You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@doris.apache.org by "wyxxxcat (via GitHub)" <gi...@apache.org> on 2026/10/23 02:13:10 UTC
[PR] [fix](cloud) Fix recycler checks for shared storage in clone and rollback scenarios [doris]
wyxxxcat opened a new pull request, #69492:
URL: https://github.com/apache/doris/pull/69492
### What problem does this PR solve?
Issue Number: close #xxx
Related PR: #xxx
Problem Summary:
### Release note
None
### Check List (For Author)
- Test <!-- At least one of them must be included. -->
- [ ] Regression test
- [ ] Unit Test
- [ ] Manual test (add detailed scripts or steps below)
- [ ] No need to test or manual test. Explain why:
- [ ] This is a refactor/code format and no logic has been changed.
- [ ] Previous test can cover this change.
- [ ] No code files have been changed.
- [ ] Other reason <!-- Add your reason? -->
- Behavior changed:
- [ ] No.
- [ ] Yes. <!-- Explain the behavior change -->
- Does this need documentation?
- [ ] No.
- [ ] Yes. <!-- Add document PR link here. eg: https://github.com/apache/doris-website/pull/1214 -->
### Check List (For Reviewer who merge this PR)
- [ ] Confirm the release note
- [ ] Confirm test cases
- [ ] Confirm document
- [ ] Add branch pick label <!-- Add branch pick label that this PR should merge into -->
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For additional commands, e-mail: commits-help@doris.apache.org
Re: [PR] [fix](cloud) Fix recycler checks for shared storage in clone and rollback scenarios [doris]
Posted by "wyxxxcat (via GitHub)" <gi...@apache.org>.
wyxxxcat commented on PR #69492:
URL: https://github.com/apache/doris/pull/69492#issuecomment-6406389788
run buildall
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For additional commands, e-mail: commits-help@doris.apache.org
Re: [PR] [fix](cloud) Fix recycler checks for shared storage in clone and rollback scenarios [doris]
Posted by "wyxxxcat (via GitHub)" <gi...@apache.org>.
wyxxxcat commented on PR #69492:
URL: https://github.com/apache/doris/pull/69492#issuecomment-6406431768
run buildall
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For additional commands, e-mail: commits-help@doris.apache.org
Re: [PR] [fix](cloud) Fix recycler checks for shared storage in clone and rollback scenarios [doris]
Posted by "wyxxxcat (via GitHub)" <gi...@apache.org>.
wyxxxcat commented on PR #69492:
URL: https://github.com/apache/doris/pull/69492#issuecomment-6412613930
run buildall
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For additional commands, e-mail: commits-help@doris.apache.org
Re: [PR] [fix](cloud) Fix recycler checks for shared storage in clone and rollback scenarios [doris]
Posted by "wyxxxcat (via GitHub)" <gi...@apache.org>.
wyxxxcat commented on PR #69492:
URL: https://github.com/apache/doris/pull/69492#issuecomment-6412623955
/review
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For additional commands, e-mail: commits-help@doris.apache.org
Re: [PR] [fix](cloud) Fix recycler checks for shared storage in clone and rollback scenarios [doris]
Posted by "github-actions[bot] (via GitHub)" <gi...@apache.org>.
github-actions[bot] commented on code in PR #69492:
URL: https://github.com/apache/doris/pull/69492#discussion_r4375817896
##########
cloud/src/recycler/checker.cpp:
##########
@@ -519,16 +531,156 @@ int key_exist(TxnKv* txn_kv, std::string_view key) {
}
}
-InstanceChecker::InstanceChecker(std::shared_ptr<TxnKv> txn_kv, const std::string& instance_id)
- : txn_kv_(txn_kv), instance_id_(instance_id) {
+SharedStorageIndex::SharedStorageIndex(std::shared_ptr<TxnKv> txn_kv,
+ std::optional<std::vector<InstanceInfoPB>> instances)
+ : txn_kv_(std::move(txn_kv)), preloaded_instances_(std::move(instances)) {}
+
+std::optional<SharedStorageIndex::StorageLocation> SharedStorageIndex::s3_location(
+ const ObjectStoreInfoPB& obj_info) {
+ // Missing connection info must not make unrelated resources match.
