Hi,
I'm running FreeBSD 6.1, mysql 5.0.24a(binaries from mysql ab) and are
running some benchmarks using sysbench, what kind of performance should
i expect?
The machine is a Dell poweredge 1800 with the cerc raid controller (two
sata disks in raid1).
sysbench command:
sysbench --test=oltp --mysql-table-engine=innodb
--oltp-table-size=5000000 --num-threads=16
--mysql-socket=/tmp/mysql.sock --mysql-user=abc--mysql-password=def
--oltp-test-mode=complex --max-requests=10000 run
On a single disk (not using the cerc controller) I get the following
results:
OLTP test statistics:
queries performed:
read: 140000
write: 50000
other: 20000
total: 210000
transactions: 10000 (413.70 per sec.)
deadlocks: 0 (0.00 per sec.)
read/write requests: 190000 (7860.35 per sec.)
other operations: 20000 (827.41 per sec.)
Test execution summary:
total time: 24.1719s
total number of events: 10000
total time taken by event execution: 384.3010
per-request statistics:
min: 0.0061s
avg: 0.0384s
max: 0.3780s
approx. 95 percentile: 0.0752s
Threads fairness:
events (avg/stddev): 625.0000/16.48
execution time (avg/stddev): 24.0188/0.01
I guess the disk cache is causing the many transactions/s?
On a raid1 using the cerc hw raid controller:
OLTP test statistics:
queries performed:
read: 140140
write: 50023
other: 20010
total: 210173
transactions: 10000 (53.48 per sec.)
deadlocks: 10 (0.05 per sec.)
read/write requests: 190163 (1016.93 per sec.)
other operations: 20010 (107.01 per sec.)
Test execution summary:
total time: 186.9973s
total number of events: 10000
total time taken by event execution: 2988.9251
per-request statistics:
min: 0.0087s
avg: 0.2989s
max: 186.9913s
approx. 95 percentile: 0.0499s
There a quite big difference between the single disk and the raid1,
should i expect more or is this normal?
// Fredrik Carlsson
--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
