hbase-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jacques <whs...@gmail.com>
Subject Are the read order of columns deterministic within a single family?
Date Mon, 13 Feb 2012 16:42:33 GMT
When doing a get or scan, are the read order of columns within a particular
family always going to be the same?

For example, if I have two columns within a family, ColumnNameA and
ColumnNameB, will ColumnNameA always come first in read order?  If so, is
this byte ordered on the column name qualifier the same as row ids?

I'm working on a scenario where I often want to read a small subset of
values from a particular row.  While it seems like this might be a good use
for a separate column family, I was thinking that
cause me problems because this information is magnitudes smaller than
the rest of the family.  I was thinking that if I made sure these values
were first in disk order, I could pull this data without making HBase
iterate over 1000's of other columns.

Possible?  Thoughts?


  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message