+ if (obj_info.endpoint().empty() || obj_info.bucket().empty()) {
+ return std::nullopt;
+ }
+ return StorageLocation {"s3", obj_info.endpoint(), obj_info.bucket(), obj_info.prefix()};
+}
+
+std::optional<SharedStorageIndex::StorageLocation> SharedStorageIndex::hdfs_location(
+ const HdfsVaultInfo& hdfs_info) {
+ if (hdfs_info.build_conf().fs_name().empty()) {
+ return std::nullopt;
+ }
+ return StorageLocation {"hdfs", hdfs_info.build_conf().fs_name(), "", hdfs_info.prefix()};
+}
+
+void SharedStorageIndex::ensure_loaded() {
+ std::call_once(load_once_, [this] {
+ load_status_ = load_all_instances();
+ if (load_status_ != 0) {
+ // Never expose a partially built index.
+ instance_resource_locations_.clear();
+ location_to_owners_.clear();
+ }
+ });
+}
+
+int SharedStorageIndex::load_all_instances() {
+ std::vector<InstanceInfoPB> instances;
+ if (preloaded_instances_.has_value()) {
+ instances = std::move(*preloaded_instances_);
+ preloaded_instances_.reset();
+ } else if (get_all_instances(txn_kv_.get(), instances) != 0) {
+ return -1;
+ }
+ // Load all instances, even those not scheduled for checking: decoupled
+ // instances can still share storage with the checked one.
+ for (const auto& instance : instances) {
+ if (load_instance(instance) != 0) {
+ return -1;
+ }
+ }
+ return 0;
+}
+
+int SharedStorageIndex::load_instance(const InstanceInfoPB& instance) {
+ const std::string& instance_id = instance.instance_id();
+ // Register the instance even if it has no resources, so that it can be told
+ // apart from an unknown instance.
+ instance_resource_locations_[instance_id];
+ // Locations of the resources in `resource_ids` are filled by the vault scan below.
+ for (const auto& resource_id : instance.resource_ids()) {
+ add_resource(instance_id, resource_id, std::nullopt);
+ }
+ for (const auto& obj_info : instance.obj_info()) {
+ add_resource(instance_id, obj_info.id(), s3_location(obj_info));
+ }
+ if (instance.resource_ids().empty()) {
+ return 0;
+ }
+ return load_storage_vaults(instance_id);
+}
+
+int SharedStorageIndex::load_storage_vaults(const std::string& instance_id) {
+ std::unique_ptr<Transaction> txn;
+ if (txn_kv_->create_txn(&txn) != TxnErrorCode::TXN_OK) {
+ return -1;
+ }
+ auto iter = txn->full_range_get(storage_vault_key({instance_id, ""}),
+ storage_vault_key({instance_id, "\xff"}));
+ for (auto entry = iter->next(); entry.has_value(); entry = iter->next()) {
+ StorageVaultPB vault;
+ if (!vault.ParseFromArray(entry->second.data(), entry->second.size())) {
+ LOG_WARNING("malformed storage vault").tag("instance_id", instance_id);
+ return -1;
+ }
+ std::optional<StorageLocation> location;
+ if (vault.has_obj_info()) {
+ location = s3_location(vault.obj_info());
+ } else if (vault.has_hdfs_info()) {
+ location = hdfs_location(vault.hdfs_info());
+ }
+ add_resource(instance_id, vault.id(), std::move(location));
+ }
+ return iter->is_valid() ? 0 : -1;
+}
+
+void SharedStorageIndex::add_resource(const std::string& instance_id,
+ const std::string& resource_id,
+ std::optional<StorageLocation> location) {
+ auto& resource_location = instance_resource_locations_[instance_id][resource_id];
+ if (!location.has_value()) {
+ return;
+ }
+ if (!resource_location.has_value()) {
+ resource_location = location;
+ }
+ location_to_owners_[std::move(*location)].push_back({instance_id, resource_id});
+}
+
+int SharedStorageIndex::get_shared_resource_owners(std::string_view instance_id,
+ SharedResourceOwners* owners) {
+ ensure_loaded();
+ owners->clear();
+ if (load_status_ != 0) {
+ return -1;
+ }
+ auto instance = instance_resource_locations_.find(std::string(instance_id));
+ if (instance == instance_resource_locations_.end()) {
+ LOG_WARNING("instance not found while checking shared resources")
+ .tag("instance_id", instance_id);
+ return -1;
+ }
+ for (const auto& [resource_id, location] : instance->second) {
+ if (location.has_value()) {
+ (*owners)[resource_id] = location_to_owners_.at(*location);
+ } else {
+ // Without a known location, the resource is only owned by its own instance.
