Reading error in cassandra -

i'm having weir error trying read data cassandra table. have single-node installation, default setup. query i'm making:

  select component_id,          reading_1,          reading_2,          reading_3,          date   component_readings   park_id=2         , component_id in (479)         , date >= '2016-04-09+0000'         , date <= '2016-05-08+0000'; 

component_readings simple table, no clustering conditions:

create table component_readings (     park_id int,     component_id int,     date timestamp,     reading_1 decimal,     reading_2 decimal,     ...     primary key ((park_id), component_id, date) ); 

with component_id values, works, , values, fails. error i'm getting:

cassandra.readfailure: code=1300 [replica(s) failed execute read]  message="operation failed - received 0 responses , 1 failures" info={'required_responses': 1, 'received_responses': 0, 'failures': 1, 'consistency': 'local_one'} 

and cassandra's system.log shows error:

error [sharedpool-worker-1] 2016-05-09 15:33:58,872 storageproxy.java:1818 -  scanned on 100001 tombstones during query 'select * xrem.component_readings park_id, component_id = 2, 479 , date >= 2016-04-09 02:00+0200 , date <= 2016-05-08 02:00+0200 limit 5000' (last scanned row partion key ((2, 479), 2016-05-04 17:30+0200)); query aborted 

the weird thing error when making query external program (via python cassandra-connector). if make directly in cqlsh shell, works perfectly.

my installation cassandra 2.2, i've upgraded 3.5, , same error.

you exceeding tombstone_failure_threshold. defaults 100'000. can either

• increase value in cassandra.yaml or
• clean tombstones

to latter alter table , set gc_grace_seconds 0:

alter table component_readings gc_grace_seconds = 0; 

then trigger compaction via nodetool. flush out tombstones.

in particular scenario of one-node-cluster leave gc_grace_seconds @ zero. if do, keep in mind undo if ever want use more 1 node!