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 Ninad Raut <hb...@gmail.com> on 2009/08/06 08:54:45 UTC
Transfer of Index Vs HTTP GET Vs Embedded Solr -- Urgent Help
Hi,
I have a search engine on Solr. Also I have a remote web application which
will be using the Solr Indexes for search.
I have three scenarios:
1) Transfer the Indexes to the Remote Application.
- This will reduce load on the actual solr server and make seraches
faster.
- Need to write some code to transfer the index
- Need to double my effort to update,merge,optimize index
2)Use HTTP GET
- Will increase load on the Solr server
- No extra code needed for transfer
3) Embedded Serach
- Use SolrJ for querying
I want to know which is the best approach.
Regards,
Ninad Raut.
Re: Transfer of Index Vs HTTP GET Vs Embedded Solr -- Urgent Help
Posted by Ninad Raut <hb...@gmail.com>.
The you should consider replicating the index to the local intranet
and still run that it as a separate app.
Will it be the same master-slave replication?? If the master is
multicore, can I specifically replicate an index of a certain core ? Thanks
for the help.
2009/8/7 Noble Paul നോബിള് नोब्ळ् <no...@corp.aol.com>
> The you should consider replicating the index to the local intranet
> and still run that it as a separate app.
>
> On Fri, Aug 7, 2009 at 10:53 AM, Ninad Raut<hb...@gmail.com>
> wrote:
> > The remote web app will be accessing the Solr server via internet. Its
> not a
> > intranet setup.
> >
> > On Fri, Aug 7, 2009 at 10:19 AM, Walter Underwood <wunder@wunderwood.org
> >wrote:
> >
> >> About the first option, caches are more effective with more traffic, so
> ten
> >> front end servers using three Solr servers will have better caching and
> >> probably better overall performance than having separate search on all
> ten
> >> servers. You can even put an HTTP cache in there and get better caching.
> >>
> >> Cached HTTP responses are usually faster than accessing disc locally.
> >>
> >> You say you have a "remote web application". How remote? If the indexes
> are
> >> big, then copying them to a remote location is a lot of traffic.
> >>
> >> wunder
> >>
> >> On Aug 6, 2009, at 8:49 PM, Ninad Raut wrote:
> >>
> >> Hi Noble,
> >>> Can you explain a bit more on how to use Solr "out of the box". I am
> >>> looking at ways to design the UI for remote application quickly and
> with
> >>> less problems.
> >>> Also could you elaborate more on what can go wrong with the first
> option?
> >>> Thanks.
> >>>
> >>> 2009/8/6 Noble Paul നോബിള് नोब्ळ् <no...@corp.aol.com>
> >>>
> >>> On Thu, Aug 6, 2009 at 12:24 PM, Ninad Raut<
> hbase.user.ninad@gmail.com>
> >>>> wrote:
> >>>>
> >>>>> Hi,
> >>>>> I have a search engine on Solr. Also I have a remote web application
> >>>>>
> >>>> which
> >>>>
> >>>>> will be using the Solr Indexes for search.
> >>>>> I have three scenarios:
> >>>>> 1) Transfer the Indexes to the Remote Application.
> >>>>>
> >>>>> - This will reduce load on the actual solr server and make seraches
> >>>>> faster.
> >>>>> - Need to write some code to transfer the index
> >>>>> - Need to double my effort to update,merge,optimize index
> >>>>>
> >>>>> 2)Use HTTP GET
> >>>>>
> >>>>> - Will increase load on the Solr server
> >>>>> - No extra code needed for transfer
> >>>>>
> >>>> This by far is the best solution. ecause you do not have to do any
> >>>> work at all. It all works out of the box and that is what everyone
> >>>> uses
> >>>>
> >>>>>
> >>>>> 3) Embedded Serach
> >>>>>
> >>>>> - Use SolrJ for querying
> >>>>>
> >>>>> I want to know which is the best approach.
> >>>>> Regards,
> >>>>> Ninad Raut.
> >>>>>
> >>>>>
> >>>>
> >>>>
> >>>> --
> >>>> -----------------------------------------------------
> >>>> Noble Paul | Principal Engineer| AOL | http://aol.com
> >>>>
> >>>>
> >>
> >
>
>
>
> --
> -----------------------------------------------------
> Noble Paul | Principal Engineer| AOL | http://aol.com
>
Re: Transfer of Index Vs HTTP GET Vs Embedded Solr -- Urgent Help
Posted by Noble Paul നോബിള് नोब्ळ् <no...@corp.aol.com>.
The you should consider replicating the index to the local intranet
and still run that it as a separate app.
On Fri, Aug 7, 2009 at 10:53 AM, Ninad Raut<hb...@gmail.com> wrote:
> The remote web app will be accessing the Solr server via internet. Its not a
> intranet setup.
