You are viewing a plain text version of this content. The canonical link for it is here.
Posted to solr-user@lucene.apache.org by Pratik Patel <pr...@semandex.net> on 2018/02/07 22:59:16 UTC
Best Practice about solr cloud schema
Hello all,
I have added some fields to default managed-schema file. I was wondering if
it is safe to take default managed-schema file as is and add your own
fields to it in production. What is the best practice for this? As I
understand, it should be safe to use default schema as base if documents
that are going to be indexed in solr will only have newly defined fields in
it. In fact, it helps because the common field types are already defined in
default schema which can be re-used. I looked through the documentation but
couldn't find the answer and more clarity on this would be helpful.
Is it safe to use default managed-schema file as base add your own fields
to it?
Thanks,
Pratik
Re: Best Practice about solr cloud schema
Posted by Pratik Patel <pr...@semandex.net>.
That makes it clear. Thanks a lot for your help.
Pratik
On Feb 7, 2018 10:33 PM, "Erick Erickson" <er...@gmail.com> wrote:
> It can pretty much be used as-is, _except_....
>
> you'll find one or more entries in your request handlers like:
> <str name="df">_text_</str>
>
> Change "_text_" to something in your schema, that's the default search
> field if you don't field-qualify your search terms.
>
> Note that if you take out, for instance, all of your non-english
> fieldTypes, you can also remove most of the stuff under the /lang
> folder.
>
> I essentially always test this out on a local, stand-alone instance
> until I can index a few documents and query them, it's faster than
> always having to remember to move them to ZooKeeper....
>
> Best,
> Erick
>
> On Wed, Feb 7, 2018 at 7:14 PM, Pratik Patel <pr...@semandex.net> wrote:
> > Hey Eric, thanks for the clarification! What about solrConfig.xml file?
> > Sure, it should be customized to suit one's needs but can it be used as a
> > base or is it best to create one from scratch ?
> >
> > Thanks,
> > Pratik
> >
> > On Wed, Feb 7, 2018 at 5:29 PM, Erick Erickson <er...@gmail.com>
> > wrote:
> >
> >> That's really the point of the default managed-schema, to be a base
> >> you use for your customizations. In fact, I often _remove_ most of the
> >> fields (and especially fieldTypes) that I don't need. This includes
> >> dynamic fields, copyFields and the like.
> >>
> >> Sometimes it's actually easier, though, to just start all over.
> >>
> >> BTW, do not delete any field that begins and ends with an underscore,
> >> e.g. _version_ unless you know exactly what the consequences are....
> >>
> >> Best,
> >> Erick
> >>
> >> On Wed, Feb 7, 2018 at 2:59 PM, Pratik Patel <pr...@semandex.net>
> wrote:
> >> > Hello all,
> >> >
> >> > I have added some fields to default managed-schema file. I was
> wondering
> >> if
> >> > it is safe to take default managed-schema file as is and add your own
> >> > fields to it in production. What is the best practice for this? As I
> >> > understand, it should be safe to use default schema as base if
> documents
> >> > that are going to be indexed in solr will only have newly defined
> fields
> >> in
> >> > it. In fact, it helps because the common field types are already
> defined
> >> in
> >> > default schema which can be re-used. I looked through the
> documentation
> >> but
> >> > couldn't find the answer and more clarity on this would be helpful.
> >> >
> >> > Is it safe to use default managed-schema file as base add your own
> fields
> >> > to it?
> >> >
> >> > Thanks,
> >> > Pratik
> >>
>
Re: Best Practice about solr cloud schema
Posted by Erick Erickson <er...@gmail.com>.
It can pretty much be used as-is, _except_....
you'll find one or more entries in your request handlers like:
<str name="df">_text_</str>
Change "_text_" to something in your schema, that's the default search
field if you don't field-qualify your search terms.
Note that if you take out, for instance, all of your non-english
fieldTypes, you can also remove most of the stuff under the /lang
folder.
I essentially always test this out on a local, stand-alone instance
until I can index a few documents and query them, it's faster than
always having to remember to move them to ZooKeeper....
Best,
Erick
On Wed, Feb 7, 2018 at 7:14 PM, Pratik Patel <pr...@semandex.net> wrote:
> Hey Eric, thanks for the clarification! What about solrConfig.xml file?
