10-04-2015 01:53 AM - edited 10-04-2015 03:09 AM
In my cluster , I am facing issue with cached memory in hadoop client server .
I have hadoop clients like oozie, hue, hive, impala ... running on this machine ( Flume and HBASE is not configured in this server ) . Also my CM services are running on the same machine.
If I look at my Memory my cache is utlized highly and this cache is not released even if i stop all my hadoop process and CM process.
Why my cache is not flused automatically . How to identify is there any memory leak ?
Memory in (GB )-
Total : 94
used : 77
free : 16
shared : 0
buffres : 0
cached : 59
-/+ buffers/cache : 17,76
Swap: 7 , 0, 7
Is this expected ?
The reason I posted this query is - my application is hanging after some stage and hope this is due to the memory unavailiablity.
Rathish A M