Does it reproducible only on large data sets? If so, then you possible hit PHOENIX-3112 bug. As a workaround, you may try to increase hbase.client.scanner.max.result.size <https://issues.apache.org/jira/browse/PHOENIX-3118> property to something big (100Mb for example).
Thanks, Sergey On Mon, Oct 30, 2017 at 12:56 PM, <[email protected]> wrote: > Hello all, > > > > I am having a confusing problem when querying a Phoenix table via a JDBC > connection (in R as well as client like Squirrel). The table has five > columns. When I query the table like so: > > > > *select * from table1 where column1 = “A” and column2 = “B” and column3 = > “C”* > > > > the query returns all five columns with “NA” as the value in the fifth > column – a column I KNOW to be populated. However, if I query: > > > > *select column5 from table1 where column1 = “A” and column2 = “B” and > column3 = “C”* > > > > the query returns the populated column5. > > > > The other odd detail – if I add a limit clause to the SELECT * query, be > it limit 1 or limit 1000 (there are far less than 1000 records total in > this table), column5 will populate. Otherwise, NA. > > > > Any insight or theories on this odd behavior would be greatly > appreciated! > > > > Thank you, > > > > M. > > > > >
