We do not have anything that is designed to address this use-case, except maybe MemoryIndex if your use-case involves indexing a single document. I don't really have suggestions except making sure that the number of indices that you maintain at any point in time is bounded. Lucene indices typically some overhead which is usually negligible once amortized across many documents. But if your use-case is more about having tons of small indices, then it might become an issue (I'm thinking about memory usage and file handles in particular). You might also want to check whether Lucene is the rigtht tool for your problem.
Le mar. 12 juil. 2016 à 17:56, Abhinav Tiwari <abhinavtiwari.1...@gmail.com> a écrit : > Any help here? > > ---------- Forwarded message ---------- > From: Abhinav Tiwari <abhinavtiwari.1...@gmail.com> > Date: Fri, Jun 10, 2016 at 1:48 PM > Subject: Caching lucene index > To: java-user@lucene.apache.org > > > Hi Team, > > We have recently implemented a search feature in our web application using > Lucene where we perform dynamic indexing. We are looking for opportunities > of caching the index so that we do not have to perform dynamic indexing > again for the same query and query can be searched using the cached index > to save latency. To explain more in detail, we have implemented a search > where we index data dynamically for each request of search for a particular > Id. Once a request comes up, we fetch the details for the request and index > it. Now, we do not want to do the same thing for all the requests of the > same Id until a new request comes up for a new Id. That is why we are > looking for a way to cache index generated for each Id for some time. > Could you please provide suggestions around how can we cache such index? > > > Thanks, > Abhinav >
