Files
impala/testdata/datasets/functional/schema_constraints.csv
Sailesh Mukil 76b674850f IMPALA-2466: Add more tests for the HDFS parquet scanner.
These tests functionally test whether the following type of files
are able to be scanned properly:

1) Add a parquet file with multiple blocks such that each node has to
   scan multiple blocks.
2) Add a parquet file with multiple blocks but only one row group
   that spans the entire file. Only one scan range should do any work
   in this case.

Change-Id: I4faccd9ce3fad42402652c8f17d4e7aa3d593368
Reviewed-on: http://gerrit.cloudera.org:8080/1500
Reviewed-by: Sailesh Mukil <sailesh@cloudera.com>
Tested-by: Internal Jenkins
2016-03-25 13:10:15 +00:00

11 KiB

1# Table level constraints:
2# Allows for defining constraints on which file formats to generate for an individual
3# table. The table name should match the base table name defined in the schema template
4# file.
5table_name:stringids, constraint:restrict_to, table_format:hbase/none/none
6table_name:hbasecolumnfamilies, constraint:restrict_to, table_format:hbase/none/none
7table_name:insertalltypesagg, constraint:restrict_to, table_format:hbase/none/none
8table_name:alltypessmallbinary, constraint:restrict_to, table_format:hbase/none/none
9table_name:insertalltypesaggbinary, constraint:restrict_to, table_format:hbase/none/none
10table_name:hbasealltypeserror, constraint:restrict_to, table_format:hbase/none/none
11table_name:hbasealltypeserrornonulls, constraint:restrict_to, table_format:hbase/none/none
12table_name:alltypesinsert, constraint:restrict_to, table_format:text/none/none
13table_name:alltypesnopart_insert, constraint:restrict_to, table_format:text/none/none
14table_name:insert_overwrite_nopart, constraint:restrict_to, table_format:text/none/none
15table_name:insert_overwrite_partitioned, constraint:restrict_to, table_format:text/none/none
16table_name:insert_string_partitioned, constraint:restrict_to, table_format:text/none/none
17table_name:alltypesinsert, constraint:restrict_to, table_format:parquet/none/none
18table_name:alltypesnopart_insert, constraint:restrict_to, table_format:parquet/none/none
19table_name:alltypesinsert, constraint:restrict_to, table_format:text/none/none
20table_name:alltypesnopart_insert, constraint:restrict_to, table_format:text/none/none
21table_name:insert_overwrite_nopart, constraint:restrict_to, table_format:text/none/none
22table_name:insert_overwrite_partitioned, constraint:restrict_to, table_format:text/none/none
23table_name:insert_string_partitioned, constraint:restrict_to, table_format:text/none/none
24table_name:alltypesinsert, constraint:restrict_to, table_format:parquet/none/none
25table_name:alltypesnopart_insert, constraint:restrict_to, table_format:parquet/none/none
26table_name:insert_overwrite_nopart, constraint:restrict_to, table_format:parquet/none/none
27table_name:insert_overwrite_partitioned, constraint:restrict_to, table_format:parquet/none/none
28table_name:insert_string_partitioned, constraint:restrict_to, table_format:parquet/none/none
29table_name:old_rcfile_table, constraint:restrict_to, table_format:rc/none/none
30table_name:bad_text_lzo, constraint:restrict_to, table_format:text/lzo/block
31table_name:bad_text_gzip, constraint:restrict_to, table_format:text/gzip/block
32table_name:bad_seq_snap, constraint:restrict_to, table_format:seq/snap/block
33table_name:bad_parquet, constraint:restrict_to, table_format:parquet/none/none
34table_name:bad_magic_number, constraint:restrict_to, table_format:parquet/none/none
35table_name:bad_metadata_len, constraint:restrict_to, table_format:parquet/none/none
36table_name:bad_dict_page_offset, constraint:restrict_to, table_format:parquet/none/none
37table_name:bad_compressed_size, constraint:restrict_to, table_format:parquet/none/none
38table_name:alltypesagg_hive_13_1, constraint:restrict_to, table_format:parquet/none/none
39table_name:kite_required_fields, constraint:restrict_to, table_format:parquet/none/none
40table_name:bad_column_metadata, constraint:restrict_to, table_format:parquet/none/none
41table_name:lineitem_multiblock, constraint:restrict_to, table_format:parquet/none/none
42table_name:lineitem_sixblocks, constraint:restrict_to, table_format:parquet/none/none
43table_name:lineitem_multiblock_one_row_group, constraint:restrict_to, table_format:parquet/none/none
44# TODO: Support Avro. Data loading currently fails for Avro because complex types
45# cannot be converted to the corresponding Avro types yet.