>
> On Fri, Aug 7, 2009 at 10:19 AM, Walter Underwood <wu...@wunderwood.org>wrote:
>
>> About the first option, caches are more effective with more traffic, so ten
>> front end servers using three Solr servers will have better caching and
>> probably better overall performance than having separate search on all ten
>> servers. You can even put an HTTP cache in there and get better caching.
>>
>> Cached HTTP responses are usually faster than accessing disc locally.
>>
>> You say you have a "remote web application". How remote? If the indexes are
>> big, then copying them to a remote location is a lot of traffic.
>>
>> wunder
>>
>> On Aug 6, 2009, at 8:49 PM, Ninad Raut wrote:
>>
>> Hi Noble,
>>> Can you explain a bit more on how to use Solr "out of the box". I am
>>> looking at ways to design the UI for remote application quickly and with
>>> less problems.
>>> Also could you elaborate more on what can go wrong with the first option?
>>> Thanks.
>>>
>>> 2009/8/6 Noble Paul നോബിള് नोब्ळ् <no...@corp.aol.com>
>>>
>>> On Thu, Aug 6, 2009 at 12:24 PM, Ninad Raut<hb...@gmail.com>
>>>> wrote:
>>>>
>>>>> Hi,
>>>>> I have a search engine on Solr. Also I have a remote web application
>>>>>
>>>> which
>>>>
>>>>> will be using the Solr Indexes for search.
>>>>> I have three scenarios:
>>>>> 1) Transfer the Indexes to the Remote Application.
>>>>>
>>>>> - This will reduce load on the actual solr server and make seraches
>>>>> faster.
>>>>> - Need to write some code to transfer the index
>>>>> - Need to double my effort to update,merge,optimize index
>>>>>
>>>>> 2)Use HTTP GET
>>>>>
>>>>> - Will increase load on the Solr server
>>>>> - No extra code needed for transfer
>>>>>
>>>> This by far is the best solution. ecause you do not have to do any
>>>> work at all. It all works out of the box and that is what everyone
>>>> uses
>>>>
>>>>>
>>>>> 3) Embedded Serach
>>>>>
>>>>> - Use SolrJ for querying
>>>>>
>>>>> I want to know which is the best approach.
>>>>> Regards,
>>>>> Ninad Raut.
>>>>>
>>>>>
>>>>
>>>>
>>>> --
>>>> -----------------------------------------------------
>>>> Noble Paul | Principal Engineer| AOL | http://aol.com
>>>>
>>>>
>>
>
--
-----------------------------------------------------
Noble Paul | Principal Engineer| AOL | http://aol.com
Re: Transfer of Index Vs HTTP GET Vs Embedded Solr -- Urgent Help
Posted by Ninad Raut <hb...@gmail.com>.
The remote web app will be accessing the Solr server via internet. Its not a
intranet setup.
On Fri, Aug 7, 2009 at 10:19 AM, Walter Underwood <wu...@wunderwood.org>wrote:
> About the first option, caches are more effective with more traffic, so ten
> front end servers using three Solr servers will have better caching and
> probably better overall performance than having separate search on all ten
> servers. You can even put an HTTP cache in there and get better caching.
>
> Cached HTTP responses are usually faster than accessing disc locally.
>
> You say you have a "remote web application". How remote? If the indexes are
> big, then copying them to a remote location is a lot of traffic.
>
> wunder
>
> On Aug 6, 2009, at 8:49 PM, Ninad Raut wrote:
>
> Hi Noble,
>> Can you explain a bit more on how to use Solr "out of the box". I am
>> looking at ways to design the UI for remote application quickly and with
>> less problems.
>> Also could you elaborate more on what can go wrong with the first option?
>> Thanks.
>>
>> 2009/8/6 Noble Paul നോബിള് नोब्ळ् <no...@corp.aol.com>
>>
>> On Thu, Aug 6, 2009 at 12:24 PM, Ninad Raut<hb...@gmail.com>
>>> wrote:
>>>
>>>> Hi,
>>>> I have a search engine on Solr. Also I have a remote web application
>>>>
>>> which
>>>
>>>> will be using the Solr Indexes for search.
>>>> I have three scenarios:
>>>> 1) Transfer the Indexes to the Remote Application.
>>>>
>>>> - This will reduce load on the actual solr server and make seraches
>>>> faster.
>>>> - Need to write some code to transfer the index
>>>> - Need to double my effort to update,merge,optimize index
>>>>
>>>> 2)Use HTTP GET
>>>>
>>>> - Will increase load on the Solr server
>>>> - No extra code needed for transfer
>>>>
>>> This by far is the best solution. ecause you do not have to do any
>>> work at all. It all works out of the box and that is what everyone
>>> uses
>>>
>>>>
>>>> 3) Embedded Serach
>>>>
>>>> - Use SolrJ for querying
>>>>
>>>> I want to know which is the best approach.
>>>> Regards,
>>>> Ninad Raut.
