You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@marmotta.apache.org by Bernhard Ortner <be...@gmx.net> on 2013/09/29 14:29:48 UTC
JDBC Driver for Marmotta
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Hi Marmotta Users
Since Im relatively new to marmotta, i have some practical questions,
regarding the data management of the project:
Is there any additional information regarding the jdbc connection,
except the wiki post at [1] ?
To be more precise here are my concrete questions:
Do you have a BulkCRUD handler, that allows the management of > 10m
triples ?
Can i use the current implementation as pooled Spring Datasource ?
Kind regards
Bernhard Ortner
[1] http://marmotta.incubator.apache.org/kiwi/triplestore.html
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.20 (MingW32)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/
iQEcBAEBAgAGBQJSSB08AAoJEGKIVFSKl71aVWAH/iQbVSPdAVXVaKLLu//tiDL1
JO6PPeX9SQEu0or5x0Mwz1weFmdHyhvOOyMNiBBnnD5rExJo1k1+ai1rf9Hf9PUj
n4HzhkTZccwlZbQfH1r3O7fXBNK8AXZ97tLzD5bMPw9LrFzVCB889w+MHfHb6TT4
boVmVRLAvsqEPtmDIwVfRlGwRDWhVmwDygiJx1px5NUyaSnVVe18QO+PwrKRKRcI
g0hR9V2YHlsl2Xz89r1tIgrV4LTjNFPgrV6v7F6eBvjlU33IBOg57Eo9nKC2s46P
p352N+Hv7sh+eGj8UerXI1gSvf+KejxgCVY9pfRUcm5aDR/FBURfYF5jxlvp31Y=
=USRW
-----END PGP SIGNATURE-----
Re: JDBC Driver for Marmotta
Posted by Sergio Fernández <se...@salzburgresearch.at>.
Dear Bernhard,
(moving this discussion to dev@marmotta, since it is more about
technical things than users' support)
On 01/10/13 21:29, Bernhard Ortner wrote:
> My Spring Use Case is:
> I want to Inject marmotta as datasource via Spring.
> To be more precise, i want to configure marmotta via Spring Annotations
> and / or a context file (as Pooled DataSource)
> Raffaele just gave me the answer, that i was searching for... so I will
> try to write / configure a Spring service for marmotta.
So, if you understood correctly, you would like to have a Spring Data
Source directly consuming the triple store by JDBC.
Strictly talking, that would be a Data Source for the KiWi triple store
[1]. Although right now it is true there is not difference, since that
one is the single triple store currently supported by Marmotta. Perhaps
we should improve how the platform is described [2]... But our roadmap
for the short-term future (hopefully for 3.2.x) is to be able to run
Marmotta over any Sesame-based triple store.
The clarification comes because in LMF (the former project) we were
thinking about something similar, providing support to Spring Data [3].
But in that case was at a resource level (so using LDP and/or SPARQL,
most likely), and not directly at a low level. But we never managed to
find time to implement such idea.
But your proposal also looks quite interesting, and it could be quite
useful in some use cases. Please, keep us updated about it :-)
Cheers,
[1] http://marmotta.incubator.apache.org/kiwi/introduction.html
[2] http://marmotta.incubator.apache.org/platform/introduction.html
[3] https://code.google.com/p/lmf/issues/detail?id=95
--
Sergio Fernández
Salzburg Research
+43 662 2288 318
Jakob-Haringer Strasse 5/II
A-5020 Salzburg (Austria)
http://www.salzburgresearch.at
Re: JDBC Driver for Marmotta
Posted by Bernhard Ortner <be...@gmx.net>.
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Hi Sebastian
My Spring Use Case is:
I want to Inject marmotta as datasource via Spring.
To be more precise, i want to configure marmotta via Spring Annotations
and / or a context file (as Pooled DataSource)
Raffaele just gave me the answer, that i was searching for... so I will
try to write / configure a Spring service for marmotta.
Have a nice evening
Bernhard
Am 30.09.2013 12:00, schrieb Sebastian Schaffert:
> Hi Bernhard,
>
> Raffaele already pointed you to an important discussion on the issue.
