Redis search with index giving inconsistent results - redis

Question

We are not able to get consistent searches to redis.
Redis server v=7.0.7 sha=00000000:0 malloc=jemalloc-5.2.1 bits=64 build=2260280010e18db8
root#redis01:~# redis-cli
127.0.0.1:6379> info modules
module:name=ReJSON,ver=20008,api=1,filters=0,usedby=[search],using=[],options=[handle-io-errors]
module:name=search,ver=20405,api=1,filters=0,usedby=[],using=[ReJSON],options=[]
We have exactly 5022786 key/value pairs, the same number of entries are on our 'idx36' index. There is no incoming traffic, so this dataset of 5022786 entries remains constant all the time.
127.0.0.1:6379> info keyspace
db0:keys=5022786,expires=5022786,avg_ttl=257039122
127.0.0.1:6379> ft.info idx36
1) index_name
2) idx36
3) index_options
4) 1) "NOFREQS"
5) index_definition
6) 1) key_type
2) HASH
3) prefixes
4) 1) 36|
5) default_score
6) "1"
7) attributes
8) 1) 1) identifier
2) CheckIn
3) attribute
4) CheckIn
5) type
6) NUMERIC
2) 1) identifier
2) HotelCode
3) attribute
4) HotelCode
5) type
6) TAG
7) SEPARATOR
8)
9) num_docs
10) "5022786"
11) max_doc_id
12) "12729866"
13) num_terms
14) "0"
15) num_records
16) "1.8446744073526942e+19"
17) inverted_sz_mb
18) "162.96730041503906"
19) vector_index_sz_mb
20) "0"
21) total_inverted_index_blocks
22) "17965200"
23) offset_vectors_sz_mb
24) "0"
25) doc_table_size_mb
26) "1843.12548828125"
27) sortable_values_size_mb
28) "0"
29) key_table_size_mb
30) "194.14376831054688"
31) records_per_doc_avg
32) "3672612012032"
33) bytes_per_record_avg
34) "9.2636185181071973e-12"
35) offsets_per_term_avg
36) "0"
37) offset_bits_per_record_avg
38) "-nan"
39) hash_indexing_failures
40) "333006"
41) indexing
42) "0"
43) percent_indexed
44) "1"
45) gc_stats
46) 1) bytes_collected
2) "264321009"
3) total_ms_run
4) "1576438"
5) total_cycles
6) "5"
7) average_cycle_time_ms
8) "315287.59999999998"
9) last_run_time_ms
10) "706929"
11) gc_numeric_trees_missed
12) "0"
13) gc_blocks_denied
14) "23251"
47) cursor_stats
48) 1) global_idle
2) (integer) 0
3) global_total
4) (integer) 0
5) index_capacity
6) (integer) 128
7) index_total
8) (integer) 0
This index has a tag called 'HotelCode' and a numeric field called 'CheckIn'.
Now we try to search all entries that contain the string 'AO-B0' within 'HotelCode' (should be all):
127.0.0.1:6379> ft.search idx36 '(#HotelCode:{AO\-B0})' limit 0 0
1) (integer) 2708499
And now try to search all entries that don't contain the string 'AO-B0' within 'HotelCode' (should be 0):
127.0.0.1:6379> ft.search idx36 '-(#HotelCode:{AO\-B0})' limit 0 0
1) (integer) 0
But these two searches don't sum the total number of entries. Even if I'm wrong and not all entries contain the 'AO-B0' string, If I repeat the first search the result changes every time:
127.0.0.1:6379> ft.search idx36 '(#HotelCode:{AO\-B0})' limit 0 0
1) (integer) 2615799
(0.50s)
127.0.0.1:6379> ft.search idx36 '(#HotelCode:{AO\-B0})' limit 0 0
1) (integer) 2442799
(0.50s)
127.0.0.1:6379> ft.search idx36 '(#HotelCode:{AO\-B0})' limit 0 0
1) (integer) 2626299
(0.50s)
127.0.0.1:6379> ft.search idx36 '(#HotelCode:{AO\-B0})' limit 0 0
1) (integer) 2694899
(0.50s)
127.0.0.1:6379> ft.search idx36 '(#HotelCode:{AO\-B0})' limit 0 0
1) (integer) 2516699
(0.50s)
If now I try this search, less restrictive, I should get more entries ... but not:
127.0.0.1:6379> ft.search idx36 '#HotelCode:{AO\-B*}' limit 0 0
1) (integer) 1806899
Maybe I'm doing something wrong ... if someone can point me to the right way ...