For the datasize you are proposing , single index should be fine .Just
give the m/c enough RAM
Distributed search involves multiple requests made between shards
which may be an unncessary overhead.
--Noble
On Wed, Jun 4, 2008 at 4:02 PM, Ritesh Ambastha [EMAIL PROTECTED] wrote:
Thanks Noble,
Thanks Noble,
That means, I can go ahead with single Index for long.
:)
Regards,
Ritesh Ambastha
Noble Paul നോബിള് नोब्ळ् wrote:
For the datasize you are proposing , single index should be fine .Just
give the m/c enough RAM
Distributed search involves multiple requests made between
Thanks Noble,
I maintain two separate indexes on my disk for two different search
services.
The index size of two are: 91MB and 615MB. I am pretty sure that these index
size will grow in future, and may reach 10GB.
My doubts :
1. When should I start partitioning my index?
2. Is there any
A lot of this also depends on the number of documents. But we have
successfully used Solr with upto 10-12 million documents.
On Wed, Jun 4, 2008 at 4:10 PM, Ritesh Ambastha [EMAIL PROTECTED]
wrote:
Thanks Noble,
That means, I can go ahead with single Index for long.
:)
Regards,
Ritesh
The number of docs I have indexed till now is : 1,633,570
I am bit afraid as the number of indexed docs will grow atleast 5-10 times
in very near future.
Regards,
Ritesh Ambastha
Shalin Shekhar Mangar wrote:
A lot of this also depends on the number of documents. But we have
successfully
Fot 16 mil docs it may not be necessary. Add the shards when you see
that perf is degrading.
--Noble
On Wed, Jun 4, 2008 at 4:17 PM, Ritesh Ambastha [EMAIL PROTECTED] wrote:
The number of docs I have indexed till now is : 1,633,570
I am bit afraid as the number of indexed docs will grow
2008/6/3 Ryan McKinley [EMAIL PROTECTED]:
This way I don't connect:
new CommonsHttpSolrServer(http://localhost:8983/solr/idxItem;)
this is how you need to connect... otherwise nothing will work.
When I try this way, I get the following exception, when trying to make an
update to my
Hi,
I wanted to use the new dismax support for more like this described in
SOLR-295 https://issues.apache.org/jira/browse/SOLR-295 but can't even get
the new syntax for dismax to work (described in
SOLR-281https://issues.apache.org/jira/browse/SOLR-281).
Any ideas if this functionality works?
On Wed, Jun 4, 2008 at 11:11 AM, Tom Morton [EMAIL PROTECTED] wrote:
I wanted to use the new dismax support for more like this described in
SOLR-295 https://issues.apache.org/jira/browse/SOLR-295 but can't even get
the new syntax for dismax to work (described in
Hi,
I'm doing some test with solr to see if it can be usefull for us, and
after deleting the index to restart from scratch the indexing, it
returns me this error when I want to post datas :
java.lang.RuntimeException: java.io.FileNotFoundException: no segments*
file found in
Cuong,
I think you will need some manipulation beyond solr queries. You should
separate the results by your site criteria after retrieving them. After
that, you could cache the results on your application and randomize the
lists every time you render the a page.
I don't know if solr has
Noble,
Thanks for continuing to assist me on trying to come up a config that works.
A couple of questions/clarifications:
1) I had to introduce the artificial comboID and the transformer because
of a conflict with a parallel entity on the id (vets and owners).
2) I don't think there is a conflict
Hi,
Thanks Yonik. That fixed that. I would be useful to change one of the
existing dismax query types in the default solrconfig.xml to use this new
syntax (Especially since DisMaxRequestHandler is being deprecared.)
Thanks again...Tom
On Wed, Jun 4, 2008 at 11:19 AM, Yonik Seeley [EMAIL
On Jun 4, 2008, at 10:07 AM, Alexander Ramos Jardim wrote:
2008/6/3 Ryan McKinley [EMAIL PROTECTED]:
This way I don't connect:
new CommonsHttpSolrServer(http://localhost:8983/solr/idxItem;)
this is how you need to connect... otherwise nothing will work.
When I try this way, I get the
Hi,
SOLR-295 https://issues.apache.org/jira/browse/SOLR-295 mentions boost
support for morelikethis and then seems to have been subsumed by
SOLR-281https://issues.apache.org/jira/browse/SOLR-281.
To be clear, I'm talking about boosts for the mlt.fl fields and how they are
ranked rather than for
Hi Julio,
The following are my assumptions after studying your given data-config
examples
1. The column id is present in all three tables -- vets, owners and pets.
2. Vets and owners are independent of each other, there is no join required
between them
3. There is a parent-child relationship
It is mapped correctly.
2008/6/4 Erik Hatcher [EMAIL PROTECTED]:
On Jun 4, 2008, at 10:07 AM, Alexander Ramos Jardim wrote:
2008/6/3 Ryan McKinley [EMAIL PROTECTED]:
This way I don't connect:
new CommonsHttpSolrServer(http://localhost:8983/solr/idxItem;)
this is how you need to
Thanks Shalin,
I'll try this asap. Yes, you did understand the sample schema I've been
playing with.