> Sure, it should be customized to suit one's needs but can it be used as a
> base or is it best to create one from scratch ?
>
> Thanks,
> Pratik
>
> On Wed, Feb 7, 2018 at 5:29 PM, Erick Erickson <er...@gmail.com>
> wrote:
>
>> That's really the point of the default managed-schema, to be a base
>> you use for your customizations. In fact, I often _remove_ most of the
>> fields (and especially fieldTypes) that I don't need. This includes
>> dynamic fields, copyFields and the like.
>>
>> Sometimes it's actually easier, though, to just start all over.
>>
>> BTW, do not delete any field that begins and ends with an underscore,
>> e.g. _version_ unless you know exactly what the consequences are....
>>
>> Best,
>> Erick
>>
>> On Wed, Feb 7, 2018 at 2:59 PM, Pratik Patel <pr...@semandex.net> wrote:
>> > Hello all,
>> >
>> > I have added some fields to default managed-schema file. I was wondering
>> if
>> > it is safe to take default managed-schema file as is and add your own
>> > fields to it in production. What is the best practice for this? As I
>> > understand, it should be safe to use default schema as base if documents
>> > that are going to be indexed in solr will only have newly defined fields
>> in
>> > it. In fact, it helps because the common field types are already defined
>> in
>> > default schema which can be re-used. I looked through the documentation
>> but
>> > couldn't find the answer and more clarity on this would be helpful.
>> >
>> > Is it safe to use default managed-schema file as base add your own fields
>> > to it?
>> >
>> > Thanks,
>> > Pratik
>>
Re: Best Practice about solr cloud schema
Posted by Pratik Patel <pr...@semandex.net>.
Hey Eric, thanks for the clarification! What about solrConfig.xml file?
Sure, it should be customized to suit one's needs but can it be used as a
base or is it best to create one from scratch ?
Thanks,
Pratik
On Wed, Feb 7, 2018 at 5:29 PM, Erick Erickson <er...@gmail.com>
wrote:
> That's really the point of the default managed-schema, to be a base
> you use for your customizations. In fact, I often _remove_ most of the
> fields (and especially fieldTypes) that I don't need. This includes
> dynamic fields, copyFields and the like.
>
> Sometimes it's actually easier, though, to just start all over.
>
> BTW, do not delete any field that begins and ends with an underscore,
> e.g. _version_ unless you know exactly what the consequences are....
>
> Best,
> Erick
>
> On Wed, Feb 7, 2018 at 2:59 PM, Pratik Patel <pr...@semandex.net> wrote:
> > Hello all,
> >
> > I have added some fields to default managed-schema file. I was wondering
> if
> > it is safe to take default managed-schema file as is and add your own
> > fields to it in production. What is the best practice for this? As I
> > understand, it should be safe to use default schema as base if documents
> > that are going to be indexed in solr will only have newly defined fields
> in
> > it. In fact, it helps because the common field types are already defined
> in
> > default schema which can be re-used. I looked through the documentation
> but
> > couldn't find the answer and more clarity on this would be helpful.
> >
> > Is it safe to use default managed-schema file as base add your own fields
> > to it?
> >
> > Thanks,
> > Pratik
>
Re: Best Practice about solr cloud schema
Posted by Erick Erickson <er...@gmail.com>.
That's really the point of the default managed-schema, to be a base
you use for your customizations. In fact, I often _remove_ most of the
fields (and especially fieldTypes) that I don't need. This includes
dynamic fields, copyFields and the like.
Sometimes it's actually easier, though, to just start all over.
BTW, do not delete any field that begins and ends with an underscore,
e.g. _version_ unless you know exactly what the consequences are....
Best,
Erick
On Wed, Feb 7, 2018 at 2:59 PM, Pratik Patel <pr...@semandex.net> wrote:
> Hello all,
>
> I have added some fields to default managed-schema file. I was wondering if
> it is safe to take default managed-schema file as is and add your own
> fields to it in production. What is the best practice for this? As I
> understand, it should be safe to use default schema as base if documents
> that are going to be indexed in solr will only have newly defined fields in
> it. In fact, it helps because the common field types are already defined in
> default schema which can be re-used. I looked through the documentation but
> couldn't find the answer and more clarity on this would be helpful.
>
> Is it safe to use default managed-schema file as base add your own fields
> to it?
>
> Thanks,
> Pratik