@timhaines yeah, i was just curious. :) glad you have a reliable way to do what you want to do.
2012-05-18 06:33:06@dizzyd Sure. The test is pretty close to just a time extension on your benchmark here at 17500 ops/sec tho right? http://t.co/jbtW4EGq
2012-05-18 06:35:32@timhaines no. What your benchmark does and what that blog entry describes are night and day different.
2012-05-18 06:46:46@timhaines thought you might be interested, pt-online-schema-change http://t.co/THaUs59c
2012-05-18 06:48:19@dizzyd mid-range SATA disks, "We inserted 100m items with numerically-sorted keys, using fairly small values of 100 bytes per item."
2012-05-18 06:48:47@dizzyd I'd genuinely like to understand the differences.
2012-05-18 06:49:14@timhaines you are benchmarking Riak; that was benchmarking LevelDB. You are doing gets/updates; that test is only inserts (no reads at all)
2012-05-18 07:01:01@dizzyd Okay. I was benchmarking inserts only too. I totally missed that was LevelDB only without Riak though. Thank you for clarifying.
2012-05-18 07:04:24