tag:blogger.com,1999:blog-9149523927864751087.post2951220323176995002..comments2024-03-26T09:43:01.052-07:00Comments on Small Datum: Examining performance for MongoDB and the insert benchmarkMark Callaghanhttp://www.blogger.com/profile/09590445221922043181noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-9149523927864751087.post-91251467637444140532018-02-01T06:58:35.034-08:002018-02-01T06:58:35.034-08:00I am sure that WT is faster than MongoRocks for so...I am sure that WT is faster than MongoRocks for some workloads. But I have a few suggestions:<br />1) I take a lot of time to explain the workloads that I use. You have provide almost no information about the workload that you used.<br />2) There are many dimensions by which something can be better. These include database size, write efficiency and response time. When you write that WT is faster I assume you are describing response time, but you make no mention of space and write efficiency. Space efficiency determines how much SSD you need to buy to store your database. Write efficiency determines how long that SSD will last. While WT frequently wins on response time, MongoRocks usually wins on space and write efficiency.<br /><br />Regardless, I like both WT and MongoRocks.Mark Callaghanhttps://www.blogger.com/profile/09590445221922043181noreply@blogger.comtag:blogger.com,1999:blog-9149523927864751087.post-64594596507836681812018-02-01T01:38:14.722-08:002018-02-01T01:38:14.722-08:00Great post.
Hi, Mark I have read most of your pos...Great post.<br /><br />Hi, Mark I have read most of your posts in Small Datum and learn a lot. <br /><br />I test MongoDB 3.4 with RocksDB and WiredTiger engine using sysbench with single thread, and found WiredTiger is 20% faster then RocksDB, and the result is reproducible. YoudongZhangnoreply@blogger.com