rocksdb/cache/cache_reservation_manager.cc

179 lines
6.7 KiB
C++
Raw Normal View History

Refactor WriteBufferManager::CacheRep into CacheReservationManager (#8506) Summary: Context: To help cap various memory usage by a single limit of the block cache capacity, we charge the memory usage through inserting/releasing dummy entries in the block cache. CacheReservationManager is such a class (non thread-safe) responsible for inserting/removing dummy entries to reserve cache space for memory used by the class user. - Refactored the inner private class CacheRep of WriteBufferManager into public CacheReservationManager class for reusability such as for https://github.com/facebook/rocksdb/pull/8428 - Encapsulated implementation details of cache key generation and dummy entries insertion/release in cache reservation as discussed in https://github.com/facebook/rocksdb/pull/8506#discussion_r666550838 - Consolidated increase/decrease cache reservation into one API - UpdateCacheReservation. - Adjusted the previous dummy entry release algorithm in decreasing cache reservation to be loop-releasing dummy entries to stay symmetric to dummy entry insertion algorithm - Made the previous dummy entry release algorithm in delayed decrease mode more aggressive for better decreasing cache reservation when memory used is less likely to increase back. Previously, the algorithms only release 1 dummy entries when new_mem_used < 3/4 * cache_allocated_size_ and cache_allocated_size_ - kSizeDummyEntry > new_mem_used. Now, the algorithms loop-releases as many dummy entries as possible when new_mem_used < 3/4 * cache_allocated_size_. - Updated WriteBufferManager's test cases to adapt to changes on the release algorithm mentioned above and left comment for some test cases for clarity - Replaced the previous cache key prefix generation (utilizing object address related to the cache client) with one that utilizes Cache->NewID() to prevent cache-key collision among dummy entry clients sharing the same cache. The specific collision we are preventing happens when the object address is reused for a new cache-key prefix while the old cache-key using that same object address in its prefix still exists in the cache. This could happen due to that, under LRU cache policy, there is a possible delay in releasing a cache entry after the cache client object owning that cache entry get deallocated. In this case, the object address related to the cache client object can get reused for other client object to generate a new cache-key prefix. This prefix generation can be made obsolete after Peter's unification of all the code generating cache key, mentioned in https://github.com/facebook/rocksdb/pull/8506#discussion_r667265255 Pull Request resolved: https://github.com/facebook/rocksdb/pull/8506 Test Plan: - Passing the added unit tests cache_reservation_manager_test.cc - Passing existing and adjusted write_buffer_manager_test.cc Reviewed By: ajkr Differential Revision: D29644135 Pulled By: hx235 fbshipit-source-id: 0fc93fbfe4a40bb41be85c314f8f2bafa8b741f7
2021-08-24 19:42:31 +00:00
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// This source code is licensed under both the GPLv2 (found in the
// COPYING file in the root directory) and Apache 2.0 License
// (found in the LICENSE.Apache file in the root directory).
//
// Copyright (c) 2011 The LevelDB Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file. See the AUTHORS file for names of contributors.
#include "cache/cache_reservation_manager.h"
#include <cassert>
#include <cstddef>
#include <cstring>
#include <memory>
#include "cache/cache_entry_roles.h"
#include "rocksdb/cache.h"
#include "rocksdb/slice.h"
#include "rocksdb/status.h"
#include "table/block_based/block_based_table_reader.h"
#include "util/coding.h"
namespace ROCKSDB_NAMESPACE {
CacheReservationManager::CacheReservationManager(std::shared_ptr<Cache> cache,
bool delayed_decrease)
Add new API CacheReservationManager::GetTotalMemoryUsage() (#9071) Summary: Note: This PR is the 2nd PR of a bigger PR stack (https://github.com/facebook/rocksdb/pull/9073). Context: `CacheReservationManager::UpdateCacheReservation(std::size_t new_memory_used)` accepts an accumulated total memory used (e.g, used 10MB so far) instead of usage change (e.g, increase by 5 MB, decrease by 5 MB). It has benefits including consolidating API for increase and decrease as described in https://github.com/facebook/rocksdb/pull/8506. However, not every `CacheReservationManager` user keeps track of this accumulated total memory usage. For example, Bloom/Ribbon Filter construction (e.g, [here](https://github.com/facebook/rocksdb/blob/822d729fcd9f7af9f371ca7168e52dbdab898e41/table/block_based/filter_policy.cc#L587) in https://github.com/facebook/rocksdb/pull/9073) does not while WriteBufferManager and compression dictionary buffering do. Considering future users might or might not keep track of this counter and implementing this counter within `CacheReservationManager` is easy due to the passed-in `std::size_t new_memory_used` in calling `CacheReservationManager::UpdateCacheReservation(std::size_t new_memory_used)`, it is proposed to add a new API `CacheReservationManager::GetTotalMemoryUsage()`. As noted in the API comments, since `CacheReservationManager` is NOT thread-safe, external synchronization is needed in calling `UpdateCacheReservation()` if you want `GetTotalMemoryUsed()` returns the indeed latest memory used. - Added and updated private counter `memory_used_` every time `CacheReservationManager::UpdateCacheReservation(std::size_t new_memory_used)` is called regardless if the call returns non-okay status - Added `CacheReservationManager::GetTotalMemoryUsage()` to return `memory_used_` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9071 Test Plan: - Passing new tests - Passing existing tests Reviewed By: ajkr Differential Revision: D31887813 Pulled By: hx235 fbshipit-source-id: 9a09f0c8683822673260362894c878b61ee60ceb
