2747 925
sergio.fernandez@salzburgresearch.at
http://www.salzburgresearch.at
Re: [lmf-users] LDCache
Posted by Ivan Ruiz Rube <ru...@gmail.com>.
Hi Sergio,
The example case is the following:
First, I added the triple (subject is local resource and object is a remote
resource:
<http://localhost:8080/LMF/resource/380014300>
owl:sameAs
<http://spi-fm.uca.es/d2r4ea/resource/projects/1377>
Then, I requested the SPARQL query:
SELECT DISTINCT ?externalResource
WHERE
{ <http://localhost:8080/LMF/resource/380014300> owl:sameAs
?externalResource
}
obtaining the expected result:
?externalResource
------------------------
<http://spi-fm.uca.es/d2r4ea/resource/projects/1377>
Then, I requested this new query:
SELECT DISTINCT ?externalResource ?name
WHERE
{ <http://localhost:8080/LMF/resource/380014300> owl:sameAs
?externalResource .
?externalResource rdfs:label ?name
}
obtaining no results.
There are no error messages in the log files. In a next step, I asked for
the following LDPath query
owl:sameAs / rdfs:label :: xsd:string
in context <http://localhost:8080/LMF/resource/380014300>
obtaining these expected results:
Field Results
----------------------------------------------------------------------------
http://www.w3.org/2002/07/owl#sameAs Proyecto MPS-G3
Meanwhile marmotta-main.log throws the following proper message:
09:06:04.306 INFO o.a.m.l.s.p.AbstractHttpProvider - retrieving resource
data for http://spi-fm.uca.es/d2r4ea/resource/projects/1377 from 'SPARQL'
endpoint, request URI is <
http://spi-fm.uca.es/d2r4ea/sparql?query=SELECT+%3Fp+%3Fo+WHERE+%7B+%3Chttp%3A%2F%2Fspi-fm.uca.es%2Fd2r4ea%2Fresource%2Fprojects%2F1377%3E+%3Fp+%3Fo+%7D
>
09:06:04.478 INFO o.a.m.l.s.p.AbstractHttpProvider - retrieved 10 triples
for resource http://spi-fm.uca.es/d2r4ea/resource/projects/1377; expiry
date: Mon Oct 21 09:06:14 CEST 2013
Finally, I requested the same query
SELECT DISTINCT ?externalResource ?name
WHERE
{ <http://localhost:8080/LMF/resource/380014300> owl:sameAs
?externalResource .
?externalResource rdfs:label ?name
}
obtaining in this time, the expected results:
externalResource name
----------------------
-------
<http://spi-fm.uca.es/d2r4ea/resource/projects/1377> "Proyecto MPS-G3"
I am using SESAME strategy, but it seems that the remote resource retrieval
is only triggered with LDPath queries, but not with SPARQL.
Thank you very much.
Iván
On Sat, Oct 19, 2013 at 11:13 AM, Sergio Fernández <
sergio.fernandez@salzburgresearch.at> wrote:
> Hi Ivan,
>
>
> On 18/10/13 19:20, Ivan Ruiz Rube wrote:
>
>> I don't finish to understand it well. When you say that SPARQL is able to
>> lively cache missing resources used in queries by switching the SPARQL
>> strategy to "sesame", is this feature implemented? or not? Because, I
>> have
>> just checked using the sesame strategy how the external triples are not
>> available in SPARQL until the LDPath engine doesn't ask for them.
>>
>
> It should. Personally I did no switch to such strategy for quite a long,
> so we may have an issue with it. Would be nice to have some more details
> from your logs as well as the query and remote resource your are missing.
>
>
> It is a shame, because I thought LDCache would offer a behavior similar
>> to the SERVICE extension of SPARQL, but avoiding include the endpoint URI
>> in the query itself.
>>
>
> Well, you could have a similar behaviour, but SPARQL Federated Query works
> over the endpoint, while LDCache works a a resource level. Not in all cases
> both methods are available, and each method has its pros and cons.
