Details
-
Type:
Improvement
-
Status: Patch Available
-
Priority:
Major
-
Resolution: Unresolved
-
Affects Version/s: v0.9.3, v0.9.4
-
Fix Version/s: None
-
Component/s: Sinks+Sources
-
Labels:None
Description
Wrapping the ReaderThread's InputStream with a BufferedInputStream results in a huge performance increase:
Before:
$ ./loggen --verbose -s 113 -r 50000 --interval 30 -i localhost 35999
average rate = 4467.76 msg/sec, count=134041, time=30.018, (last) msg size=113, bandwidth=493.02 kB/sec
After:
$ ./loggen --verbose -s 113 -r 50000 --interval 30 -i localhost 35999
average rate = 50776.43 msg/sec, count=1523293, time=30.000, (last) msg size=113, bandwidth=5603.26 kB/sec
Tested with the standard syslogTcp() source and a null sink. See attached patch.