It's a description of how many of the compacted SSTables the rows were spread across prior to compaction.  In your case, 15 rows were spread across two of the four sstables, 68757 rows were spread across three of the four sstables, and 6865 were spread across all four.


On Fri, Mar 1, 2013 at 11:07 AM, Jabbar Azam <ajazam@gmail.com> wrote:
Hello,

I'm seeing compaction statistics which look like the following

 INFO 17:03:09,216 Compacted 4 sstables to [/var/lib/cassandra/data/studata/datapoints/studata-datapoints-ib-629,].  420,807,293 bytes to 415,287,150 (~98% of original) in 341,690ms = 1.159088MB/s.  233,761 total rows, 75,637 unique.  Row merge counts were {1:0, 2:15, 3:68757, 4:6865, }

Does anybody know what "Row merge counts were {1:0, 2:15, 3:68757, 4:6865, }" means?

--
Thanks

 A Jabbar Azam



--
Tyler Hobbs
DataStax