>
>
> Cheers,
>
> --
> Sergio Fernández
> Senior Researcher
> Knowledge and Media Technologies
> Salzburg Research Forschungsgesellschaft mbH
> Jakob-Haringer-Straße 5/3 | 5020 Salzburg, Austria
> T: +43 662 2288 318 | M: +43 660 2747 925
> sergio.fernandez@**salzburgresearch.at<se...@salzburgresearch.at>
> http://www.salzburgresearch.at
>
Re: [lmf-users] LDCache
Posted by Sergio Fernández <se...@salzburgresearch.at>.
Hi Ivan,
On 18/10/13 19:20, Ivan Ruiz Rube wrote:
> I don't finish to understand it well. When you say that SPARQL is able to
> lively cache missing resources used in queries by switching the SPARQL
> strategy to "sesame", is this feature implemented? or not? Because, I have
> just checked using the sesame strategy how the external triples are not
> available in SPARQL until the LDPath engine doesn't ask for them.
It should. Personally I did no switch to such strategy for quite a long,
so we may have an issue with it. Would be nice to have some more details
from your logs as well as the query and remote resource your are missing.
> It is a shame, because I thought LDCache would offer a behavior similar
> to the SERVICE extension of SPARQL, but avoiding include the endpoint URI
> in the query itself.
Well, you could have a similar behaviour, but SPARQL Federated Query
works over the endpoint, while LDCache works a a resource level. Not in
all cases both methods are available, and each method has its pros and cons.
Cheers,
--
Sergio Fernández
Senior Researcher
Knowledge and Media Technologies
Salzburg Research Forschungsgesellschaft mbH
Jakob-Haringer-Straße 5/3 | 5020 Salzburg, Austria
T: +43 662 2288 318 | M: +43 660 2747 925
sergio.fernandez@salzburgresearch.at
http://www.salzburgresearch.at
Re: [lmf-users] LDCache
Posted by Ivan Ruiz Rube <ru...@gmail.com>.
Sebastian, Sergio, thank you very much for your explanations.
I don't finish to understand it well. When you say that SPARQL is able to
lively cache missing resources used in queries by switching the SPARQL
strategy to "sesame", is this feature implemented? or not? Because, I have
just checked using the sesame strategy how the external triples are not
available in SPARQL until the LDPath engine doesn't ask for them. It is a
shame, because I thought LDCache would offer a behavior similar to the
SERVICE extension of SPARQL, but avoiding include the endpoint URI in the
query itself.
Nice weekend
Regards
Iván
On Fri, Oct 18, 2013 at 3:26 PM, Sergio Fernández <
sergio.fernandez@salzburgresearch.at> wrote:
> Hi Ivan,
>
>
> On 18/10/13 14:27, Ivan Ruiz Rube wrote:
>
>> I think LD Caching is intended for SPARQL queries, too. In the
>> introduction
>> of "Linked Data Caching" appears: "Apache Marmotta can optionally be
>> extended with a Linked Data Caching Module that transparently retrieves
>> resources from the Linked Data Cloud when they are needed (e.g. when
>> querying a relation to a remote resource) and caches them transparently.
>> Linked Data Caching is integrated at the triple store level and thus
>> available to all services accessing the triple store, including the SPARQL
>> endpoint."
>>
>
> Well, this is an erroneous interpretation, maybe because a wrong wording
> in the documentation. We need to improve it a lot to help our users, we are
> aware of such issue. So such kind a comments are welcomed!
>
> But what that actually tries to say is, whatever data was previously
> retrieved by the cache system, could be used in your SPARQL queries.
> Internally all the cached triples are stored at an independent named graph,
> so fully available by SPARQL.
>
> What it is less true, and Sebastian has already provided you some details,
> is that SPARQL is able to lively cache missing resources used in queries.
> This would be possible by switching the SPARQL strategy, but it may have a
> big impact on the performance.
>
>
> Besides, when I see the existing resources in the DataView page appear,
>> for
>> example, the resource with identifier
>> http://localhost:8080/LMF/**resource/Book<http://localhost:8080/LMF/resource/Book>has
>> http://localhost:8080/**resource/Book<http://localhost:8080/resource/Book>as the real URI, resulting hence in a
>> 404 HTTP Error. It is rare. I have checked that the parameter kiwi.context
>> is set to http://localhost:8080/LMF/
>>
>
> The question here would be: how did you import that resource? Because it
> just looks like a wrong base URI when parsing. By default Marmotta, hence
> LMF, takes the URI where it's running to parse files.