46table_name:allcomplextypes, constraint:restrict_to, table_format:text/none/none
47table_name:allcomplextypes, constraint:restrict_to, table_format:parquet/none/none
48table_name:allcomplextypes, constraint:restrict_to, table_format:hbase/none/none
49table_name:functional, constraint:restrict_to, table_format:text/none/none
50table_name:complextypes_fileformat, constraint:restrict_to, table_format:text/none/none
51table_name:complextypes_fileformat, constraint:restrict_to, table_format:parquet/none/none
52table_name:complextypes_fileformat, constraint:restrict_to, table_format:avro/snap/block
53table_name:complextypes_fileformat, constraint:restrict_to, table_format:rc/snap/block
54table_name:complextypes_fileformat, constraint:restrict_to, table_format:seq/snap/block
55table_name:complextypes_multifileformat, constraint:restrict_to, table_format:text/none/none
56# TODO: Avro
57table_name:complextypestbl, constraint:restrict_to, table_format:parquet/none/none
58table_name:alltypeserror, constraint:exclude, table_format:parquet/none/none
59table_name:alltypeserrornonulls, constraint:exclude, table_format:parquet/none/none
60table_name:unsupported_types, constraint:exclude, table_format:parquet/none/none
61table_name:escapechartesttable, constraint:exclude, table_format:parquet/none/none
62table_name:TblWithRaggedColumns, constraint:exclude, table_format:parquet/none/none
63# the text_ tables are for testing test delimiters and escape chars in text files
64table_name:text_comma_backslash_newline, constraint:restrict_to, table_format:text/none/none
65table_name:text_dollar_hash_pipe, constraint:restrict_to, table_format:text/none/none
66table_name:text_thorn_ecirc_newline, constraint:restrict_to, table_format:text/none/none
67table_name:bad_serde, constraint:restrict_to, table_format:text/none/none
68table_name:rcfile_lazy_binary_serde, constraint:restrict_to, table_format:rc/none/none
69table_name:unsupported_partition_types, constraint:restrict_to, table_format:text/none/none
70table_name:nullformat_custom, constraint:exclude, table_format:parquet/none/none
71table_name:alltypes_view, constraint:restrict_to, table_format:text/none/none
72table_name:allcomplextypes_view, constraint:restrict_to, table_format:text/none/none
73table_name:alltypes_view, constraint:restrict_to, table_format:seq/snap/block
74table_name:alltypes_hive_view, constraint:restrict_to, table_format:text/none/none
75table_name:alltypes_view_sub, constraint:restrict_to, table_format:text/none/none
76table_name:alltypes_view_sub, constraint:restrict_to, table_format:seq/snap/block
77table_name:alltypes_parens, constraint:restrict_to, table_format:text/none/none
78table_name:complex_view, constraint:restrict_to, table_format:text/none/none
79table_name:complex_view, constraint:restrict_to, table_format:seq/snap/block
80table_name:view_view, constraint:restrict_to, table_format:text/none/none
81table_name:view_view, constraint:restrict_to, table_format:seq/snap/block
82# liketbl and tblwithraggedcolumns all have
83# NULLs in primary key columns. hbase does not support
84# writing NULLs to primary key columns.
85table_name:liketbl, constraint:exclude, table_format:hbase/none/none
86table_name:tblwithraggedcolumns, constraint:exclude, table_format:hbase/none/none
87# Tables with only one column are not supported in hbase.
88table_name:greptiny, constraint:exclude, table_format:hbase/none/none
89table_name:tinyinttable, constraint:exclude, table_format:hbase/none/none
90# overflow has a bigint that's too big. hbase may lose precision, hence this
91# table cannot be loaded.