We do not have a bulk handler at the moment (this is still planned), but
the performance has been improved considerably so that 10 million
triples should be no problem to handle. We have installations of more
than 150 million triples that are running fine (but be careful with
SPARQL queries in such environments...). You might want to split your
dataset into smaller files before importing, though, because otherwise
the transaction size will be too big.
>
> I don't fully understand the question with the SpringDatasource. What
is your use case you have in mind?
>
> Greetings,
>
> Sebastian
>
>
>
> 2013/9/30 Raffaele Palmieri <raffaele.palmieri@gmail.com
<ma...@gmail.com>>
>
> Hi Bernhard, I 'd suggest you to look following thread
http://markmail.org/message/gj6oh7askhepi7yj#query:+page:1+mid:staylfo6gg37wyei+state:results
.
> In that post, you'll find some useful informations to understand
data life cycle.
>
>
> On 29 September 2013 14:29, Bernhard Ortner
<bernhardortner@gmx.net <ma...@gmx.net>> wrote:
>
>
> Hi Marmotta Users
>
> Since Im relatively new to marmotta, i have some practical questions,
> regarding the data management of the project:
> Is there any additional information regarding the jdbc connection,
> except the wiki post at [1] ?
>
> To be more precise here are my concrete questions:
> Do you have a BulkCRUD handler, that allows the management of > 10m
> triples ?
>
>
> > The process of bulk import should pass for loading in memory of
triples and then writing in a such format to be processed from DBMS.
>
>
> Can i use the current implementation as pooled Spring Datasource ?
>
>
> > I don't think so, because Marmotta actually doesn't use Spring.
>
>
>
> Kind regards
>
> Bernhard Ortner
>
> [1] http://marmotta.incubator.apache.org/kiwi/triplestore.html
>
> Regards,
> Raffaele.
>
>
- --
yours sincerely
Bernhard Ortner, BSc.
- ------------------
Phone: +43 664 2538 738
Skype: be_ortner
E-Mail: bernhardortner@gmx.net
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.20 (MingW32)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/
iQEcBAEBAgAGBQJSSyKrAAoJEGKIVFSKl71auqMH+wcw28z0GrmmV5wZnTVh8FuB
CLY+J1Bq4ekNHj+iopGRhOApZ2jERJRNex9ajbXNmb0WgyUyRJAHN6YctvILgPi7
MUrvi5yFvOTEINxLnTjvHrp+xM0hh6jbFvOxRqm3ui7czerc3saaE+e3nCjIk3Bt
znVVdpB2aeHTTRJzFsexNwSIPGUt2K6iTphjvIZIo9g/vcZLJaj2I/zMNHFjO8jL
kWmg9h4WTiwELfHZKqPRW6jrH5qno/jm62fk+/GULg0huDLrJ8aawH1K5Qgni97u
vHlhKVJUJqxBB9Kb9czbc3aSROO26uq91oI4U2f814qfgwlv6mqH50MRTAZJGNU=
=7OWB
-----END PGP SIGNATURE-----
Re: JDBC Driver for Marmotta
Posted by Sebastian Schaffert <se...@gmail.com>.
Hi Bernhard,
Raffaele already pointed you to an important discussion on the issue. We do
not have a bulk handler at the moment (this is still planned), but the
performance has been improved considerably so that 10 million triples
should be no problem to handle. We have installations of more than 150
million triples that are running fine (but be careful with SPARQL queries
in such environments...). You might want to split your dataset into smaller
files before importing, though, because otherwise the transaction size will
be too big.
I don't fully understand the question with the SpringDatasource. What is
your use case you have in mind?
Greetings,
Sebastian
2013/9/30 Raffaele Palmieri <ra...@gmail.com>
> Hi Bernhard, I 'd suggest you to look following thread
> http://markmail.org/message/gj6oh7askhepi7yj#query:+page:1+mid:staylfo6gg37wyei+state:results
> .
> In that post, you'll find some useful informations to understand data life
> cycle.
>
>
> On 29 September 2013 14:29, Bernhard Ortner <be...@gmx.net>wrote:
>
>>
>> -----BEGIN PGP SIGNED MESSAGE-----
>> Hash: SHA1
>>
>> Hi Marmotta Users
>>
>> Since Im relatively new to marmotta, i have some practical questions,
>> regarding the data management of the project:
>> Is there any additional information regarding the jdbc connection,
>> except the wiki post at [1] ?
