这是一个可重复的例子:
library(doMC)
library(doParallel)
registerDoMC(4)
timing <- system.time( fitall <- foreach(i=1:1000, .combine = "c") %dopar% {
print(i)
})
我启动R
并查看流程表:
> system("ps -efl")
F S UID PID PPID C PRI NI ADDR SZ WCHAN STIME TTY TIME CMD
4 S chbr 1 0 5 80 0 - 21399 wait 10:58 ? 00:00:00 /usr/local/lib/R/bin/exec/R --no-save --no-restore
0 S chbr 9 1 0 80 0 - 1113 wait 10:58 ? 00:00:00 sh -c ps -efl
0 R chbr 10 9 0 80 0 - 4294 - 10:58 ? 00:00:00 ps -efl
如果我使用前面提到的简单for循环doMC
或doParallel
,则会留下僵尸进程。运行循环后输出ps -efl
:
> system("ps -efl")
F S UID PID PPID C PRI NI ADDR SZ WCHAN STIME TTY TIME CMD
4 S chbr 1 0 4 80 0 - 25256 wait 11:00 ? 00:00:00 /usr/local/lib/R/b
1 Z chbr 10 1 0 80 0 - 0 exit 11:00 ? 00:00:00 [R] <defunct>
0 S chbr 12 1 0 80 0 - 1113 wait 11:00 ? 00:00:00 sh -c ps -efl
0 R chbr 13 12 0 80 0 - 4294 - 11:00 ? 00:00:00 ps -efl
如果我重复循环而不再发出registerDoMC(4)
则不会创建额外的僵尸进程。但是,如果我发出registerDoMC(4)
,则会创建一个额外的僵尸进程:
> system("ps -efl")
F S UID PID PPID C PRI NI ADDR SZ WCHAN STIME TTY TIME CMD
4 S chbr 1 0 0 80 0 - 25554 wait 11:00 ? 00:00:01 /usr/local/lib/R/b
1 Z chbr 21 1 0 80 0 - 0 exit 11:02 ? 00:00:00 [R] <defunct>
1 Z chbr 22 1 0 80 0 - 0 exit 11:02 ? 00:00:00 [R] <defunct>
0 S chbr 26 1 0 80 0 - 1113 wait 11:03 ? 00:00:00 sh -c ps -efl
0 R chbr 27 26 0 80 0 - 4294 - 11:03 ? 00:00:00 ps -efl
我认为它可能是doMC
,它正在做一些不应该做的事情。如果doMC导致这种情况,有办法阻止doMC
让僵尸进程落后吗? (stopCluster()
不起作用,因为首先没有创建集群。)
> sessionInfo()
R Under development (unstable) (2014-08-16 r66404)
Platform: x86_64-unknown-linux-gnu (64-bit)
locale:
[1] LC_CTYPE=en_IE.UTF-8 LC_NUMERIC=C
[3] LC_TIME=en_IE.UTF-8 LC_COLLATE=en_IE.UTF-8
[5] LC_MONETARY=en_IE.UTF-8 LC_MESSAGES=en_IE.UTF-8
[7] LC_PAPER=en_IE.UTF-8 LC_NAME=C
[9] LC_ADDRESS=C LC_TELEPHONE=C
[11] LC_MEASUREMENT=en_IE.UTF-8 LC_IDENTIFICATION=C
attached base packages:
[1] parallel stats graphics grDevices utils datasets methods
[8] base
other attached packages:
[1] doParallel_1.0.8 doMC_1.3.3 iterators_1.0.7 foreach_1.4.2
loaded via a namespace (and not attached):
[1] codetools_0.2-8 compiler_3.2.0
答案 0 :(得分:4)
这与foreach或doMC无关;正如Steve Weston在回答其他StackOverflow查询时指出的那样,doMC基本上只是mclapply的包装器,你可以看到通过简单调用mclapply创建的僵尸进程:
library(parallel)
mclapply(rep(5,4), rnorm)
在我的系统上,这会留下两个僵尸进程:
[richcalaway@richcalaway-pc ~]$ ps -efl | grep defunct
1 Z 1660945517 28701 28624 0 77 0 - 0 exit 12:00 pts/1 00:00:00 [R] <defunct>
1 Z 1660945517 28702 28624 0 78 0 - 0 exit 12:00 pts/1 00:00:00 [R] <defunct>
0 S 1660945517 28704 28308 0 78 0 - 15306 pipe_w 12:00 pts/2 00:00:00 grep defunct
在正常情况下,这些僵尸进程不会造成任何麻烦,并且当R会话结束时它们确实会消失。您可以使用doParallel和fork集群而不是使用doMC来避免它们。
干杯,
Rich Calaway
首席项目经理
革命分析