Comparison between Sedna, eXist and BaseX?
2014-05-17 18:47:20 GMT
While interesting, I am always skeptical of these types of analysis. 1. Tuning: a. Sedna is written in C/C++, the other 2 systems are Java based. Java based systems always require some type of Java VM option tuning based on the problem space. b. Sedna in a production environment also requires some tuning. I have spent effort in turning Sedna and eXistDb. c. Including tuning options (outside of indexing) is important. 2. Alternatives. There are also other free options, such as DB2 express. I've talked to several people that prefer DB2 express for XQuery. 3. Test scenarios a. Are the XQuery drivers equivalent? Did you use REST for eXistDB & xBase and not Sedna? Sedna driver is lean and mean, REST has some inherit overhead. b. All the documents are stored as separate documents? What about the scenario of a single <Outletinfos> parent document with Outletinfo documents as children? I've pushed a single 2 GB XML document, with Outletinfo type documents as children. c. What about batch updating? d. Is the full test suite available for review? 4. The project activity test(Continue reading)