>>>>
>>>>
>>>
>>>
>>> --
>>> -----------------------------------------------------
>>> Noble Paul | Principal Engineer| AOL | http://aol.com
>>>
>>>
>
Re: Transfer of Index Vs HTTP GET Vs Embedded Solr -- Urgent Help
Posted by Walter Underwood <wu...@wunderwood.org>.
About the first option, caches are more effective with more traffic,
so ten front end servers using three Solr servers will have better
caching and probably better overall performance than having separate
search on all ten servers. You can even put an HTTP cache in there and
get better caching.
Cached HTTP responses are usually faster than accessing disc locally.
You say you have a "remote web application". How remote? If the
indexes are big, then copying them to a remote location is a lot of
traffic.
wunder
On Aug 6, 2009, at 8:49 PM, Ninad Raut wrote:
> Hi Noble,
> Can you explain a bit more on how to use Solr "out of the box". I am
> looking at ways to design the UI for remote application quickly and
> with
> less problems.
> Also could you elaborate more on what can go wrong with the first
> option?
> Thanks.
>
> 2009/8/6 Noble Paul നോബിള് नोब्ळ्
> <no...@corp.aol.com>
>
>> On Thu, Aug 6, 2009 at 12:24 PM, Ninad Raut<hbase.user.ninad@gmail.com
>> >
>> wrote:
>>> Hi,
>>> I have a search engine on Solr. Also I have a remote web application
>> which
>>> will be using the Solr Indexes for search.
>>> I have three scenarios:
>>> 1) Transfer the Indexes to the Remote Application.
>>>
>>> - This will reduce load on the actual solr server and make seraches
>>> faster.
>>> - Need to write some code to transfer the index
>>> - Need to double my effort to update,merge,optimize index
>>>
>>> 2)Use HTTP GET
>>>
>>> - Will increase load on the Solr server
>>> - No extra code needed for transfer
>> This by far is the best solution. ecause you do not have to do any
>> work at all. It all works out of the box and that is what everyone
>> uses
>>>
>>> 3) Embedded Serach
>>>
>>> - Use SolrJ for querying
>>>
>>> I want to know which is the best approach.
>>> Regards,
>>> Ninad Raut.
>>>
>>
>>
>>
>> --
>> -----------------------------------------------------
>> Noble Paul | Principal Engineer| AOL | http://aol.com
>>
Re: Transfer of Index Vs HTTP GET Vs Embedded Solr -- Urgent Help
Posted by Ninad Raut <hb...@gmail.com>.
Hi Noble,
Can you explain a bit more on how to use Solr "out of the box". I am
looking at ways to design the UI for remote application quickly and with
less problems.
Also could you elaborate more on what can go wrong with the first option?
Thanks.
2009/8/6 Noble Paul നോബിള് नोब्ळ् <no...@corp.aol.com>
> On Thu, Aug 6, 2009 at 12:24 PM, Ninad Raut<hb...@gmail.com>
> wrote:
> > Hi,
> > I have a search engine on Solr. Also I have a remote web application
> which
> > will be using the Solr Indexes for search.
> > I have three scenarios:
> > 1) Transfer the Indexes to the Remote Application.
> >
> > - This will reduce load on the actual solr server and make seraches
> > faster.
> > - Need to write some code to transfer the index
> > - Need to double my effort to update,merge,optimize index
> >
> > 2)Use HTTP GET
> >
> > - Will increase load on the Solr server
> > - No extra code needed for transfer
> This by far is the best solution. ecause you do not have to do any
> work at all. It all works out of the box and that is what everyone
> uses
> >
> > 3) Embedded Serach
> >
> > - Use SolrJ for querying
> >
> > I want to know which is the best approach.
> > Regards,
> > Ninad Raut.
> >
>
>
>
> --
> -----------------------------------------------------
> Noble Paul | Principal Engineer| AOL | http://aol.com
>
Re: Transfer of Index Vs HTTP GET Vs Embedded Solr -- Urgent Help
Posted by Noble Paul നോബിള് नोब्ळ् <no...@corp.aol.com>.
On Thu, Aug 6, 2009 at 12:24 PM, Ninad Raut<hb...@gmail.com> wrote:
> Hi,
> I have a search engine on Solr. Also I have a remote web application which
> will be using the Solr Indexes for search.
> I have three scenarios:
> 1) Transfer the Indexes to the Remote Application.
>
> - This will reduce load on the actual solr server and make seraches
> faster.
> - Need to write some code to transfer the index
> - Need to double my effort to update,merge,optimize index
>
> 2)Use HTTP GET
>
> - Will increase load on the Solr server
> - No extra code needed for transfer
This by far is the best solution. ecause you do not have to do any
work at all. It all works out of the box and that is what everyone
uses
>
> 3) Embedded Serach
>
> - Use SolrJ for querying
>
> I want to know which is the best approach.
> Regards,
> Ninad Raut.
>
--
-----------------------------------------------------
Noble Paul | Principal Engineer| AOL | http://aol.com