[ https://issues.apache.org/jira/browse/CASSANDRA-20830 ]
Caleb Rackliffe deleted comment on CASSANDRA-20830:
---------------------------------------------
was (Author: maedhroz):
Didn't take long to find our first failing seed. This is from a new test,
{{MultiNodeTableWalkWithMutationTrackingTest}}, that I'll push up shortly in a
new branch...
{noformat}
accord.utils.Property$PropertyError: Property error detected:
Seed = 3447971527803381056
Examples = 10
Pure = true
Error: Unexpected results for query: SELECT * FROM ks1.tbl WHERE token(pk0,
pk1) = token('✒묭쮯ꪠ뱷', '1988-02-17T00:33:34.107Z') PER PARTITION LIMIT 207
Steps: 400
Values:
State:
Setup:
CREATE KEYSPACE IF NOT EXISTS ks1 WITH replication = {'class':
'SimpleStrategy', 'replication_factor': 3} AND replication_type='tracked';
CREATE TABLE ks1.tbl (
pk0 text,
pk1 timestamp,
ck0 smallint,
s0 bigint static,
s1 uuid static,
v0 vector<frozen<list<timestamp>>, 2>,
v1 frozen<map<frozen<list<uuid>>, frozen<tuple<date,
bigint, tinyint>>>>,
v2 frozen<tuple<frozen<map<timeuuid, time>>>>,
v4 frozen<list<vector<smallint, 2>>>,
v3 list<frozen<map<timestamp, uuid>>>,
PRIMARY KEY ((pk0, pk1), ck0)
) WITH CLUSTERING ORDER BY (ck0 ASC)
AND additional_write_policy = '99p'
AND allow_auto_snapshot = true
AND bloom_filter_fp_chance = 0.1
AND caching = {'keys': 'NONE', 'rows_per_partition':
'1396934264'}
AND cdc = false
AND comment = ''
AND compaction = {'class':
'org.apache.cassandra.db.compaction.LeveledCompactionStrategy', 'fanout_size':
'76', 'max_threshold': '32', 'min_threshold': '4',
'provide_overlapping_tombstones': 'NONE', 'single_sstable_uplevel': 'false',
'tombstone_threshold': '0.085422635', 'unchecked_tombstone_compaction': 'false'}
AND compression = {'chunk_length_in_kb': '16', 'class':
'org.apache.cassandra.io.compress.DeflateCompressor'}
AND memtable = 'default'
AND crc_check_chance = 1.0
AND default_time_to_live = 0
AND extensions = {}
AND gc_grace_seconds = 864000
AND incremental_backups = true
AND max_index_interval = 2048
AND memtable_flush_period_in_ms = 0
AND min_index_interval = 128
AND read_repair = 'NONE'
AND speculative_retry = '99p';:
org.apache.cassandra.distributed.test.cql3.MultiNodeTableWalkBase.MultiNodeState
History:
1: DELETE FROM ks1.tbl USING TIMESTAMP 1 WHERE pk0 = '' AND
pk1 = '2015-08-01T09:58:09.924Z' -- on node3
2: INSERT INTO ks1.tbl (pk0, pk1, ck0, s0, s1, v0, v1, v2, v4,
v3) VALUES ('✒묭쮯ꪠ뱷', '1988-02-17T00:33:34.107Z', 22995 - -10195,
3564315193935804201, 00000000-0000-4c00-b500-000000000000,
[['2037-08-28T10:32:11.741Z', '2007-07-30T06:28:57.726Z',
'1971-04-03T20:05:23.920Z'], ['2018-06-23T17:43:48.668Z',
'2034-11-08T04:15:45.848Z', '2066-02-19T18:32:56.931Z']],
{[00000000-0000-4e00-bf00-000000000000, 00000000-0000-4f00-8200-000000000000]:
('-4209731-09-08', -1154931185171508000, 116)},
({00000000-0000-1600-b400-000000000000: '01:35:05.370667439',
00000000-0000-1a00-a800-000000000000: '01:46:23.763460552'}), [[-9555, 30469]],
[{'2033-05-14T17:10:32.208Z': 00000000-0000-4700-b900-000000000000,
'2041-01-18T20:49:16.728Z': 00000000-0000-4600-a100-000000000000,
'2063-01-09T17:24:08.781Z': 00000000-0000-4a00-b300-000000000000},
{'2038-06-27T03:53:13.595Z': 00000000-0000-4400-8a00-000000000000}]) USING
TIMESTAMP 2 -- on node3
3: DELETE FROM ks1.tbl USING TIMESTAMP 3 WHERE pk0 = '' AND
pk1 = '2015-08-01T09:58:09.924Z' -- on node2
4: SELECT * FROM ks1.tbl PER PARTITION LIMIT 341 LIMIT 572 --