+ (*owners)[resource_id] = {{instance->first, resource_id}};
+ }
+ }
+ return 0;
+}
+
+InstanceChecker::InstanceChecker(std::shared_ptr<TxnKv> txn_kv, const std::string& instance_id,
+ std::shared_ptr<ResourceManager> resource_mgr,
+ std::shared_ptr<SharedStorageIndex> shared_storage_index)
+ : txn_kv_(txn_kv),
+ instance_id_(instance_id),
+ resource_mgr_(std::move(resource_mgr)),
+ shared_storage_index_(std::move(shared_storage_index)) {
snapshot_manager_ = create_snapshot_manager(txn_kv);
- resource_mgr_ = std::make_shared<ResourceManager>(std::move(txn_kv));
- resource_mgr_->init();
+ if (!shared_storage_index_) {
+ shared_storage_index_ = std::make_shared<SharedStorageIndex>(txn_kv);
+ }
+ if (!resource_mgr_) {
+ resource_mgr_ = std::make_shared<ResourceManager>(std::move(txn_kv));
+ resource_mgr_->init();
+ }
}
int InstanceChecker::init(const InstanceInfoPB& instance) {
table_stream_versioned_write_ = instance.multi_version_status() == MULTI_VERSION_WRITE_ONLY ||
instance.multi_version_status() == MULTI_VERSION_READ_WRITE;
+ resource_mgr_->refresh_instance(instance_id_, instance);
Review Comment:
[P2] Populate intermediate ancestor links before clone-chain lookup
The new fallback still fails for supported multilevel clones because this refresh registers only the checked instance. For `C -> B -> A`, let C commit a nonempty rowset for an inherited tablet whose versioned tablet index remains in A. After a checker restart, `CloneChainReader::get_tablet_index()` searches C and B, then returns `TXN_KEY_NOT_FOUND` because B's parent link is absent from this manager. This persists if B is filtered out or a DELETED tombstone: both scheduling and inspection skip it, so healthy C checks keep failing. Populate the full lineage before filtering/scheduling, or fetch missing ancestors on demand, including the HTTP check path. Calling `ResourceManager::init()` alone is insufficient because it does not populate source-snapshot links. Add a checker case with C's local rowset, the index only in A, and an unscheduled B.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For additional commands, e-mail: commits-help@doris.apache.org
Re: [PR] [fix](cloud) Fix recycler checks for shared storage in clone and rollback scenarios [doris]
Posted by "hello-stephen (via GitHub)" <gi...@apache.org>.
hello-stephen commented on PR #69492:
URL: https://github.com/apache/doris/pull/69492#issuecomment-6355741597
Thank you for your contribution to Apache Doris.
Don't know what should be done next? See [How to process your PR](https://doris.apache.org/community/how-to-contribute/review-and-ci).
Please clearly describe your PR:
1. What problem was fixed (it's best to include specific error reporting information). How it was fixed.
2. Which behaviors were modified. What was the previous behavior, what is it now, why was it modified, and what possible impacts might there be.
3. What features were added. Why was this function added?
4. Which code was refactored and why was this part of the code refactored?
5. Which functions were optimized and what is the difference before and after the optimization?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For additional commands, e-mail: commits-help@doris.apache.org
Re: [PR] [fix](cloud) Fix recycler checks for shared storage in clone and rollback scenarios [doris]
Posted by "hello-stephen (via GitHub)" <gi...@apache.org>.
hello-stephen commented on PR #69492:
URL: https://github.com/apache/doris/pull/69492#issuecomment-6413433779
# Cloud UT Coverage Report
Increment line coverage `27.61% (45/163)` :tada:
[Increment coverage report](http://coverage.selectdb-in.cc/coverage/25b7d982cce673a52f30c23db763eb5bc7cb3478_25b7d982cce673a52f30c23db763eb5bc7cb3478_cloud/increment_report/index.html)
[Complete coverage report](http://coverage.selectdb-in.cc/coverage/25b7d982cce673a52f30c23db763eb5bc7cb3478_25b7d982cce673a52f30c23db763eb5bc7cb3478_cloud/report/index.html)
| Category | Coverage |
|-------------------|--------------------|
| Function Coverage | 76.64% (2100/2740) |
| Line Coverage | 65.98% (38728/58694) |
| Region Coverage | 53.60% (36266/67656) |
| Branch Coverage | 57.17% (11824/20682) |
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For additional commands, e-mail: commits-help@doris.apache.org
Re: [PR] [fix](cloud) Fix recycler checks for shared storage in clone and rollback scenarios [doris]
Posted by "wyxxxcat (via GitHub)" <gi...@apache.org>.
wyxxxcat merged PR #69492:
URL: https://github.com/apache/doris/pull/69492
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For additional commands, e-mail: commits-help@doris.apache.org