From 38ec00acc7b721acf0eea8be42f7d19a4314aba5 Mon Sep 17 00:00:00 2001
From: robot-clickhouse <robot-clickhouse@users.noreply.github.com>
Date: Tue, 10 Sep 2024 14:09:45 +0000
Subject: [PATCH] Backport #68584 to 24.3: Avoid detached covered-by-broken
 part duplicates

---
 src/Storages/MergeTree/IMergeTreeDataPart.cpp |   1 +
 .../test_covered_by_broken_exists/__init__.py |   0
 .../test_covered_by_broken_exists/test.py     | 103 ++++++++++++++++++
 3 files changed, 104 insertions(+)
 create mode 100644 tests/integration/test_covered_by_broken_exists/__init__.py
 create mode 100644 tests/integration/test_covered_by_broken_exists/test.py

diff --git a/src/Storages/MergeTree/IMergeTreeDataPart.cpp b/src/Storages/MergeTree/IMergeTreeDataPart.cpp
index a26e2b725be..933b556934a 100644
--- a/src/Storages/MergeTree/IMergeTreeDataPart.cpp
+++ b/src/Storages/MergeTree/IMergeTreeDataPart.cpp
@@ -1985,6 +1985,7 @@ DataPartStoragePtr IMergeTreeDataPart::makeCloneInDetached(const String & prefix
     IDataPartStorage::ClonePartParams params
     {
         .copy_instead_of_hardlink = isStoredOnRemoteDiskWithZeroCopySupport() && storage.supportsReplication() && storage_settings->allow_remote_fs_zero_copy_replication,
+        .keep_metadata_version = prefix == "covered-by-broken",
         .make_source_readonly = true,
         .external_transaction = disk_transaction
     };
diff --git a/tests/integration/test_covered_by_broken_exists/__init__.py b/tests/integration/test_covered_by_broken_exists/__init__.py
new file mode 100644
index 00000000000..e69de29bb2d
diff --git a/tests/integration/test_covered_by_broken_exists/test.py b/tests/integration/test_covered_by_broken_exists/test.py
new file mode 100644
index 00000000000..caa091fdd2d
--- /dev/null
+++ b/tests/integration/test_covered_by_broken_exists/test.py
@@ -0,0 +1,103 @@
+import pytest
+import logging
+import time
+from helpers.cluster import ClickHouseCluster
+from helpers.test_tools import TSV
+from helpers.test_tools import assert_eq_with_retry
+
+cluster = ClickHouseCluster(__file__)
+node1 = cluster.add_instance("node1", stay_alive=True, with_zookeeper=True)
+node2 = cluster.add_instance("node2", with_zookeeper=True)
+
+instance = node1
+q = node1.query
+
+path_to_data = "/var/lib/clickhouse/"
+
+
+@pytest.fixture(scope="module")
+def started_cluster():
+    try:
+        cluster.start()
+        yield cluster
+    finally:
+        cluster.shutdown()
+
+
+def wait_merged_part(table, part_name, retries=100):
+    q("OPTIMIZE TABLE {} FINAL".format(table))
+    for i in range(retries):
+        result = q(
+            "SELECT name FROM system.parts where table='{}' AND name='{}'".format(
+                table, part_name
+            )
+        )
+        if result:
+            return True
+        time.sleep(0.5)
+    else:
+        return False
+
+
+def test_make_clone_covered_by_broken_detached_dir_exists(started_cluster):
+    q("DROP TABLE IF EXISTS test_make_clone_cvbdde SYNC")
+
+    q(
+        "CREATE TABLE test_make_clone_cvbdde(n int, m String) ENGINE=ReplicatedMergeTree('/test_make_clone_cvbdde', '1') ORDER BY n SETTINGS old_parts_lifetime=3600, min_age_to_force_merge_seconds=1, min_age_to_force_merge_on_partition_only=0"
+    )
+    path = path_to_data + "data/default/test_make_clone_cvbdde/"
+
+    q("INSERT INTO test_make_clone_cvbdde VALUES (0, 'hbl')")
+
+    q("INSERT INTO test_make_clone_cvbdde VALUES (1, 'hbl')")
+    if not (wait_merged_part("test_make_clone_cvbdde", "all_0_1_1")):
+        assert False, "Part all_0_1_1 doesn't appeared in system.parts"
+
+    q("INSERT INTO test_make_clone_cvbdde VALUES (2, 'hbl')")
+    if not (wait_merged_part("test_make_clone_cvbdde", "all_0_2_2")):
+        assert False, "Part all_0_2_2 doesn't appeared in system.parts"
+
+    q("INSERT INTO test_make_clone_cvbdde VALUES (3, 'hbl')")
+    if not (wait_merged_part("test_make_clone_cvbdde", "all_0_3_3")):
+        assert False, "Part all_0_3_3 doesn't appeared in system.parts"
+
+    res = str(instance.exec_in_container(["ls", path]).strip().split("\n"))
+
+    # broke the merged parts
+    instance.exec_in_container(
+        [
+            "bash",
+            "-c",
+            "echo 'broken' > {}".format(path + "all_0_1_1/data.bin"),
+        ]
+    )
+
+    instance.exec_in_container(
+        [
+            "bash",
+            "-c",
+            "echo 'broken' > {}".format(path + "all_0_2_2/data.bin"),
+        ]
+    )
+
+    instance.exec_in_container(
+        [
+            "bash",
+            "-c",
+            "echo 'broken' > {}".format(path + "all_0_3_3/data.bin"),
+        ]
+    )
+
+    instance.restart_clickhouse(kill=True)
+
+    assert [
+        "broken-on-start_all_0_1_1",
+        "broken-on-start_all_0_2_2",
+        "broken-on-start_all_0_3_3",
+        "covered-by-broken_all_0_0_0",
+        "covered-by-broken_all_1_1_0",
+        "covered-by-broken_all_2_2_0",
+        "covered-by-broken_all_3_3_0",
+    ] == sorted(
+        instance.exec_in_container(["ls", path + "detached/"]).strip().split("\n")
+    )