JVM如何获取当前容器的资源限制

最近同事说到Java的 ParallelGCThreads 参数,我翻了下jdk8的代码,发现 ParallelGCThreads 的参数默认值如下:

  • 如果cpu核心数目少于等于8,则GC线程数量和CPU数一致
  • 如果cpu核心数大于8,则前8个核,每个核心对应一个GC线;其他核,每8个核对应5个GC线程

但是被提醒,发现即使在分配4核的容器上,GC线程数也为38。然后就想到应该和容器的资源限制有关——jvm可能无法觉察到当前容器的资源限制。

翻了下代码,发现最新版本的java是能感知容器的资源限制的,就按照jdk版本再翻了下代码:

线上的jdk(jdk8u144)

写一个sleep 1000s的程序,然后运行:

./jdk1.8.0_144/bin/java -XX:+UseG1GC -XX:+ParallelRefProcEnabled Main 

然后查看GC线程数目:

$ jstack $pid | grep 'Parallel GC Threads' | wc -l
38

一算就知道物理机器有56个核心(8+(56-8)*5/8=38)

然后使用+PrintFlagsFinal看下参数:

$ ./jdk1.8.0_144/bin/java -XX:+UnlockDiagnosticVMOptions -XX:+PrintFlagsFinal  -version | grep ParallelGCThreads
uintx ParallelGCThreads          =38                     {product} 

看来jdk8u144并无法读取容器配额。

jdk 8u191

然后发现,从jdk 8u191版本开始,Java就可以读取容器配额了:

运行同样的程序:

./jre1.8.0_191/bin/java -XX:+UseG1GC -XX:+ParallelRefProcEnabled Main
$ jstack $pid | grep 'Parallel GC Threads' | wc -l
4

查看实际参数:

$ ./jre1.8.0_191/bin/java -XX:+UnlockDiagnosticVMOptions -XX:+PrintFlagsFinal  -version | grep ParallelGCThreads
uintx ParallelGCThreads         =4                   {product} 

另外,jdk 8u191引入了PrintContainerInfo参数:

$ ./jre1.8.0_191/bin/java -XX:+UnlockDiagnosticVMOptions -XX:+PrintContainerInfo -version
OSContainer::init: Initializing Container Support
Path to /memory.limit_in_bytes is /sys/fs/cgroup/memory/memory.limit_in_bytes
Memory Limit is: 10737418240
Path to /cpu.cfs_quota_us is /sys/fs/cgroup/cpu,cpuacct/cpu.cfs_quota_us
CPU Quota is: 400000
Path to /cpu.cfs_period_us is /sys/fs/cgroup/cpu,cpuacct/cpu.cfs_period_us
CPU Period is: 100000
Path to /cpu.shares is /sys/fs/cgroup/cpu,cpuacct/cpu.shares
CPU Shares is: 681
CPU Quota count based on quota/period: 4
CPU Share count based on shares: 1
OSContainer::active_processor_count: 4
Path to /cpu.cfs_quota_us is /sys/fs/cgroup/cpu,cpuacct/cpu.cfs_quota_us
CPU Quota is: 400000
Path to /cpu.cfs_period_us is /sys/fs/cgroup/cpu,cpuacct/cpu.cfs_period_us
CPU Period is: 100000
Path to /cpu.shares is /sys/fs/cgroup/cpu,cpuacct/cpu.shares
CPU Shares is: 681
CPU Quota count based on quota/period: 4
CPU Share count based on shares: 1
OSContainer::active_processor_count: 4
……
java version "1.8.0_191"
Java(TM) SE Runtime Environment (build 1.8.0_191-b12)
Java HotSpot(TM) 64-Bit Server VM (build 25.191-b12, mixed mode)

可以看到,获取的内存限制、可用CPU数目都是对的了。

如何获取容器资源配额呢?

结合这个日志和代码,我们也可以看到如何获取容器配额:

首先从/proc/self/mounts中读取对应的资源的mount位置,比如cpu就是在/sys/fs/cgroup/cpu,cpuacct

$ cat /proc/mounts  | grep -E -w '(cpu|memory)'
cgroup /sys/fs/cgroup/memory cgroup ro,nosuid,nodev,noexec,relatime,memory 0 0
cgroup /sys/fs/cgroup/cpu,cpuacct cgroup ro,nosuid,nodev,noexec,relatime,cpu,cpuacct 0 0

对于内存:

$ cat /sys/fs/cgroup/memory/memory.limit_in_bytes
10737418240 

对于cpu资源:

其一,可以通过quota/period来算:

$ cat /sys/fs/cgroup/cpu,cpuacct/cpu.cfs_period_us #单CPU总时间片配额,微秒 
100000
$ cat /sys/fs/cgroup/cpu,cpuacct/cpu.cfs_quota_us #时间片内,容器可占用的CPU时间
400000

比如上面就表示分配了4核。

也可以通过cpu.shares来获取:

$ cat /sys/fs/cgroup/cpu,cpuacct/cpu.shares
681

不过这个值是cpu占用份额,无法根据这个算出来可用cpu数量,所以基本没用…

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.