>
> Hope we've clarified some of your questions.
>
> Have a nice weekend.
>
>
> Cheers,
>
> --
> Sergio Fernández
> Senior Researcher
> Knowledge and Media Technologies
> Salzburg Research Forschungsgesellschaft mbH
> Jakob-Haringer-Straße 5/3 | 5020 Salzburg, Austria
> T: +43 662 2288 318 | M: +43 660 2747 925
> sergio.fernandez@**salzburgresearch.at<se...@salzburgresearch.at>
> http://www.salzburgresearch.at
>
Re: [lmf-users] LDCache
Posted by Sergio Fernández <se...@salzburgresearch.at>.
Hi Ivan,
On 18/10/13 14:27, Ivan Ruiz Rube wrote:
> I think LD Caching is intended for SPARQL queries, too. In the introduction
> of "Linked Data Caching" appears: "Apache Marmotta can optionally be
> extended with a Linked Data Caching Module that transparently retrieves
> resources from the Linked Data Cloud when they are needed (e.g. when
> querying a relation to a remote resource) and caches them transparently.
> Linked Data Caching is integrated at the triple store level and thus
> available to all services accessing the triple store, including the SPARQL
> endpoint."
Well, this is an erroneous interpretation, maybe because a wrong wording
in the documentation. We need to improve it a lot to help our users, we
are aware of such issue. So such kind a comments are welcomed!
But what that actually tries to say is, whatever data was previously
retrieved by the cache system, could be used in your SPARQL queries.
Internally all the cached triples are stored at an independent named
graph, so fully available by SPARQL.
What it is less true, and Sebastian has already provided you some
details, is that SPARQL is able to lively cache missing resources used
in queries. This would be possible by switching the SPARQL strategy, but
it may have a big impact on the performance.
> Besides, when I see the existing resources in the DataView page appear, for
> example, the resource with identifier
> http://localhost:8080/LMF/resource/Book has
> http://localhost:8080/resource/Book as the real URI, resulting hence in a
> 404 HTTP Error. It is rare. I have checked that the parameter kiwi.context
> is set to http://localhost:8080/LMF/
The question here would be: how did you import that resource? Because it
just looks like a wrong base URI when parsing. By default Marmotta,
hence LMF, takes the URI where it's running to parse files.
Hope we've clarified some of your questions.
Have a nice weekend.
Cheers,
--
Sergio Fernández
Senior Researcher
Knowledge and Media Technologies
Salzburg Research Forschungsgesellschaft mbH
Jakob-Haringer-Straße 5/3 | 5020 Salzburg, Austria
T: +43 662 2288 318 | M: +43 660 2747 925
sergio.fernandez@salzburgresearch.at
http://www.salzburgresearch.at
Re: [lmf-users] LDCache
Posted by Sebastian Schaffert <ss...@apache.org>.
Hi Ivan,
let me first say that a complete implementation of SPARQL over the Linked
Data Cloud is obviously not possible due to conceptual discrepancies:
SPARQL is a closed-world database-inspired query language while Linked Data
is an open-world concept with the additional restriction that a a resource
only knows its outgoing RDF links but not its incoming RDF links. Even a
simple query like SELECT * WHERE {?s ?p ?o} can obviously not be answered
with a conceptually complete result. What will work in certain cases (and
this is what LDCache implements) is retrieval of triples where the subject
is a known URI resource, so a SPARQL query like SELECT * WHERE {<http://...>
?p1 ?o1 . ?o1 ?p2 ?o2 } can in theory be answered correctly and completely.
There are even some other more complex cases where this might be possible
(this has been described also in a paper by Olaf Hartig). However, we
explicitly make no such guarantees, because the question if the result of a
certain SPARQL query will be complete or not is not easy to answer from a
user side, cannot be decided on purely syntactical characteristics and
requires expert knowledge. We therefore provide our own custom query
language (called LDPath), which is capable of simulating the way a user
would navigate through the Linked Data Cloud by following paths starting at
a certain subject. LDPath fully works together with LDCache, you can even
use it on the command line without the whole Marmotta platform.
