Thursday, June 23, 2011

Query Multiple rows with range in cassandra using hector api 0.7


The RangeSlicesQuery method functions similarly to MultigetSliceQuery. The difference is that RangeSlicesQuery uses a contiguous range of keys as opposed to the specific set of keys used by MultigetSliceQuery. If the OrderPreservingPartitioner (OPP) were configured, the same query from the MultigetSliceQuery example above could be done with the RangeSlicesQuery:
RangeSlicesQuery<String, String, String> rangeSlicesQuery =
HFactory.createRangeSlicesQuery(keyspace, stringSerializer, stringSerializer, stringSerializer);
rangeSlicesQuery.setKeys("fake_key_0", "fake_key_4");
rangeSlicesQuery.setRange("", "", false, 3);
Result<OrderedRows<String, String, String>> result = rangeSlicesQuery.execute();
If you did not want to explicitly set the column names, or as above, get the first three columns in order, but instead wanted to return all columns starting with a common prefix, you could change the RangeSlicesQuery from the example above as follows:
rangeSlicesQuery.setRange("fake_column_", "", false, 3);
If you were using OPP the endKey could be provided as well to further limit the results.
It is important to note that unless you are using OPP in your storage configuration, the results returned from a RangeSlicesQuery will not be in any key order. Therefore using a RangeSlicesQuery with partitioners other than OPP is really only useful when you want to iterate over the entire keyspace. See the Cassandra API documentation for further discussion on partitioner selection and the resulting effects on queries.

No comments:

Post a Comment