On 26/02/2016 10:17, Ancona Francesco wrote: > Hello, > so if Lucene or Solr have a problem or are busy for some reasons, we can't > search anything, if i understand.
As already pointed out out by Michael, other that you plan to leverage Solr specific features I would stick to lucene. The query engine is able to serve queries while it indexs or re-indexes so you should not have failures on such side. Of course by the fact that Lucene, for example, is asynchronous the information returned by it may be slightly out do date as the new updates were not being processed already. > So, i imagine, we have to be very careful to the search engine that is a > potential single point of failure if it goes down or if loose index and so it > has to make a full reindex. Not familiar with the Solr side. But I guess that yes, if the Solr server goes down it will be not able to serve queries. With all the other approaches indexes are embedded in the product (oak-core) so if it's not available then something major has happened. If you miss an index definition; it means someone deleted it and it's a different things. > What kind of topology (application and search engine) do you suggest to > mitigate this problem ? As already said, other than you plan to deliver any Solr specific feature I would stick initially with Lucene. If you have to leverage solr I would go with usual approach for balancing and multiple solr servers for resilience. Davide