92table_name:overflow, constraint:exclude, table_format:hbase/none/none
93# widerow has a single column with a single row containing a 10MB string. hbase doesn't
94# seem to like this.
95table_name:widerow, constraint:exclude, table_format:hbase/none/none
96# nullformat_custom is used in null-insert tests, which user insert overwrite,
97# which is not supported in hbase. The schema is also specified in HIVE_CREATE
98# with no corresponding LOAD statement.
99table_name:nullformat_custom, constraint:exclude, table_format:hbase/none/none
100table_name:unsupported_types, constraint:exclude, table_format:hbase/none/none
101# On CDH4, decimal can only be tested on formats Impala can write to (text and parquet)
102# TODO: add Avro once Hive or Impala can write Avro decimals
103table_name:decimal_tbl, constraint:restrict_to, table_format:text/none/none
104table_name:decimal_tiny, constraint:restrict_to, table_format:text/none/none
105table_name:decimal_tbl, constraint:restrict_to, table_format:parquet/none/none
106table_name:decimal_tiny, constraint:restrict_to, table_format:parquet/none/none
107table_name:avro_decimal_tbl, constraint:restrict_to, table_format:avro/snap/block
108# TODO first set of tests are for text/none/none
109table_name:chars_tiny, constraint:restrict_to, table_format:text/none/none
110# invalid_decimal_part_tbl[1,2,3] tables are used for testing invalid decimal
111# partition key values (see IMPALA-1040)
112table_name:invalid_decimal_part_tbl1, constraint:restrict_to, table_format:text/none/none
113table_name:invalid_decimal_part_tbl2, constraint:restrict_to, table_format:text/none/none
114table_name:invalid_decimal_part_tbl3, constraint:restrict_to, table_format:text/none/none
115table_name:avro_decimal_tbl, constraint:restrict_to, table_format:avro/snap/block
116# testescape tables are used for testing text scanner delimiter handling
117table_name:table_no_newline, constraint:restrict_to, table_format:text/none/none
118table_name:table_no_newline_part, constraint:restrict_to, table_format:text/none/none
119table_name:testescape_16_lf, constraint:restrict_to, table_format:text/none/none
120table_name:testescape_16_crlf, constraint:restrict_to, table_format:text/none/none
121table_name:testescape_17_lf, constraint:restrict_to, table_format:text/none/none
122table_name:testescape_17_crlf, constraint:restrict_to, table_format:text/none/none
123table_name:testescape_32_lf, constraint:restrict_to, table_format:text/none/none
124table_name:testescape_32_crlf, constraint:restrict_to, table_format:text/none/none
125# alltimezones is used to verify that impala properly deals with timezones
126table_name:alltimezones, constraint:restrict_to, table_format:text/none/none
127# Avro schema is inferred from the column definitions (IMPALA-1136)
128table_name:no_avro_schema, constraint:restrict_to, table_format:avro/snap/block
129table_name:avro_unicode_nulls, constraint:restrict_to, table_format:avro/snap/block
130# test single and multi stream bz2 files
131table_name:bzip2_tbl, constraint:restrict_to, table_format:text/bzip/block
132table_name:large_bzip2_tbl, constraint:restrict_to, table_format:text/bzip/block
133table_name:multistream_bzip2_tbl, constraint:restrict_to, table_format:text/bzip/block
134table_name:large_multistream_bzip2_tbl, constraint:restrict_to, table_format:text/bzip/block
135# Kudu can't handle certain types such as timestamp so we pick and choose the tables
136# we actually use for Kudu related tests.
137table_name:testtbl, constraint:only, table_format:kudu/none/none
138table_name:dimtbl, constraint:only, table_format:kudu/none/none
139table_name:text_comma_backslash_newline, constraint:only, table_format:kudu/none/none
140table_name:tinytable, constraint:only, table_format:kudu/none/none
141table_name:tinyinttable, constraint:only, table_format:kudu/none/none
142table_name:zipcode_incomes, constraint:only, table_format:kudu/none/none