2021-11-09 16:15:29 +00:00
: delayed_decrease_(delayed_decrease),
cache_allocated_size_(0),
memory_used_(0) {
Refactor WriteBufferManager::CacheRep into CacheReservationManager (#8506) Summary: Context: To help cap various memory usage by a single limit of the block cache capacity, we charge the memory usage through inserting/releasing dummy entries in the block cache. CacheReservationManager is such a class (non thread-safe) responsible for inserting/removing dummy entries to reserve cache space for memory used by the class user. - Refactored the inner private class CacheRep of WriteBufferManager into public CacheReservationManager class for reusability such as for https://github.com/facebook/rocksdb/pull/8428 - Encapsulated implementation details of cache key generation and dummy entries insertion/release in cache reservation as discussed in https://github.com/facebook/rocksdb/pull/8506#discussion_r666550838 - Consolidated increase/decrease cache reservation into one API - UpdateCacheReservation. - Adjusted the previous dummy entry release algorithm in decreasing cache reservation to be loop-releasing dummy entries to stay symmetric to dummy entry insertion algorithm - Made the previous dummy entry release algorithm in delayed decrease mode more aggressive for better decreasing cache reservation when memory used is less likely to increase back. Previously, the algorithms only release 1 dummy entries when new_mem_used < 3/4 * cache_allocated_size_ and cache_allocated_size_ - kSizeDummyEntry > new_mem_used. Now, the algorithms loop-releases as many dummy entries as possible when new_mem_used < 3/4 * cache_allocated_size_. - Updated WriteBufferManager's test cases to adapt to changes on the release algorithm mentioned above and left comment for some test cases for clarity - Replaced the previous cache key prefix generation (utilizing object address related to the cache client) with one that utilizes Cache->NewID() to prevent cache-key collision among dummy entry clients sharing the same cache. The specific collision we are preventing happens when the object address is reused for a new cache-key prefix while the old cache-key using that same object address in its prefix still exists in the cache. This could happen due to that, under LRU cache policy, there is a possible delay in releasing a cache entry after the cache client object owning that cache entry get deallocated. In this case, the object address related to the cache client object can get reused for other client object to generate a new cache-key prefix. This prefix generation can be made obsolete after Peter's unification of all the code generating cache key, mentioned in https://github.com/facebook/rocksdb/pull/8506#discussion_r667265255 Pull Request resolved: https://github.com/facebook/rocksdb/pull/8506 Test Plan: - Passing the added unit tests cache_reservation_manager_test.cc - Passing existing and adjusted write_buffer_manager_test.cc Reviewed By: ajkr Differential Revision: D29644135 Pulled By: hx235 fbshipit-source-id: 0fc93fbfe4a40bb41be85c314f8f2bafa8b741f7
2021-08-24 19:42:31 +00:00
assert(cache != nullptr);
cache_ = cache;
std::memset(cache_key_, 0, kCacheKeyPrefixSize + kMaxVarint64Length);
EncodeVarint64(cache_key_, cache_->NewId());
}
CacheReservationManager::~CacheReservationManager() {
for (auto* handle : dummy_handles_) {
cache_->Release(handle, true);
}
}
template <CacheEntryRole R>
Status CacheReservationManager::UpdateCacheReservation(
std::size_t new_mem_used) {
Add new API CacheReservationManager::GetTotalMemoryUsage() (#9071) Summary: Note: This PR is the 2nd PR of a bigger PR stack (https://github.com/facebook/rocksdb/pull/9073). Context: `CacheReservationManager::UpdateCacheReservation(std::size_t new_memory_used)` accepts an accumulated total memory used (e.g, used 10MB so far) instead of usage change (e.g, increase by 5 MB, decrease by 5 MB). It has benefits including consolidating API for increase and decrease as described in https://github.com/facebook/rocksdb/pull/8506. However, not every `CacheReservationManager` user keeps track of this accumulated total memory usage. For example, Bloom/Ribbon Filter construction (e.g, [here](https://github.com/facebook/rocksdb/blob/822d729fcd9f7af9f371ca7168e52dbdab898e41/table/block_based/filter_policy.cc#L587) in https://github.com/facebook/rocksdb/pull/9073) does not while WriteBufferManager and compression dictionary buffering do. Considering future users might or might not keep track of this counter and