我用这个提交了我的jar文件：
-verbose:gc -XX:+PrintGCDetails -XX:+PrintGCTimeStamps -XX:+UseCompressedOops您可以看到我按照官方文档中的建议添加了stdout部分。但是，当我在Web UI中查看从属设备“stderr和log4j.properties.template时，我看不到垃圾收集。
是否阻止显示gc日志的log4j设置？我的spark conf目录中只有import csv
from datetime import datetime
today = datetime.now().date().strftime("%d-%B-%Y")
while True:
with open("InputFile.csv") as d:
has_birthday = [user for user, birthday, gender in csv.reader(d,skipinitialspace=True) if birthday == today]
if len(has_birthday) < 1:
exit ()
else:
with open("InputFile.csv") as f:
has_birthday = [user for user, birthday, gender in csv.reader(f,skipinitialspace=True) if birthday == today]
with open("InputFile.csv") as g:
male = [user for user, birthday, gender in csv.reader(g, delimiter=',', skipinitialspace=True) if birthday == today and gender == 'M']
with open("InputFile.csv") as h:
female = [user for user, birthday, gender in csv.reader(h, delimiter=',', skipinitialspace=True) if birthday == today and gender == 'F']
if len(has_birthday) == 1 and male:
frm = "{} has his birthday today."
elif len(has_birthday) == 1 and female:
frm = "{} has her birthday today."
else:
frm = "{} have their birthdays today."
test = (frm.format(" and ".join([", ".join(has_birthday[:-1]), has_birthday[-1]]))) if len(has_birthday) > 1 else frm.format(" and ".join(has_birthday))
print test
文件。
关于什么是错的任何建议？谢谢。
答案 0 :(得分：0)
根据https://databricks.com/blog/2015/05/28/tuning-java-garbage-collection-for-spark-applications.html - GC日志写入：
$SPARK_HOME/work/$app_id/$executor_id/stdout
尝试使用此页面上建议的设置配置您的spark应用，并查看它是否按预期工作。