And so I think. Actually I hope that I can do something like that: 1) tell the Solr to prepare for searching 2) start my very fast filtering routine 3) send asynchronoussly IDs of filtered documents to the Solr and expect that Solr is ranging them in the parallel 4) get the result quickly
On 11 October 2010 21:25, Gora Mohanty <g...@mimirtech.com> wrote: > On Mon, Oct 11, 2010 at 8:20 PM, Sergey Bartunov <sbos....@gmail.com> wrote: >> Whether it will be enough effective if the subset is really large? > [...] > > If the subset of IDs is large, and disjoint (so that you cannot use ranges), > the query might look ugly, but generating it should not be much of a > problem if you are using some automated method to create the query. > > If you mean whether it will be efficient enough, the only way is to try > it out, and measure performance. Offhand, I do not think that it should > increase the query response time by a lot. > > Regards, > Gora >
