I added concurrent_writes: 48
to the cqlstress-example.yaml, which is recommended as No. of cores * 8. But no improvement is made when trying 1000000 writes test. Since cassandra.yaml specifies concurrent_writes, but cqlstress-example.yaml doesn't. BTW, the following warning was shown as the first line of the test. WARN 10:25:44 You listed localhost/0:0:0:0:0:0:0:1:9042 in your contact points, but it wasn't found in the control host's system.peers at startup how to fix it? many thanks On 7 July 2016 at 09:45, Stone Fang <cnstonef...@gmail.com> wrote: > there is a simple config file for testing."cqlstress-example.yaml" under > the tool folder. > you can customize this file to achieve your test > > stone. > > On Thu, Jul 7, 2016 at 4:28 PM, Daiyue Weng <daiyuew...@gmail.com> wrote: > >> Hi, I am wondering what's the default .yaml file that cassandra-stress >> uses when testing writes and reads when command 'profile=' is not >> specified. Is it the cassandra.yaml? Does it affect the performance of >> cassandra-stress test by modifying/tuning it? >> >> ps.I am running cassandra instances on Linux, the path to >> cassandra.yaml that I found is >> >> /etc/cassandra/cassandra.yaml >> >> thanks >> > >
