Chunked array repro

[pzuev]

Changelog category

• Not for changelog (changelog entry is not required)

Description for reviewers

Use with kqprun: ./kqprun -c dq.conf -s s1.sql -p query_dq_hash_combine_q10.sql -C query --log KQP_YQL=trace -T script --result-file result_dq_hash_combine_q10.json

Query SQL (tpch q10 + DqHashCombine):

pragma ydb.UseDqHashCombine = "true";
pragma kikimr.UseLlvm = "false";
-- TPC-H/TPC-R Returned Item Reporting Query (Q10)
-- TPC TPC-H Parameter Substitution (Version 2.17.2 build 0)
-- using 1680793381 as a seed to the RNG

$border = Date("1993-12-01");
$join1 = (
select
    c.c_custkey as c_custkey,
    c.c_name as c_name,
    c.c_acctbal as c_acctbal,
    c.c_address as c_address,
    c.c_phone as c_phone,
    c.c_comment as c_comment,
    c.c_nationkey as c_nationkey,
    o.o_orderkey as o_orderkey
from
    `customer` as c
join
    `orders` as o
on
    c.c_custkey = o.o_custkey
where
    o.o_orderdate >= $border
    and o.o_orderdate < ($border + Interval("P90D"))
);
$join2 = (
select
    j.c_custkey as c_custkey,
    j.c_name as c_name,
    j.c_acctbal as c_acctbal,
    j.c_address as c_address,
    j.c_phone as c_phone,
    j.c_comment as c_comment,
    j.c_nationkey as c_nationkey,
    l.l_extendedprice as l_extendedprice,
    l.l_discount as l_discount
from
    $join1 as j
join
    `lineitem` as l
on
    l.l_orderkey = j.o_orderkey
where
    l.l_returnflag = 'R'
);
$join3 = (
select
    j.c_custkey as c_custkey,
    j.c_name as c_name,
    j.c_acctbal as c_acctbal,
    j.c_address as c_address,
    j.c_phone as c_phone,
    j.c_comment as c_comment,
    j.c_nationkey as c_nationkey,
    j.l_extendedprice as l_extendedprice,
    j.l_discount as l_discount,
    n.n_name as n_name
from
    $join2 as j
join
    `nation` as n
on
    n.n_nationkey = j.c_nationkey
);
select
    c_custkey,
    c_name,
    sum(l_extendedprice * (1 - l_discount)) as revenue,
    c_acctbal,
    n_name,
    c_address,
    c_phone,
    c_comment
from
    $join3
group by
    c_custkey,
    c_name,
    c_acctbal,
    c_phone,
    n_name,
    c_address,
    c_comment
order by
    revenue desc
limit 20;

Schema SQL:

CREATE EXTERNAL DATA SOURCE tpc WITH (
    SOURCE_TYPE="ObjectStorage",
    LOCATION="https://storage.yandexcloud.net/tpc/",
    AUTH_METHOD="NONE"
);

CREATE EXTERNAL TABLE customer (
    c_acctbal Double NOT NULL,
    c_mktsegment String NOT NULL,
    c_phone String NOT NULL,
    c_nationkey Int32 NOT NULL,
    c_custkey Int32 NOT NULL,
    c_name String NOT NULL,
    c_comment String NOT NULL,
    c_address String NOT NULL
) WITH (
    DATA_SOURCE="tpc",
    LOCATION="/h/s1/parquet/customer/",
    FORMAT="parquet"
);

CREATE EXTERNAL TABLE lineitem (
    l_orderkey Int64 NOT NULL,
    l_partkey Int32 NOT NULL,
    l_suppkey Int32 NOT NULL,
    l_linenumber Int32 NOT NULL,
    l_quantity Double NOT NULL,
    l_extendedprice Double NOT NULL,
    l_discount Double NOT NULL,
    l_tax Double NOT NULL,
    l_returnflag String NOT NULL,
    l_linestatus String NOT NULL,
    l_shipdate Date NOT NULL,
    l_commitdate Date NOT NULL,
    l_receiptdate Date NOT NULL,
    l_shipinstruct String NOT NULL,
    l_shipmode String NOT NULL,
    l_comment String NOT NULL
) WITH (
    DATA_SOURCE="tpc",
    LOCATION="/h/s1/parquet/lineitem/",
    FORMAT="parquet"
);

CREATE EXTERNAL TABLE nation (
    n_nationkey Int32 NOT NULL,
    n_name String NOT NULL,
    n_regionkey Int32 NOT NULL,
    n_comment String NOT NULL
) WITH (
    DATA_SOURCE="tpc",
    LOCATION="/h/s1/parquet/nation/",
    FORMAT="parquet"
);

CREATE EXTERNAL TABLE orders (
    o_orderkey Int64 NOT NULL,
    o_custkey Int32 NOT NULL,
    o_orderstatus String NOT NULL,
    o_totalprice Double NOT NULL,
    o_orderdate Date NOT NULL,
    o_orderpriority String NOT NULL,
    o_clerk String NOT NULL,
    o_shippriority Int32 NOT NULL,
    o_comment String NOT NULL
) WITH (
    DATA_SOURCE="tpc",
    LOCATION="/h/s1/parquet/orders/",
    FORMAT="parquet"
);