Now to the concrete question you have: in theory, LMF/Marmotta can work in
the way you describe. In practice, it does not because in the default
distribution we are taking a "short path" for SPARQL queries by translating
them directly into SQL on the underlying database. The disadvantage is that
certain repository layers (like the Linked Data Caching) will not be
triggered, the advantage is that SPARQL queries do not run over the
expensive in-memory API and are therefore much faster. However, it is
possible to turn off this kind of optimization by setting the configuration
option "sparql.strategy" to "sesame" instead of "native". Just don't do it
with too big amounts of data and too complex SPARQL queries :-)
Greetings,
Sebastian
2013/10/18 Ivan Ruiz Rube <ru...@gmail.com>
> Dear Sergio,
>
> I think LD Caching is intended for SPARQL queries, too. In the
> introduction of "Linked Data Caching" appears: "Apache Marmotta can
> optionally be extended with a Linked Data Caching Module that transparently
> retrieves resources from the Linked Data Cloud when they are needed (e.g.
> when querying a relation to a remote resource) and caches them
> transparently. Linked Data Caching is integrated at the triple store level
> and thus available to all services accessing the triple store, including
> the SPARQL endpoint."
>
> Besides, when I see the existing resources in the DataView page appear,
> for example, the resource with identifier
> http://localhost:8080/LMF/resource/Book has
> http://localhost:8080/resource/Book as the real URI, resulting hence in a
> 404 HTTP Error. It is rare. I have checked that the parameter kiwi.context
> is set to http://localhost:8080/LMF/
>
> Could be this the source of the problem with the LDCache?
>
> Regards
>
> --
> Sie erhalten diese Nachricht, weil Sie Mitglied der Google Groups-Gruppe
> "LMF Users" sind.
> Um Ihr Abonnement für diese Gruppe zu beenden und keine E-Mails mehr von
> dieser Gruppe zu erhalten, senden Sie eine Email an
> lmf-users+unsubscribe@googlegroups.com.
> Weitere Optionen: https://groups.google.com/groups/opt_out
>
Re: [lmf-users] LDCache
Posted by Ivan Ruiz Rube <ru...@gmail.com>.
Dear Sergio,
I think LD Caching is intended for SPARQL queries, too. In the introduction
of "Linked Data Caching" appears: "Apache Marmotta can optionally be
extended with a Linked Data Caching Module that transparently retrieves
resources from the Linked Data Cloud when they are needed (e.g. when
querying a relation to a remote resource) and caches them transparently.
Linked Data Caching is integrated at the triple store level and thus
available to all services accessing the triple store, including the SPARQL
endpoint."
Besides, when I see the existing resources in the DataView page appear, for
example, the resource with identifier
http://localhost:8080/LMF/resource/Book has
http://localhost:8080/resource/Book as the real URI, resulting hence in a
404 HTTP Error. It is rare. I have checked that the parameter kiwi.context
is set to http://localhost:8080/LMF/
Could be this the source of the problem with the LDCache?
Regards
Re: [lmf-users] LDCache
Posted by Sergio Fernández <se...@salzburgresearch.at>.
Hi Ivan,
On 18/10/13 12:20, Ivan Ruiz Rube wrote:
> I have checked that when I use the webservice /{BASE}/cache/cached the
> resources are correctly fetched and saved into the cache context. However,
> when I use the SNORQL interfaces the triples are not received.
Jakob may correctly me if I'm wrong, but I think the Marmotta SPARQL
implementation is not cache-aware; i.e., SPARQL queries would not
automatically cache resources not locally stored before query time.
Try to browse the data:
http://host/marmotta/resource?uri=http://example.org/your/uri
which should make use of LDCache to transparently retrieve missing
resources.
If you'd need to query, LDPath does cache in query time.
Hope this helps.
Cheers,
--
Sergio Fernández
Senior Researcher
Knowledge and Media Technologies
Salzburg Research Forschungsgesellschaft mbH
Jakob-Haringer-Straße 5/3 | 5020 Salzburg, Austria
T: +43 662 2288 318 | M: +43 660 2747 925
sergio.fernandez@salzburgresearch.at
http://www.salzburgresearch.at
Re: [lmf-users] LDCache
Posted by Ivan Ruiz Rube <ru...@gmail.com>.
