Need to split the logstash output to hadoop, per minute.
My output looks like that:
webhdfs {
host => "127.0.0.1" # (required)
port => 50070 # (optional, default: 50070)
path => "/user/logstash/dt=%{+YYYY-MM-dd}/logstash-%{+HH:mm}.log" # (required)
user => "vphadoop" # (required)
codec => "json"
}
I am getting the following error:
[2017-06-08T13:22:48,149][ERROR][logstash.outputs.webhdfs ] Max write retries reached. Events will be discarded. Exception: {"RemoteException":{"exception":"InvalidPathException","javaClassName":"org.apache.hadoop.fs.**InvalidPathException**","message":"Invalid path name Invalid file name: /user/logstash/dt=2017-03-01/logstash-13:04.log"}}
When looking on the hadoop hdfs I see:
root@vpelastic:[/var/log]$ hdfs dfs -ls /user
Found 2 items
drwxr-xr-x - vpwrk1 supergroup 0 2017-06-05 11:16 /user/db
drwxrwxrwx - vphadoop supergroup 0 2017-06-01 15:05 /user/logstash
I expect the logstash to create the rest of the directories.
Thanks
Sharon.