[
https://issues.apache.org/jira/browse/CASSANDRA-920?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12855149#action_12855149
]
Ryan King commented on CASSANDRA-920:
-------------------------------------
I've rewritten this in terms of our cassandra.gem test suite and can't seem to
find a bug in the client. It fails on the last line.
{code}
@twitter.insert(:StatusRelationships, key, {'user_timelines' =>
{...@uuids[0] => 'Item 0', @uuids[1] => 'Item 1', @uuids[2] => 'Item 2'}})
@twitter.remove(:StatusRelationships, key)
items = @twitter.get(:StatusRelationships, key, 'user_timelines')
assert_equal 0, items.size
@twitter.insert(:StatusRelationships, key,
{'user_timelines' => {...@uuids[2] => 'Item 2', @uuids[3] => 'Item 3',
@uuids[4] => 'Item 4'}})
items = @twitter.get(:StatusRelationships, key, 'user_timelines')
assert_equal 3, items.size
items = @twitter.get(:StatusRelationships,key,'user_timelines',:count => 3)
assert_equal 3, items.size
{code}
And the server logs for this whole test:
{code}
DEBUG - batch_mutate
DEBUG - insert writing local key test_get_super_sub_keys_with_count_after_remove
DEBUG - remove
DEBUG - insert writing local key test_get_super_sub_keys_with_count_after_remove
DEBUG - multiget_slice
DEBUG - weakreadlocal reading SliceFromReadCommand(table='Twitter',
key='test_get_super_sub_keys_with_count_after_remove',
column_parent='QueryPath(columnFamilyName='StatusRelationships',
superColumnName='[...@746a63d3', columnName='null')', start='', finish='',
reversed=false, count=100)
DEBUG - batch_mutate
DEBUG - insert writing local key test_get_super_sub_keys_with_count_after_remove
DEBUG - multiget_slice
DEBUG - weakreadlocal reading SliceFromReadCommand(table='Twitter',
key='test_get_super_sub_keys_with_count_after_remove',
column_parent='QueryPath(columnFamilyName='StatusRelationships',
superColumnName='[...@3bd840d9', columnName='null')', start='', finish='',
reversed=false, count=100)
DEBUG - collecting
93814000-b668-11b2-8e6a-06dc05f11207:false:6...@1270760311002397
DEBUG - collecting
13814000-4eff-11b3-88b0-351600af16aa:false:6...@1270760311002397
DEBUG - collecting
13814000-802c-11b4-8782-bede78b81183:false:6...@1270760311069304
DEBUG - collecting
13814000-e286-11b6-9ea9-48a9edd974b6:false:6...@1270760311069304
DEBUG - collecting
13814000-a73a-11bb-8a14-b8d289de6d53:false:6...@1270760311069304
DEBUG - multiget_slice
DEBUG - weakreadlocal reading SliceFromReadCommand(table='Twitter',
key='test_get_super_sub_keys_with_count_after_remove',
column_parent='QueryPath(columnFamilyName='StatusRelationships',
superColumnName='[...@6e37d490', columnName='null')', start='', finish='',
reversed=false, count=3)
DEBUG - collecting
93814000-b668-11b2-8e6a-06dc05f11207:false:6...@1270760311002397
DEBUG - collecting
13814000-4eff-11b3-88b0-351600af16aa:false:6...@1270760311002397
DEBUG - collecting
13814000-802c-11b4-8782-bede78b81183:false:6...@1270760311069304
DEBUG - GC for ParNew: 7 ms, 20659648 reclaimed leaving 19819056 used; max is
1211826176
{code}
And the thrift structures for the failing multiget_slice
{code}
<CassandraThrift::ColumnParent column_family:"StatusRelationships",
super_column:"user_timelines">
<CassandraThrift::SlicePredicate slice_range:<CassandraThrift::SliceRange
start:"", finish:"", reversed:false, count:3>>
{code}
> Deleting and re-inserting row causes error in get_slice count parameter
> -----------------------------------------------------------------------
>
> Key: CASSANDRA-920
> URL: https://issues.apache.org/jira/browse/CASSANDRA-920
> Project: Cassandra
> Issue Type: Bug
> Affects Versions: 0.5, 0.6
> Environment: Mac OS/ Java 6
> Reporter: Bob Florian
> Assignee: Brandon Williams
> Priority: Minor
> Fix For: 0.6.1
>
> Attachments: 0001_add_system_test.txt
>
>
> I've found that when I delete an entire row in a column family with super
> columns, and then re-insert values with the same row and super column keys,
> the count parameter to the get_slice call no longer works properly. Its like
> it is still counting the deleted columns, but only returning the new columns.
> The following example uses the Ruby Cassandra client (see link below), but
> I've seen the same behavior with the Java Thrift interface.
> Test code:
> --------------
> require 'rubygems'
> require 'cassandra'
> cc = Cassandra.new('Keyspace1')
> cc.insert(:Super1,'test-key1',{'bucket1' => {'1' => 'Item 1', '2' => 'Item
> 2', '5' => 'Item 5'}})
> items = cc.get(:Super1,'test-key1','bucket1')
> puts "returned #{items.size} items, should be 3"
> cc.remove(:Super1,'test-key1')
> items = cc.get(:Super1,'test-key1','bucket1')
> puts "returned #{items.size} items, should be 0"
> cc.insert(:Super1,'test-key1',{'bucket1' => {'3' => 'Item 3', '4' => 'Item
> 4', '6' => 'Item 6'}})
> items = cc.get(:Super1,'test-key1','bucket1')
> puts "returned #{items.size} items, should be 3"
> items = cc.get(:Super1,'test-key1','bucket1',:count => 3)
> puts "returned #{items.size} items, should be 3"
> items = cc.get(:Super1,'test-key1','bucket1',:count => 4)
> puts "returned #{items.size} items, should be 3"
> items = cc.get(:Super1,'test-key1','bucket1',:count => 5)
> puts "returned #{items.size} items, should be 3"
> items = cc.get(:Super1,'test-key1','bucket1',:count => 6)
> puts "returned #{items.size} items, should be 3"
> Output:
> returned 3 items, should be 3
> returned 0 items, should be 0
> returned 3 items, should be 3
> returned 1 items, should be 3
> returned 2 items, should be 3
> returned 2 items, should be 3
> returned 3 items, should be 3
> Ruby library link:
> http://blog.evanweaver.com/files/doc/fauna/cassandra/files/README_rdoc.html
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