Hello,
I have checked that when I use the webservice /{BASE}/cache/cached the
resources are correctly fetched and saved into the cache context. However,
when I use the SNORQL interfaces the triples are not received. I have
modified the params ldcache.expiry and ldcache.minexpiry but I am not be
able to get it working.
Thanks.
El martes, 8 de octubre de 2013 12:46:07 UTC+2, Ivan Ruiz Rube escribió:
>
> Thank you Jakob.
>
> I am using the version 3.0.0 of LMF. I have seen how the triples are well
> received by the webservice "live". However, when I use the SNORQL interface
> I can not retrieve this triples. It seems like the collected triples are
> not stored in the cache.
>
>
>
>
>
> El martes, 8 de octubre de 2013 11:53:22 UTC+2, Jakob Frank escribió:
>>
>> [ cross-posting to users@marmotta.incubator.apache.org as LDCache has
>> moved to Apache Marmotta ]
>>
>> Hi Ivan,
>>
>> did you try explicitly requesting a resource from LDCache, e.g. by
>> querying to "live" webservice?
>> e.g.
>> <
>> http://localhost:8080/LMF/cache/live?uri=http://dbpedia.org/resource/Berlin>
>>
>>
>> Plus: which version are you currently using?
>>
>> Best,
>> Jakob
>>
>> On Tue, 2013-10-08 at 02:26 -0700, Ivan Ruiz Rube wrote:
>> > Dear LMF Users,
>> >
>> > I have a problem using the cache capabilities of LMF. I can not
>> retrieve
>> > resources from the external endpoints configurated in my LMF
>> installation,
>> > neither with the DBPedia SPARQL sample endpoint.
>> >
>> > The property ldcache.enabled is activated and I have found interesting
>> > nothing about it in the logs marmotta-main, catalina and access-log.
>> >
>> > Please, any idea?
>> >
>> > Thanks in advance.
>> >
>> > Regards.
>> >
>> >
>> >
>>
>> --
>> DI Jakob Frank
>> Knowledge and Media Technologies
>>
>> Salzburg Research Forschungsgesellschaft mbH
>> Jakob Haringer-Strasse 5/3 | 5020 Salzburg, Austria
>> T: +43.662.2288-419
>> jakob.frank@salzburgresearch.at
>> http://www.salzburgresearch.at
>>
>>
Re: [lmf-users] LDCache
Posted by Ivan Ruiz Rube <ru...@gmail.com>.
Thank you Jakob.
I am using the version 3.0.0 of LMF. I have seen how the triples are well
received by the webservice "live". However, when I use the SNORQL interface
I can not retrieve this triples. It seems like the collected triples are
not stored in the cache.
El martes, 8 de octubre de 2013 11:53:22 UTC+2, Jakob Frank escribió:
>
> [ cross-posting to users@marmotta.incubator.apache.org as LDCache has
> moved to Apache Marmotta ]
>
> Hi Ivan,
>
> did you try explicitly requesting a resource from LDCache, e.g. by
> querying to "live" webservice?
> e.g.
> <
> http://localhost:8080/LMF/cache/live?uri=http://dbpedia.org/resource/Berlin>
>
>
> Plus: which version are you currently using?
>
> Best,
> Jakob
>
> On Tue, 2013-10-08 at 02:26 -0700, Ivan Ruiz Rube wrote:
> > Dear LMF Users,
> >
> > I have a problem using the cache capabilities of LMF. I can not retrieve
> > resources from the external endpoints configurated in my LMF
> installation,
> > neither with the DBPedia SPARQL sample endpoint.
> >
> > The property ldcache.enabled is activated and I have found interesting
> > nothing about it in the logs marmotta-main, catalina and access-log.
> >
> > Please, any idea?
> >
> > Thanks in advance.
> >
> > Regards.
> >
> >
> >
>
> --
> DI Jakob Frank
> Knowledge and Media Technologies
>
> Salzburg Research Forschungsgesellschaft mbH
> Jakob Haringer-Strasse 5/3 | 5020 Salzburg, Austria
> T: +43.662.2288-419
> jakob.frank@salzburgresearch.at
> http://www.salzburgresearch.at
>
>