June 19, 2013

Post: Sphinx search performance optimization: attribute-based filters

…skewed boolean attribute “ancient” which identifies books published before year 1500 and after, and now I will run some different queries against the two: sphinx> select * from catalog where ancient = 1; | total_found | 1499266 | 14% | time | 0.552 | sphinx> select * …

Post: Sphinx: Going Beyond full text search

… of data handling hundreds of thousands of search queries per day. The count of forum posts being … this case as we’re doing kind of “boolean full text search”. Plus we can make index … forum from the global sphinx index. So in general we find parallel processing using sphinx pretty good solution …

Comment: Full text search for all MySQL Storage Engines

… underlying reason why this is not efficient. WHERE FT_QUERY_1 AND FT_QUERY_2 seems to be OK, but “OR” seems… helps someone. Also, reasonable results can be achieved by using BOOLEAN queries to restrict, and expressions to sort. AFAIK, natural FTS can… reasonable success with MySQL 5.1 and 100k+ product rows. Sphinx is better in some ways, but FT is fairly mature…

Comment: Full text search for all MySQL Storage Engines

…=date (in boolean mode sorting by date descending returning top 200 matches): — PROFILE — root: 8.81, self: 0.00 query_load_dir… no boolean search available in sphinx, defaulting to SPH_MATCH_ALL) returning top 200 matches): — PROFILE — root: 12.86, self: 0.00 query…: 0.04 query_match: 12.82, self: 11.51 read_hits: 1.31 ————— Even if the new SPH_MATCH_BOOLEAN is an…

Post: FaceBook Search, Search for social networks

… think this is serious limitation. But anyway if you do boolean search without phrase matching and relevance ranking it is rather… post sorting, meaning we simply use search engine to do “boolean search” (or other ranking) and than fetch first N results… some usages. We also already make sure Sphinx can be used efficiently to perform queries going beyond full text search. Such as…

Comment: Full text search for all MySQL Storage Engines

… SPH_MATCH_BOOLEAN if you do not care about relevance end just want to check if there is the match. Sphinx would… slow. About sphinx – it would be interesting to see which performance do you get if you’re using boolean match mode as it looks like what you need. Also benchmark using real queries from your site…

Comment: Full text search for all MySQL Storage Engines

7. peter: All results have been performed in boolean search mode. That is because in my case the client … there exist any other matches which still fulfill the search query because the fulltext engine has on its own made up… words” search the mysql fulltext in boolean mode takes more than 800 seconds on this hardware, sphinx takes approx 16 seconds, bitvector…

Comment: Full text search for all MySQL Storage Engines

… SQL query, for full text search systems results are often different, depending on relevance computing method. If you’ve been using sphinx… based search you should use boolean match mode – this bypasses all word position check logic. Note Sphinx at this moment builds index… pretty interesting to see what numbers do you get for boolean search in sphinx.

Comment: Full text search for all MySQL Storage Engines

… results have been performed in boolean search mode. Does that mean using -b command line option to Sphinx search utility? > fulltext engine…’t know what’s your experience with other engines, but Sphinx never does that. In it’s default mode, it will… case running time. However an average query won’t typically be that hard to evaluate. > sphinx takes approx 16 seconds, My thinking…

Comment: Sphinx search performance optimization: attribute-based filters

…explanations. This didn’t seem obvious from the Sphinx docs. It seems this is a side-effect…be able to simply disable). Too bad the query match(‘solar @ancientkey ancientkey_0′) can’t …’). By default, match(‘solar @ancientkey ancientkey_0′) uses boolean AND right? That’s why thought the optmization…