full table scan, on node3
5: UPDATE ks1.tbl USING TIMESTAMP 4 SET
v0=[['2006-06-04T21:16:53.597Z'], ['2028-11-05T21:32:09.490Z',
'2014-08-30T10:11:03.456Z', '2002-04-15T09:41:39.941Z']], v4=[[-20852, -124],
[-19879, -18707]], v2=({00000000-0000-1800-b100-000000000000:
'07:17:25.295656788', 00000000-0000-1c00-a700-000000000000:
'16:14:45.747481991', 00000000-0000-1e00-9d00-000000000000:
'08:58:56.011314154'}), s1=00000000-0000-4b00-8500-000000000000 WHERE pk0 =
'梈ꆜ痳' AND pk1 = '2067-08-15T22:19:59.400Z' AND ck0 = 10008 -- on node3
6: DELETE s0, s1 FROM ks1.tbl USING TIMESTAMP 5 WHERE pk0 =
'✒묭쮯ꪠ뱷' AND pk1 = '1988-02-17T00:33:34.107Z' -- on node2
7: SELECT * FROM ks1.tbl WHERE token(pk0, pk1) =
token('✒묭쮯ꪠ뱷', '1988-02-17T00:33:34.107Z') PER PARTITION LIMIT 207 -- by
token, on node2
at accord.utils.Property$StatefulBuilder.check(Property.java:504)
at
org.apache.cassandra.distributed.test.cql3.SingleNodeTableWalkTest.test(SingleNodeTableWalkTest.java:368)
Caused by: java.lang.AssertionError: Unexpected results for query: SELECT *
FROM ks1.tbl WHERE token(pk0, pk1) = token('✒묭쮯ꪠ뱷',
'1988-02-17T00:33:34.107Z') PER PARTITION LIMIT 207
Caused by: java.lang.AssertionError: Unexpected rows found:
pk0 | pk1 | ck0 | s0 | s1
| v0
| v1 | v2
| v3 | v4
'梈ꆜ痳' | '2067-08-15T22:19:59.400Z' | 10008 | null |
00000000-0000-4b00-8500-000000000000 | [['2006-06-04T21:16:53.597Z'],
['2028-11-05T21:32:09.490Z', '2014-08-30T10:11:03.456Z',
'2002-04-15T09:41:39.941Z']] | null | ({00000000-0000-1800-b100-000000000000:
'07:17:25.295656788', 00000000-0000-1c00-a700-000000000000:
'16:14:45.747481991', 00000000-0000-1e00-9d00-000000000000:
'08:58:56.011314154'}) | null | [[-20852, -124], [-19879, -18707]]
Expected:
pk0 | pk1 | ck0 | s0 | s1 | v0
|
v1
| v2
| v3
| v4
'✒묭쮯ꪠ뱷' | '1988-02-17T00:33:34.107Z' | -32346 | null | null |
[['2037-08-28T10:32:11.741Z', '2007-07-30T06:28:57.726Z',
'1971-04-03T20:05:23.920Z'], ['2018-06-23T17:43:48.668Z',
'2034-11-08T04:15:45.848Z', '2066-02-19T18:32:56.931Z']] |
{[00000000-0000-4e00-bf00-000000000000, 00000000-0000-4f00-8200-000000000000]:
('-4209731-09-08', -1154931185171508000, 116)} |
({00000000-0000-1600-b400-000000000000: '01:35:05.370667439',
00000000-0000-1a00-a800-000000000000: '01:46:23.763460552'}) |
[{'2033-05-14T17:10:32.208Z': 00000000-0000-4700-b900-000000000000,
'2041-01-18T20:49:16.728Z': 00000000-0000-4600-a100-000000000000,
'2063-01-09T17:24:08.781Z': 00000000-0000-4a00-b300-000000000000},
{'2038-06-27T03:53:13.595Z': 00000000-0000-4400-8a00-000000000000}] | [[-9555,
30469]]
{noformat}
> AST Fuzz testing support for tables w/ mutation tracking enabled
> ----------------------------------------------------------------
>
> Key: CASSANDRA-20830
> URL: https://issues.apache.org/jira/browse/CASSANDRA-20830
> Project: Apache Cassandra
> Issue Type: Task
> Components: Consistency/Coordination
> Reporter: Caleb Rackliffe
> Assignee: Caleb Rackliffe
> Priority: Normal
> Fix For: NA
>
>
> We need to start exposing the mutation tracking work to the AST fuzz tester.
> Specifically, I think we can start with something similar to
> {{MultiNodeTableWalkWithoutReadRepairTest}}, just with tracking enabled
> rather than RR disabled, etc. The fun thing, from my initial conversations
> with [~bdeggleston], is that historical problems like CASSANDRA-19007 might
> not be possible due to the way filtering queries work with MT.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
