So your code is running slow.  How can you figure out what's wrong?

machine code
    -- sampling/statistical vs. complete/deterministic/instrumentation
	-- clock precision issues
	-- scheduling issues
	-- call delay issues (event occurs, then slight delay to make gettime
	    call)
    -- gprof
	-- compilation flags
	-- gprof & gmon.out
	-- other tools exist...
	-- timing info
	-- cumulative times
	-- call graphs
    -- profile, cProfile Python modules
    -- dtrace/instruments(mac dtrace)/netbeans dtrace
    -- performance counters
    -- time command

cpu overuse
    -- ps
    -- top

network
    -- ping
    -- netstat
    -- ntop
    -- ethereal
    -- congestion issues

disk/fs
    -- iostat
    -- iozone
    -- nfsstat

vm
    -- vmstat
    -- slabtop
    -- pmap
    -- /proc/meminfo

db
    -- changing cache sizes
    -- query optimization
    -- proper indexing
    -- 

framework issues (depend on slow web proxy)

mention sysinternals