Description
During loading of the example buckets (beer-sample) the chart shows that my cluster had a max at roughly 250 ops/sec (at a second run locally on my mac laptop running on SSD with gigs of memory available I got that up to ~500 ops/sec), and running "cbc-pillowfight" to the server afterwards that was increased to 15k ops/sec (single threaded asynchronous ops), and mcbasher kicked it up to 50k ops/sec. I think its safe to assume that its not the server that is slowing down the load of the bucket.
Being just 7k objects in beer-sample, I would expect the entire bucket to be loaded in less than a second (time used to populate the data, not the entire time to create the bucket, load it, create views etc).
For people evaluating Couchbase it is the first thing they might see after installing the product, and I wouldn't be very amazed of its performance if I saw this.
To reproduce the issue: run cluster_run -n 1, don't install the beer-sample and don't let default-bucket occupy the entire memory when you click through the configuration wizard. Then press settings and import the beer-sample and go back to the overview and watch the graph to see the import performance..
http://www.couchbase.com/issues/browse/MB-10714 should be fixed at the same time.