最近开发的一个模块需要根据机房各节点的负载情况(如网卡IO、load average等指标)做任务调度,刚开始对Linux机器load average这项指标不是很清楚,经过调研,终于搞清楚了其计算方法和影响因素,作为笔记,记录于此。
1. load average
当在shell终端键入top命令时,默认情况下,在输出内容的第一行会有load average这项指标值,如下所示:
top - 19:10:32 up 626 days, 4:58, 1 user, load average: 7.74, 5.62, 6.51
Tasks: 181 total, 8 running, 173 sleeping, 0 stopped, 0 zombie
Cpu(s): 4.0% us, 0.5% sy, 0.0% ni, 95.4% id, 0.0% wa, 0.0% hi, 0.0% si
同样,输入uptime命令,load average也会被输出:
19:15:10 up 129 days, 5:12, 15 users, load average: 0.01, 0.09, 0.05
根据man uptime的说明可知, load average包含的3个值分别表示past 1, 5 and 15 minutes内的系统平均负载。
那么,这3个值是怎么计算出来的?下面从Linux源码中寻找答案。
2. linux机器load average的计算过程
wikipedia在对load的解释(
参见这里)中,提到了linux系统对load的计算方法,为亲自验证,我check了linux源码(linux kernel 2.6.9)中的相关代码,自顶向下的验证过程如下。
在源码树kernel/timer.c文件中,计算系统load的函数代码如下:
// 源码树路径:kernel/timer.c
/*
* Hmm.. Changed this, as the GNU make sources (load.c) seems to
* imply that avenrun[] is the standard name for this kind of thing.
* Nothing else seems to be standardized: the fractional size etc
* all seem to differ on different machines.
*
* Requires xtime_lock to access.
*/
unsigned long avenrun[3]; /*
* calc_load - given tick count, update the avenrun load estimates.
* This is called while holding a write_lock on xtime_lock.
*/
static inline void calc_load(unsigned long ticks)
{
unsigned long active_tasks; /* fixed-point */
static int count = LOAD_FREQ; count -= ticks;
if (count < 0) {
count += LOAD_FREQ;
active_tasks = count_active_tasks();
CALC_LOAD(avenrun[0], EXP_1, active_tasks);
CALC_LOAD(avenrun[1], EXP_5, active_tasks);
CALC_LOAD(avenrun[2], EXP_15, active_tasks);
}
}
从上面的代码可知,定义的数组avenrun[]包含3个元素,分别用于存放past 1, 5 and 15 minutes的load average值。calc_load则是具体的计算函数,其参数ticks表示采样间隔。函数体中,获取当前的活跃进程数(active tasks),然后以其为参数,调用CALC_LOAD分别计算3种load average。
沿着函数调用链,可以看到count_active_tasks()定义如下(也在kernel/timer.c文件中):
/*
* Nr of active tasks - counted in fixed-point numbers
*/
static unsigned long count_active_tasks(void)
{
return (nr_running() + nr_uninterruptible()) * FIXED_1;
}
由源码可见,count_active_tasks()返回当前的活跃进程数,其中活跃进程包括:1)当前正在运行的进程(nr_running);2)不可中断的sleeping进程(如正在执行IO操作的被挂起进程)。
关于nr_running进程和nr_uninterruptible进程的计算方法,可以在源码树kernel/schde.c中看到相关代码:
// 源码树路径:kernel/sched.c
/*
* nr_running, nr_uninterruptible and nr_context_switches:
*
* externally visible scheduler statistics: current number of runnable
* threads, current number of uninterruptible-sleeping threads, total
* number of context switches performed since bootup.
*/
unsigned long nr_running(void)
{
unsigned long i, sum = 0; for (i = 0; i < NR_CPUS; i++)
sum += cpu_rq(i)->nr_running; return sum;
} unsigned long nr_uninterruptible(void)
{
unsigned long i, sum = 0; for_each_cpu(i)
sum += cpu_rq(i)->nr_uninterruptible; return sum;
}
继续沿着函数调用链查看,可在include/linux/sched.h中看到CALC_LOAD的定义:
// 源码树路径:include/linux/sched.h
/*
* These are the constant used to fake the fixed-point load-average
* counting. Some notes:
* - 11 bit fractions expand to 22 bits by the multiplies: this gives
* a load-average precision of 10 bits integer + 11 bits fractional
* - if you want to count load-averages more often, you need more
* precision, or rounding will get you. With 2-second counting freq,
* the EXP_n values would be 1981, 2034 and 2043 if still using only
* 11 bit fractions.
*/
extern unsigned long avenrun[]; /* Load averages */ #define FSHIFT 11 /* nr of bits of precision */
#define FIXED_1 (1<<FSHIFT) /* 1.0 as fixed-point */
#define LOAD_FREQ (5*HZ) /* 5 sec intervals */
#define EXP_1 1884 /* 1/exp(5sec/1min) as fixed-point */
#define EXP_5 2014 /* 1/exp(5sec/5min) */
#define EXP_15 2037 /* 1/exp(5sec/15min) */ #define CALC_LOAD(load,exp,n) \
load *= exp; \
load += n*(FIXED_1-exp); \
load >>= FSHIFT;
可以看到,CALC_LOAD是一个宏定义,load average的值与3个参数相关,但若只考虑某1项指标值(如past 5 minutes的load average),则该值只受当前活跃进程数(active tasks)的影响,而活跃进程数包括两种:当前正在运行的进程和不可中断的挂起进程。
这符合我的观察结果:三台硬件配置相同的linux机器(8 cup, 16GB memory, 1.8T disk),在当前总进程数相差不多(均为170+)的情况下,其中1台机器有1个普通进程(这里的"普通"是指既非CPU型又非IO型)在运行,其余均sleeping;第2台机器有5个cpu型进程,cpu占用率均达到99%,其余进程sleeping;第3台机器2个进程读写硬盘,其余sleeping。很明显地可以看到:第3台机器的load average指标的3个值均为最大,第2台机器次之,第1台机器的3个值均接近0。
由此,还可以推断出:与running类型的进程相比,uninterruptible类型的进程(如正在进行IO操作)对系统load的影响较大。(
注:该推断暂无数据或代码支撑,若有误,欢迎指正)
3. 理解load average背后的含义
上面介绍了load average的概念及linux系统对该指标的计算过程,那么,这个指标值到底怎么解读呢?这篇文章给出了详细且形象的说明,此处不再赘述。
【参考资料】
1. wikipedia: Load (computing)
2. linux源码(内核版本2.6.9)
3.
Understanding Linux CPU Load - when should you be worried?
================== EOF ===================