The goal of this series on maximizing the performance
of Linux systems is to offer some relatively simple yet highly
effective tuning ideas. Moreover, all such suggestions must adhere
to the KISS (keep it simple, stupid) ideal. All tuning techniques
must be both easily implemented and easily measured. Our ideal
technique must fit this model: perform a baseline benchmark,
implement the proposed change, perform a comparative benchmark and
end with a big smile.
In part one of this series, we discussed enterprise-level
performance with databases like Oracle, and we obtained over a
1000% improvement in database load times and a 50% improvement in
the number of transactions per second. Although some of the
techniques discussed in part one are overly basic in nature (i.e.,
DBA 101 and Linux System Admin 101), it was necessary to start at
the beginning so as not to miss any low-hanging fruits along the
way. With the foundation in place, it's now time to tackle slightly
more interesting and challenging tuning concepts.
This time we'll examine Linux multi-devices (RAID), logical
volume managers (LVM), cooked versus raw filesystems and regular
versus journalized filesystems. I guarantee that some of the
enclosed benchmark results and conclusions will surprise you.