CREATE EXTERNAL TABLE part (
    p_partkey Int32 NOT NULL,
    p_name String NOT NULL,
    p_mfgr String NOT NULL,
    p_brand String NOT NULL,
    p_type String NOT NULL,
    p_size Int32 NOT NULL,
    p_container String NOT NULL,
    p_retailprice Double NOT NULL,
    p_comment String NOT NULL
) WITH (
    DATA_SOURCE="tpc",
    LOCATION="/h/s1/parquet/part/",
    FORMAT="parquet"
);

CREATE EXTERNAL TABLE partsupp (
    ps_partkey Int32 NOT NULL,
    ps_suppkey Int32 NOT NULL,
    ps_availqty Int32 NOT NULL,
    ps_supplycost Double NOT NULL,
    ps_comment String NOT NULL
) WITH (
    DATA_SOURCE="tpc",
    LOCATION="/h/s1/parquet/partsupp/",
    FORMAT="parquet"
);

CREATE EXTERNAL TABLE region (
    r_regionkey Int32 NOT NULL,
    r_name String NOT NULL,
    r_comment String NOT NULL
) WITH (
    DATA_SOURCE="tpc",
    LOCATION="/h/s1/parquet/region/",
    FORMAT="parquet"
);

CREATE EXTERNAL TABLE supplier (
    s_suppkey Int32 NOT NULL,
    s_name String NOT NULL,
    s_address String NOT NULL,
    s_nationkey Int32 NOT NULL,
    s_phone String NOT NULL,
    s_acctbal Double NOT NULL,
    s_comment String NOT NULL
) WITH (
    DATA_SOURCE="tpc",
    LOCATION="/h/s1/parquet/supplier/",
    FORMAT="parquet"
);

[github-actions]

⚪ 2025-10-24 15:11:00 UTC Pre-commit check linux-x86_64-release-asan for 0238831 has started.
⚪ 2025-10-24 15:11:15 UTC Artifacts will be uploaded here
⚪ 2025-10-24 15:14:13 UTC ya make is running...
🟡 2025-10-24 17:21:08 UTC Some tests failed, follow the links below. This fail is not in blocking policy yet

Ya make output | Test bloat

TESTS	PASSED	ERRORS	FAILED	SKIPPED	MUTED?
16201	15727	0	253	203	18

🟢 2025-10-24 17:21:13 UTC Build successful.
🟢 2025-10-24 17:21:38 UTC ydbd size 3.8 GiB changed* by +65.2 KiB, which is < 100.0 KiB vs main: OK

ydbd size dash	main: 03c8f25	merge: 0238831	diff	diff %
ydbd size	4 055 667 448 Bytes	4 055 734 264 Bytes	+65.2 KiB	+0.002%
ydbd stripped size	1 506 010 176 Bytes	1 506 037 216 Bytes	+26.4 KiB	+0.002%

^{*please be aware that the difference is based on comparing your commit and the last completed build from the post-commit, check comparation}

[github-actions]

⚪ 2025-10-24 15:11:10 UTC Pre-commit check linux-x86_64-relwithdebinfo for 0238831 has started.
⚪ 2025-10-24 15:11:25 UTC Artifacts will be uploaded here
⚪ 2025-10-24 15:14:54 UTC ya make is running...
🟡 2025-10-24 17:08:53 UTC Some tests failed, follow the links below. Going to retry failed tests...

Ya make output | Test bloat

TESTS	PASSED	ERRORS	FAILED	SKIPPED	MUTED?
39736	36694	0	220	2798	24

⚪ 2025-10-24 17:09:03 UTC ya make is running... (failed tests rerun, try 2)
🟡 2025-10-24 17:41:22 UTC Some tests failed, follow the links below. Going to retry failed tests...

Ya make output | Test bloat | Test bloat

TESTS	PASSED	ERRORS	FAILED	SKIPPED	MUTED?
2352 (only retried tests)	2087	0	203	40	22

⚪ 2025-10-24 17:41:27 UTC ya make is running... (failed tests rerun, try 3)
🔴 2025-10-24 18:19:28 UTC Some tests failed, follow the links below.

Ya make output | Test bloat | Test bloat | Test bloat

TESTS	PASSED	ERRORS	FAILED	SKIPPED	MUTED?
1982 (only retried tests)	1729	0	190	41	22

🟢 2025-10-24 18:19:30 UTC Build successful.
🟢 2025-10-24 18:20:00 UTC ydbd size 2.3 GiB changed* by +9.8 KiB, which is < 100.0 KiB vs main: OK

ydbd size dash	main: 7090690	merge: 0238831	diff	diff %
ydbd size	2 437 327 344 Bytes	2 437 337 400 Bytes	+9.8 KiB	+0.000%
ydbd stripped size	518 128 456 Bytes	518 131 336 Bytes	+2.8 KiB	+0.001%

^{*please be aware that the difference is based on comparing your commit and the last completed build from the post-commit, check comparation}

[github-actions]

🟢 2025-10-24 15:12:08 UTC The validation of the Pull Request description is successful.