Database Reference
In-Depth Information
cessing to avoid garbage collection activities. So you can run out of memory off-heap
or on-heap. For more discussion about off-heap and on-heap, please refer the “CPU
and Memory Utilization” and “Off-Heap vs. On-Heap” sections in
Chapter 8
. Addi-
tionally, you can check client-side code in case the application is reading something
like entire rows or large columns.
Too Much Garbage Collection Activity
If garbage collection (GC) takes longer than a few seconds, check the
system.log
file for a message like one of the following:
INFO [ScheduledTasks:1] 2014-01-29 02:41:16,579
GCInspector.java (line 116) GC for ConcurrentMarkSweep:
341 ms for 1 collections, 8001582816 used; max is
8126464000
INFO [ScheduledTasks:1] 2014-01-29 02:41:29,135
GCInspector.java (line 116) GC for ConcurrentMarkSweep:
350 ms for 1 collections, 8027555576 used; max is
8126464000
Long garbage collection pauses can create a stop-the-world scenario that brings
data nodes to a freeze state for some time. In such scenarios, it is highly possible that a
portion of the JVM that is not in use would be swapped out. It is recommended to keep
this off using
swapoff -all
Running this command over a Linux terminal will disable swapping for all devices.
These are a few common problems you are likely to encounter and their solutions.
Next let's discuss the road ahead with Cassandra.
Road Ahead with Cassandra
Apache Cassandra is still under active development, and an extremely active commu-
nity keeps the momentum going. We have already discussed the multiple tools and
