Fix optimize_skip_unused_shards_rewrite_in for non-UInt64 types

[azat]

Changelog category (leave one):

• Bug Fix

Changelog entry (a user-readable short description of the changes that goes to CHANGELOG.md):
Fix optimize_skip_unused_shards_rewrite_in for non-UInt64 types (may select incorrect shards eventually or throw Cannot infer type of an empty tuple or Function tuple requires at least one argument)

Reported-by: @amosbird

P.S. the fix is a3add4f, other is a tiny refactoring.

[azat]

Integration tests (asan) — fail: 2, passed: 1471, flaky: 16
Integration tests (release) — fail: 3, passed: 1472, flaky: 15

Integration tests (thread) — fail: 61, passed: 824, flaky: 10

Seems that container does not have enough resources (after parallel run of integration tests had been enabled), and clickhouse-server got killed, which eventually leads to container is not running error.

Performance — Errors while building the report.

Broken by #25773

[azat]

Integration tests (thread) — fail: 14, passed: 1469, flaky: 47

Container is not running issue. Not enough resources for such level of parallelism?

E               Exception: Command ['docker', 'exec', 'roottestdistributedqueriesstress_node1_r1_1', 'bash', '-c', "echo 'select * from dist_two_over_dist where key = 0;\n    select * from dist_two_over_dist where key = 1;\n    select * from dist_two_over_dist where key = 2;\n    select * from dist_two_over_dist where key = 3;\n    select * from dist_two_over_dist;' | clickhouse benchmark --concurrency=100 --cumulative --delay=0 --timelimit=3 --hedged_connection_timeout_ms=200 --connect_timeout_with_failover_ms=200 --connections_with_failover_max_tries=5 --prefer_localhost_replica=0 --optimize_skip_unused_shards=1"] return non-zero code 1: Error response from daemon: Container 39c984334f7280985a2fdfae8d7275c6e137f40bc572cda67336e5aa83bc61e9 is not running

2021.07.05 03:11:57.466621 [ 1 ] {} <Fatal> Application: Child process was terminated by signal 9 (KILL). If it is not done by 'forcestop' command or manually, the possible cause is OOM Kille
r (see 'dmesg' and look at the '/var/log/kern.log' for the details)

And AFAICS this test (test_distributed_queries_stress) hadn't passed, and it should be marked as FAIL?

• integration_run_parallel2_1.txt and integration_run_parallel2_2.txt contains only error - runner: error: argument -t/--tests_list: expected at least one argument
• and the row with OK contains failed test.

[azat]

Functional stateless tests (release, DatabaseReplicated) — Some tests restarted, fail: 1, passed: 3076, skipped: 111

2021-07-04 22:51:33 01502_long_log_tinylog_deadlock_race:                                   [ FAIL ] 31.80 sec. - having exception:
2021-07-04 22:51:33 Testing TinyLog
2021-07-04 22:51:33 Done TinyLog
2021-07-04 22:51:33 Testing StripeLog
2021-07-04 22:51:33 Code: 999. DB::Exception: Received from localhost:9000. DB::Exception: Session expired (Session expired). 
2021-07-04 22:51:33 Done StripeLog
2021-07-04 22:51:33 Testing Log
2021-07-04 22:51:33 Done Log

Functional stateless tests (release, wide parts enabled) — fail: 1, passed: 3186, skipped: 1

01781_merge_tree_deduplication - it is not flaky (like report says), but failed.

2021-07-04 22:36:27 01781_merge_tree_deduplication:                                         [ FAIL ] 0.47 sec. - result differs with reference:
2021-07-04 22:36:27 --- /usr/share/clickhouse-test/queries/0_stateless/01781_merge_tree_deduplication.reference	2021-07-04 21:30:02.000000000 +0300
2021-07-04 22:36:27 +++ /tmp/clickhouse-test/0_stateless/01781_merge_tree_deduplication.stdout	2021-07-04 22:36:27.750998547 +0300
2021-07-04 22:36:27 @@ -80,6 +80,7 @@
2021-07-04 22:36:27  1	1
2021-07-04 22:36:27  1	1
2021-07-04 22:36:27  1	1
2021-07-04 22:36:27 +1	1
2021-07-04 22:36:27  2	2
2021-07-04 22:36:27  3	3
2021-07-04 22:36:27  4	4

[azat]

And AFAICS this test (test_distributed_queries_stress) hadn't passed, and it should be marked as FAIL?

Checked in test_results.tsv, and according to it the test marked correctly as FLAKY (I see retries there).

[azat]

Integration tests (thread) — fail: 20, passed: 616, flaky: 7

E               toomanyrequests: You have reached your pull rate limit. You may increase the limit by authenticating and upgrading: https://www.docker.com/increase-rate-limit

[alexey-milovidov]

E               Pulling hdfs1 (sequenceiq/hadoop-docker:2.7.0)...
E               toomanyrequests: You have reached your pull rate limit. You may increase the limit by authenticating and upgrading: https://www.docker.com/increase-rate-limit