Execute an SQL statement in the hive, I use the hive - hiveconf hive. Root. The logger=DEBUG, the console print out its log, as follows:
14/10/27 09:51:57 INFO ql. Driver: & lt; PERFLOG method=releaseLocks>
14/10/27 09:51:57 INFO ql. Driver: & lt;/PERFLOG method=releaseLocks start=1414374717140 end=1414374717140 duration=0 & gt;
14/10/27 09:51:57 INFO ql. Driver: & lt;/PERFLOG method=Driver. Run start=1414374717102 end=1414374717140 duration=38 & gt;
14/10/27 09:51:57 INFO CliDriver: Time seems: 0.038 seconds
Hive> Select settle_date, count (1) the from t_ods_cash_payment where settle_date_month='201410' group by settle_date order by settle_date;
14/10/27 09:54:01 INFO ql. Driver: & lt; PERFLOG method=Driver. Run>
14/10/27 09:54:01 INFO ql. Driver: & lt; PERFLOG method=TimeToSubmit>
14/10/27 09:54:01 INFO ql. Driver: & lt; PERFLOG method=compile>
May submit to the resourcemanager cli from client, it took 3 minutes time,
Then check the firewall (closed), the condition of HDFS and yarn is good, the child nodes running is normal (via a web page view), interface machine number of threads is transferred to the big enough, the machine of memory, CPU and disk load is not large, the network is normal, the configuration file is normal,
Then he himself wrote a wordcount, in cluster to perform, or submit the task process is slow, can eliminate the problem of hive seems to be the problem of yarn, but to restart a service or not, do you have encountered such a situation?
CodePudding user response:
How the configuration parameters of the yarn, feeling caused by allocating resourcesCodePudding user response:
The best related configuration are also posted