From 13e273fdbee9c6c9d2b55bdfcfae6f2ceb63d3f6 Mon Sep 17 00:00:00 2001
From: Vlad Lazar <vlad@redpanda.com>
Date: Thu, 19 Oct 2023 13:51:31 +0100
Subject: [PATCH] admin: expose time point for last completed scrub

When a full scrub of the log finishes, via one or more scrubs, persist
the timestamp at which this happened. This timestamp is then exposed via
the admin API. We will use this in our ducktape tests for end of test
scrubbing.
---
 src/v/cloud_storage/partition_manifest.cc         | 4 ++++
 src/v/cloud_storage/types.cc                      | 3 +++
 src/v/cloud_storage/types.h                       | 6 +++++-
 src/v/redpanda/admin/api-doc/shadow_indexing.json | 4 ++++
 src/v/redpanda/admin_server.cc                    | 4 ++++
 tests/rptest/tests/cloud_storage_scrubber_test.py | 2 ++
 6 files changed, 22 insertions(+), 1 deletion(-)

diff --git a/src/v/cloud_storage/partition_manifest.cc b/src/v/cloud_storage/partition_manifest.cc
index 998eead05e1db..8a6612384ad47 100644
--- a/src/v/cloud_storage/partition_manifest.cc
+++ b/src/v/cloud_storage/partition_manifest.cc
@@ -2732,6 +2732,10 @@ void partition_manifest::process_anomalies(
     _last_partition_scrub = scrub_timestamp;
     _last_scrubbed_offset = last_scrubbed_offset;
 
+    if (!_last_scrubbed_offset) {
+        _detected_anomalies.last_complete_scrub = scrub_timestamp;
+    }
+
     vlog(
       cst_log.debug,
       "[{}] Anomalies processed: {{ detected: {}, last_partition_scrub: {}, "
diff --git a/src/v/cloud_storage/types.cc b/src/v/cloud_storage/types.cc
index 18572ac6a631a..6ef3dfdfd1df4 100644
--- a/src/v/cloud_storage/types.cc
+++ b/src/v/cloud_storage/types.cc
@@ -264,6 +264,9 @@ anomalies& anomalies::operator+=(anomalies&& other) {
       std::make_move_iterator(other.segment_metadata_anomalies.begin()),
       std::make_move_iterator(other.segment_metadata_anomalies.end()));
 
+    last_complete_scrub = std::max(
+      last_complete_scrub, other.last_complete_scrub);
+
     return *this;
 }
 
diff --git a/src/v/cloud_storage/types.h b/src/v/cloud_storage/types.h
index 99e9100012ff1..da042f2b6bb3f 100644
--- a/src/v/cloud_storage/types.h
+++ b/src/v/cloud_storage/types.h
@@ -391,13 +391,17 @@ struct anomalies
     absl::node_hash_set<segment_meta> missing_segments;
     // Segments that have metadata anomalies (e.g. gaps or overlaps)
     segment_meta_anomalies segment_metadata_anomalies;
+    // Optional timestamp indicating the last time point at which
+    // the scrub of the full log completed.
+    std::optional<model::timestamp> last_complete_scrub;
 
     auto serde_fields() {
         return std::tie(
           missing_partition_manifest,
           missing_spillover_manifests,
           missing_segments,
-          segment_metadata_anomalies);
+          segment_metadata_anomalies,
+          last_complete_scrub);
     }
 
     bool has_value() const;
diff --git a/src/v/redpanda/admin/api-doc/shadow_indexing.json b/src/v/redpanda/admin/api-doc/shadow_indexing.json
index 7e23f5ddb140e..b1bfebbb8d27c 100644
--- a/src/v/redpanda/admin/api-doc/shadow_indexing.json
+++ b/src/v/redpanda/admin/api-doc/shadow_indexing.json
@@ -518,6 +518,10 @@
           "type": "array",
           "items": {"type": "metadata_anomaly"},
           "nullable": true
+        },
+        "last_complete_scrub_at": {
+          "type": "long",
+          "nullable": true
         }
       }
     }
diff --git a/src/v/redpanda/admin_server.cc b/src/v/redpanda/admin_server.cc
index 675f99e56e093..88ea21f68192b 100644
--- a/src/v/redpanda/admin_server.cc
+++ b/src/v/redpanda/admin_server.cc
@@ -5296,6 +5296,10 @@ map_anomalies_to_json(
     json.partition = ntp.tp.partition();
     json.revision_id = initial_rev();
 
+    if (detected.last_complete_scrub) {
+        json.last_complete_scrub_at = detected.last_complete_scrub->value();
+    }
+
     if (detected.missing_partition_manifest) {
         json.missing_partition_manifest = true;
     }
diff --git a/tests/rptest/tests/cloud_storage_scrubber_test.py b/tests/rptest/tests/cloud_storage_scrubber_test.py
index ed0582c7e6fdb..f9d89a3f61a8d 100644
--- a/tests/rptest/tests/cloud_storage_scrubber_test.py
+++ b/tests/rptest/tests/cloud_storage_scrubber_test.py
@@ -118,6 +118,8 @@ def _collect_anomalies(self):
                                                           topic=self.topic,
                                                           partition=pid)
 
+            anomalies.pop("last_complete_scrub_at", None)
+
             ntpr = NTPR(ns=anomalies["ns"],
                         topic=anomalies["topic"],
                         partition=anomalies["partition"],