Well.. Just FYI.. 10 lakhs in each normalized tables.. Query time to fetch If linked together won't be big? And data is growing. On 25 Jun 2013 20:49, "Walter Underwood" <wun...@wunderwood.org> wrote:
> With only 10K records (lahks), a regular RDBMS should be just fine. I > don't see any need for Solr with a small dataset like that. > > Increase the caches sizes on your RDBMS so that all the tables fit in > memory. Even with 10Kbytes per record, that is only 100Mbytes of data. > > wunder > > On Jun 25, 2013, at 7:28 AM, Otis Gospodnetic wrote: > > > Hi Pradeep, > > > > 5-6 hours between email and "Any help?" == "not enough patience" :) > > > > The advantage of something like Solr over RDBMS with star schema may > > be that it is easier to scale horizontally than MySQL, or at least > > that was the case I last looked at horizontal RDBMS partitioning. But > > if you are planning to have both RDBMS w/ star schema for reporting > > AND Solr for reporting (via facets and such), that seems redundant. > > You need just one of these two. > > > > Otis > > -- > > Solr & ElasticSearch Support -- http://sematext.com/ > > Performance Monitoring -- http://sematext.com/spm > > > > > > > > On Tue, Jun 25, 2013 at 9:50 AM, pradeep kumar <pradeepkuma...@gmail.com> > wrote: > >> Any help? > >> On 25 Jun 2013 13:35, "pradeep kumar" <pradeepkuma...@gmail.com> wrote: > >> > >>> Sure, > >>> > >>> First of all thanks a lot everyone for very quick reply. > >>> > >>> We have a Ordering system which has a lakhs of records so far in a > >>> normalized RDBMS tables, say Order, Item, Details etc. We are planning > to > >>> have a offline database (star schema) and develop reports, data > analytical > >>> charts with drill down, dashboard with data from offline database. > >>> > >>> I am planning to propose solr as a solution instead of a offline > database > >>> ie through DIH to import data from DB into solr indexes. Since Solr > indexes > >>> are stored denormalized manner and querying is faster, faceting > search, i > >>> assumed that solr can be used to solve my requirement. > >>> > >>> Please correct me if i am wrong. > >>> > >>> Thanks, > >>> Pradeep > >>> > >>> > >>> > >>> > >>> On Tue, Jun 25, 2013 at 3:43 AM, Otis Gospodnetic < > >>> otis.gospodne...@gmail.com> wrote: > >>> > >>>> Yeah, perhaps.... yet people keep using it for this. So, Pradeep, it > >>>> may work for you and if you share some numbers with us we may be able > >>>> to tell you "no way" or "very likely OK". :) > >>>> > >>>> > >>>> Otis > >>>> -- > >>>> Solr & ElasticSearch Support -- http://sematext.com/ > >>>> Performance Monitoring -- http://sematext.com/spm > >>>> > >>>> > >>>> > >>>> On Mon, Jun 24, 2013 at 4:14 PM, Walter Underwood < > wun...@wunderwood.org> > >>>> wrote: > >>>>> I expect it won't be fast enough for general use. Most analytics > stores > >>>> implement functions inside the server to aggregate large amounts of > data. > >>>> There is always some query that returns the whole database in order to > >>>> calculate an average. > >>>>> > >>>>> I'm sure it will work fine for some things and for small data sets, > but > >>>> it probably won't scale for most real analytics applications. > >>>>> > >>>>> wunder > >>>>> > >>>>> On Jun 24, 2013, at 12:47 PM, pradeep kumar wrote: > >>>>> > >>>>>> Hello everyone, > >>>>>> > >>>>>> Apart from text search, can we use Solr as data store to serve data > to > >>>> form > >>>>>> analytics with drilldown charts or charts to add as widgets on > >>>> dashboards? > >>>>>> > >>>>>> Any suggestion, examples? > >>>>>> > >>>>>> Thanks, > >>>>>> Pradeep > >>>>> > >>>>> > >>>>> > >>>>> > >>>> > >>> > >>> > > -- > Walter Underwood > wun...@wunderwood.org > > > >