Just a couple of questions to clarify for my own understanding your
proposal.
1) the column comboId doesn't exist on the dB (yet it is specified as a
separate column for both owners and vets in
1. Yes you can add virtual columns. Any column/name which does not exist in
schema.xml are used for joins but not added to the document. The column
attribute is the key using which data is read from the entity's Map. The
name attribute is the solr field to which data is written. Also note that
Another thing to note, the parentDeltaQuery is of no use if deltaQuery is
not specified. In my experience, DataImportHandler is pretty fast and delta
queries may not be needed at all if your dataset is small. We use it without
delta queries even with millions of solr documents and it completes
Hi,
Just recently upgraded w/ trunk version, Solr works fine but Luke
Lucli are showing this:
lucli index /www/solr/test/index
Lucene CLI. Using directory '/www/solr/test/index'. Type 'help' for
instructions.
Error:org.apache.lucene.index.CorruptIndexException: Unknown format
version:
I got this error when trying to use an index generated by an old version
with trunk Solr.
2008/6/4 Jon Baer [EMAIL PROTECTED]:
Hi,
Just recently upgraded w/ trunk version, Solr works fine but Luke Lucli
are showing this:
lucli index /www/solr/test/index
Lucene CLI. Using directory
Jon,
Lucli is ancient and, as far as I recall, has its own Lucene jar, which you
could try replacing. Luke might be better at this point, but even with Luke
you may have to use your own Lucene jar (the one you used with Solr+DIH). The
error indicates index format incompatibility (likely old
Are you using a recent version of multi-core?
Do you have the /update RequestHandler mapped in solrconfig.xml?
Since multicore support is new, it does not support the @deprecated /
update servlet
ryan
On Jun 4, 2008, at 10:07 AM, Alexander Ramos Jardim wrote:
2008/6/3 Ryan McKinley
You will more than likely need to replace the Lucene jars in Luke with
the ones used in Solr. We have upgraded Solr's Lucene dependencies in
the trunk.
I think, this means getting just the Luke jar (instead of Luke all)
and then using it with the Lucene lib here. Double check the Luke
Thanks ... that worked. Much appreciated.
- Jon
On Jun 4, 2008, at 4:51 PM, Grant Ingersoll wrote:
You will more than likely need to replace the Lucene jars in Luke
with the ones used in Solr. We have upgraded Solr's Lucene
dependencies in the trunk.
I think, this means getting just
Hello - I was wondering if there is a work around with POSTing repeated fields
to Solr. I am using Jetty as my container with Solr 1.2.
I tried something like:
Hi,
I have a question about the HttpDataSource (DataImportHandler) ...
Is it possible add common values *explicitly*, something like:
field column=column value=value commonField=true /
Im blanking on if xpath has a command / option to return back just a
string literal (vs. node).
Thanks.
: I am using function queries to rank the results,
: if some/ allthe fields (used in the function ) are missing from the document
: what will be the ranking behavior for such documents?
Off the top of my head, I believe it's zero, but an easy way to check is
to run a simple linear function
: I dunno... for something like fl, it still seems a bit verbose to
: list every field separately.
: Some of these things feel like trade offs in ease of readability
: manually typing of URLs vs ease of programmatic manipulation.
I don't even think the primary issue is programmatic
: Hoss:Your are right. It has a version byte written first. This can be
: used for any changes that come later..So , when we introduce any
Cool .. i just wanted to make sure the issue here was the defualt format
changing from XML to binary as part of the development cycle and not an
expected
On 4-Jun-08, at 2:22 PM, Andrew Nagy wrote:
Hello - I was wondering if there is a work around with POSTing
repeated fields to Solr. I am using Jetty as my container with Solr
1.2.
I tried something like:
commonField=true can be added in any field when you are using an
XPathEntityProcessor.But you will never need to do so because only xml
has such a requirement. If you wish to add a string literal use a
TemplateTransformer and keep the field column=column value=value
template=my-string-literal/
The
attachment did not work try this
http://www.nabble.com/Re%3A-How-to-describe-2-entities-in-dataConfig-for-the-DataImporter--p17577610.html
--Noble
On Thu, Jun 5, 2008 at 9:37 AM, Noble Paul നോബിള് नोब्ळ्
[EMAIL PROTECTED] wrote:
commonField=true can be added in any field when you are using an
Hi folks,
I have a scenario as follows:
I have a CMS where in I'm storing all the contents. I need to index all these
contents and have a search on these indexes. For indexing, I can define a
schema for all the contents. Some of the properties are like title, headline,
body, keywords,
hi ,
use multi-core Solr. Each core can have its own schema.
If possible the DISCLAIMER can be dropped.
--Noble
On Thu, Jun 5, 2008 at 11:13 AM, Sachit P. Menon
[EMAIL PROTECTED] wrote:
Hi folks,
I have a scenario as follows:
I have a CMS where in I'm storing all the contents. I need to
Hi Sachit,
I think what you could do is to create all the core fields of your models
such as username, role, title, body, images... You can name them with prefix
like user.username, user.role, article.title, article.body... If you want to
dynamically add more fields to your schema, you can use
37 matches
Mail list logo