>>
>> To be more precise here are my concrete questions:
>> Do you have a BulkCRUD handler, that allows the management of > 10m
>> triples ?
>>
>
> The process of bulk import should pass for loading in memory of triples
> and then writing in a such format to be processed from DBMS.
>
>
>> Can i use the current implementation as pooled Spring Datasource ?
>>
>
> I don't think so, because Marmotta actually doesn't use Spring.
>
>
>>
>> Kind regards
>>
>> Bernhard Ortner
>>
>> [1] http://marmotta.incubator.apache.org/kiwi/triplestore.html
>> -----BEGIN PGP SIGNATURE-----
>> Version: GnuPG v2.0.20 (MingW32)
>> Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/
>>
>> iQEcBAEBAgAGBQJSSB08AAoJEGKIVFSKl71aVWAH/iQbVSPdAVXVaKLLu//tiDL1
>> JO6PPeX9SQEu0or5x0Mwz1weFmdHyhvOOyMNiBBnnD5rExJo1k1+ai1rf9Hf9PUj
>> n4HzhkTZccwlZbQfH1r3O7fXBNK8AXZ97tLzD5bMPw9LrFzVCB889w+MHfHb6TT4
>> boVmVRLAvsqEPtmDIwVfRlGwRDWhVmwDygiJx1px5NUyaSnVVe18QO+PwrKRKRcI
>> g0hR9V2YHlsl2Xz89r1tIgrV4LTjNFPgrV6v7F6eBvjlU33IBOg57Eo9nKC2s46P
>> p352N+Hv7sh+eGj8UerXI1gSvf+KejxgCVY9pfRUcm5aDR/FBURfYF5jxlvp31Y=
>> =USRW
>> -----END PGP SIGNATURE-----
>>
>> Regards,
> Raffaele.
>
Re: JDBC Driver for Marmotta
Posted by Raffaele Palmieri <ra...@gmail.com>.
Hi Bernhard, I 'd suggest you to look following thread
http://markmail.org/message/gj6oh7askhepi7yj#query:+page:1+mid:staylfo6gg37wyei+state:results
.
In that post, you'll find some useful informations to understand data life
cycle.
On 29 September 2013 14:29, Bernhard Ortner <be...@gmx.net> wrote:
>
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Hi Marmotta Users
>
> Since Im relatively new to marmotta, i have some practical questions,
> regarding the data management of the project:
> Is there any additional information regarding the jdbc connection,
> except the wiki post at [1] ?
>
> To be more precise here are my concrete questions:
> Do you have a BulkCRUD handler, that allows the management of > 10m
> triples ?
>
The process of bulk import should pass for loading in memory of triples and
then writing in a such format to be processed from DBMS.
> Can i use the current implementation as pooled Spring Datasource ?
>
I don't think so, because Marmotta actually doesn't use Spring.
>
> Kind regards
>
> Bernhard Ortner
>
> [1] http://marmotta.incubator.apache.org/kiwi/triplestore.html
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v2.0.20 (MingW32)
> Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/
>
> iQEcBAEBAgAGBQJSSB08AAoJEGKIVFSKl71aVWAH/iQbVSPdAVXVaKLLu//tiDL1
> JO6PPeX9SQEu0or5x0Mwz1weFmdHyhvOOyMNiBBnnD5rExJo1k1+ai1rf9Hf9PUj
> n4HzhkTZccwlZbQfH1r3O7fXBNK8AXZ97tLzD5bMPw9LrFzVCB889w+MHfHb6TT4
> boVmVRLAvsqEPtmDIwVfRlGwRDWhVmwDygiJx1px5NUyaSnVVe18QO+PwrKRKRcI
> g0hR9V2YHlsl2Xz89r1tIgrV4LTjNFPgrV6v7F6eBvjlU33IBOg57Eo9nKC2s46P
> p352N+Hv7sh+eGj8UerXI1gSvf+KejxgCVY9pfRUcm5aDR/FBURfYF5jxlvp31Y=
> =USRW
> -----END PGP SIGNATURE-----
>
> Regards,
Raffaele.