lucene-solr-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Erick Erickson <erickerick...@gmail.com>
Subject Re: Unable to perform search query after changing uniqueKey
Date Tue, 31 Mar 2015 05:58:12 GMT
I would never put spaces in my field names! Frankly I have no clue
what Solr does with that, but it can't be good. Solr explicitly
supports Java naming conventions, camel case, underscores and numbers.
Special symbols are frowned upon, I never use anything but upper case,
lower case and underscores. Actually, I don't use upper case either
but that's a personal preference. Other things might work, but only by
chance.

Best,
Erick

On Mon, Mar 30, 2015 at 8:59 PM, Zheng Lin Edwin Yeo
<edwinyeozl@gmail.com> wrote:
> Latest information that I've found for this is that the error only occurs
> for shard2.
>
> If I do a search for just shard1, those records that are assigned to shard1
> will be able to be displayed. Only when I search for shard2 will the
> NullPointerException error occurs. Previously I was doing a search for both
> shards.
>
> Is there any settings that I required to do for shard2 in order to solve
> this issue? Currently I have not made any changes to the shards since I
> created it using
> http://localhost:8983/solr/admin/collections?action=CREATE&name=nps1&numShards=2&collection.configName=collection1
>
>
> Regards,
> Edwin
>
> On 31 March 2015 at 09:42, Zheng Lin Edwin Yeo <edwinyeozl@gmail.com> wrote:
>
>> Hi Erick,
>>
>> I've changed the uniqueKey from id to Item No.
>>
>> <uniqueKey>Item No</uniqueKey>
>>
>>
>> Below are my definitions for both the id and Item No.
>>
>> <field name="id" type="string" indexed="true" stored="true"
>> required="false" multiValued="false" />
>> <field name="Item No" type="text_general" indexed="true" stored="true"/>
>>
>> Regards,
>> Edwin
>>
>>
>> On 30 March 2015 at 23:05, Erick Erickson <erickerickson@gmail.com> wrote:
>>
>>> Well, let's see the definition of your ID field, 'cause I'm puzzled.
>>>
>>> It's definitely A Bad Thing to have it be any kind of tokenized field
>>> though, but that's a shot in the dark.
>>>
>>> Best,
>>> Erick
>>>
>>> On Mon, Mar 30, 2015 at 2:17 AM, Zheng Lin Edwin Yeo
>>> <edwinyeozl@gmail.com> wrote:
>>> > Hi Mostafa,
>>> >
>>> > Yes, I've defined all the fields in schema.xml. It is able to work on
>>> the
>>> > version without SolrCloud, but it is not working for the one with
>>> SolrCloud.
>>> > Both of them are using the same schema.xml.
>>> >
>>> > Regards,
>>> > Edwin
>>> >
>>> >
>>> >
>>> > On 30 March 2015 at 14:34, Mostafa Gomaa <mostafa.gomaa89@gmail.com>
>>> wrote:
>>> >
>>> >> Hi Zheng,
>>> >>
>>> >> It's possible that there's a problem with your schema.xml. Are all
>>> fields
>>> >> defined and have appropriate options enabled?
>>> >>
>>> >> Regards,
>>> >>
>>> >> Mostafa.
>>> >>
>>> >> On Mon, Mar 30, 2015 at 7:49 AM, Zheng Lin Edwin Yeo <
>>> edwinyeozl@gmail.com
>>> >> >
>>> >> wrote:
>>> >>
>>> >> > Hi Erick,
>>> >> >
>>> >> > I've tried that, and removed the data directory from both the
>>> shards. But
>>> >> > the same problem still occurs, so we probably can rule out the
>>> "memory"
>>> >> > issue.
>>> >> >
>>> >> > Regards,
>>> >> > Edwin
>>> >> >
>>> >> > On 30 March 2015 at 12:39, Erick Erickson <erickerickson@gmail.com>
>>> >> wrote:
>>> >> >
>>> >> > > I meant shut down Solr and physically remove the entire data
>>> >> > > directory. Not saying this is the cure, but it can't hurt
to rule
>>> out
>>> >> > > the index having "memory"...
>>> >> > >
>>> >> > > Best,
>>> >> > > Erick
>>> >> > >
>>> >> > > On Sun, Mar 29, 2015 at 6:35 PM, Zheng Lin Edwin Yeo
>>> >> > > <edwinyeozl@gmail.com> wrote:
>>> >> > > > Hi Erick,
>>> >> > > >
>>> >> > > > I used the following query to delete all the index.
>>> >> > > >
>>> >> > > > http://localhost:8983/solr/update?stream.body=
>>> >> > > <delete><query>*:*</query></delete>
>>> >> > > http://localhost:8983/solr/update?stream.body=<commit/>
>>> >> > > >
>>> >> > > >
>>> >> > > > Or is it better to physically delete the entire data
directory?
>>> >> > > >
>>> >> > > >
>>> >> > > > Regards,
>>> >> > > > Edwin
>>> >> > > >
>>> >> > > >
>>> >> > > > On 28 March 2015 at 02:27, Erick Erickson <
>>> erickerickson@gmail.com>
>>> >> > > wrote:
>>> >> > > >
>>> >> > > >> You say you re-indexed, did you _completely_ remove
the data
>>> >> directory
>>> >> > > >> first, i.e. the parent of the "index" and, maybe,
"tlog"
>>> >> directories?
>>> >> > > >> I've occasionally seen remnants of old definitions
"pollute"
>>> the new
>>> >> > > >> one, and since the <uniqueKey> key is so fundamental
I can see
>>> it
>>> >> > > >> being a problem.
>>> >> > > >>
>>> >> > > >> Best,
>>> >> > > >> Erick
>>> >> > > >>
>>> >> > > >> On Fri, Mar 27, 2015 at 1:42 AM, Andrea Gazzarini
<
>>> >> > > a.gazzarini@gmail.com>
>>> >> > > >> wrote:
>>> >> > > >> > Hi Edwin,
>>> >> > > >> > please provide some other detail about your
context, (e.g.
>>> >> complete
>>> >> > > >> > stacktrace, query you're issuing)
>>> >> > > >> >
>>> >> > > >> > Best,
>>> >> > > >> > Andrea
>>> >> > > >> >
>>> >> > > >> >
>>> >> > > >> > On 03/27/2015 09:38 AM, Zheng Lin Edwin Yeo
wrote:
>>> >> > > >> >>
>>> >> > > >> >> Hi everyone,
>>> >> > > >> >>
>>> >> > > >> >> I've changed my uniqueKey to another name,
instead of using
>>> id,
>>> >> on
>>> >> > > the
>>> >> > > >> >> schema.xml.
>>> >> > > >> >>
>>> >> > > >> >> However, after I have done the indexing
(the indexing is
>>> >> > successful),
>>> >> > > >> I'm
>>> >> > > >> >> not able to perform a search query on it.
I gives the error
>>> >> > > >> >> java.lang.NullPointerException.
>>> >> > > >> >>
>>> >> > > >> >> Is there other place which I need to configure,
besides
>>> changing
>>> >> > the
>>> >> > > >> >> uniqueKey field in scheam.xml?
>>> >> > > >> >>
>>> >> > > >> >> Regards,
>>> >> > > >> >> Edwin
>>> >> > > >> >>
>>> >> > > >> >
>>> >> > > >>
>>> >> > >
>>> >> >
>>> >>
>>>
>>
>>

Mime
View raw message