implementing this counter within `CacheReservationManager` is easy due to the passed-in `std::size_t new_memory_used` in calling `CacheReservationManager::UpdateCacheReservation(std::size_t new_memory_used)`, it is proposed to add a new API `CacheReservationManager::GetTotalMemoryUsage()`. As noted in the API comments, since `CacheReservationManager` is NOT thread-safe, external synchronization is needed in calling `UpdateCacheReservation()` if you want `GetTotalMemoryUsed()` returns the indeed latest memory used. - Added and updated private counter `memory_used_` every time `CacheReservationManager::UpdateCacheReservation(std::size_t new_memory_used)` is called regardless if the call returns non-okay status - Added `CacheReservationManager::GetTotalMemoryUsage()` to return `memory_used_` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9071 Test Plan: - Passing new tests - Passing existing tests Reviewed By: ajkr Differential Revision: D31887813 Pulled By: hx235 fbshipit-source-id: 9a09f0c8683822673260362894c878b61ee60ceb
2021-11-09 16:15:29 +00:00
memory_used_ = new_mem_used;
Refactor WriteBufferManager::CacheRep into CacheReservationManager (#8506) Summary: Context: To help cap various memory usage by a single limit of the block cache capacity, we charge the memory usage through inserting/releasing dummy entries in the block cache. CacheReservationManager is such a class (non thread-safe) responsible for inserting/removing dummy entries to reserve cache space for memory used by the class user. - Refactored the inner private class CacheRep of WriteBufferManager into public CacheReservationManager class for reusability such as for https://github.com/facebook/rocksdb/pull/8428 - Encapsulated implementation details of cache key generation and dummy entries insertion/release in cache reservation as discussed in https://github.com/facebook/rocksdb/pull/8506#discussion_r666550838 - Consolidated increase/decrease cache reservation into one API - UpdateCacheReservation. - Adjusted the previous dummy entry release algorithm in decreasing cache reservation to be loop-releasing dummy entries to stay symmetric to dummy entry insertion algorithm - Made the previous dummy entry release algorithm in delayed decrease mode more aggressive for better decreasing cache reservation when memory used is less likely to increase back. Previously, the algorithms only release 1 dummy entries when new_mem_used < 3/4 * cache_allocated_size_ and cache_allocated_size_ - kSizeDummyEntry > new_mem_used. Now, the algorithms loop-releases as many dummy entries as possible when new_mem_used < 3/4 * cache_allocated_size_. - Updated WriteBufferManager's test cases to adapt to changes on the release algorithm mentioned above and left comment for some test cases for clarity - Replaced the previous cache key prefix generation (utilizing object address related to the cache client) with one that utilizes Cache->NewID() to prevent cache-key collision among dummy entry clients sharing the same cache. The specific collision we are preventing happens when the object address is reused for a new cache-key prefix while the old cache-key using that same object address in its prefix still exists in the cache. This could happen due to that, under LRU cache policy, there is a possible delay in releasing a cache entry after the cache client object owning that cache entry get deallocated. In this case, the object address related to the cache client object can get reused for other client object to generate a new cache-key prefix. This prefix generation can be made obsolete after Peter's unification of all the code generating cache key, mentioned in https://github.com/facebook/rocksdb/pull/8506#discussion_r667265255 Pull Request resolved: https://github.com/facebook/rocksdb/pull/8506 Test Plan: - Passing the added unit tests cache_reservation_manager_test.cc - Passing existing and adjusted write_buffer_manager_test.cc Reviewed By: ajkr Differential Revision: D29644135 Pulled By: hx235 fbshipit-source-id: 0fc93fbfe4a40bb41be85c314f8f2bafa8b741f7
2021-08-24 19:42:31 +00:00
std::size_t cur_cache_allocated_size =
cache_allocated_size_.load(std::memory_order_relaxed);
if (new_mem_used == cur_cache_allocated_size) {
return Status::OK();
} else if (new_mem_used > cur_cache_allocated_size) {
Status s = IncreaseCacheReservation<R>(new_mem_used);
return s;
} else {
// In delayed decrease mode, we don't decrease cache reservation
// untill the memory usage is less than 3/4 of what we reserve
// in the cache.
// We do this because
// (1) Dummy entry insertion is expensive in block cache
// (2) Delayed releasing previously inserted dummy entries can save such
// expensive dummy entry insertion on memory increase in the near future,
// which is likely to happen when the memory usage is greater than or equal
// to 3/4 of what we reserve
if (delayed_decrease_ && new_mem_used >= cur_cache_allocated_size / 4 * 3) {
return Status::OK();
} else {
Status s = DecreaseCacheReservation(new_mem_used);
return s;
}
}
}
// Explicitly instantiate templates for "CacheEntryRole" values we use.
// This makes it possible to keep the template definitions in the .cc file.
template Status CacheReservationManager::UpdateCacheReservation<
CacheEntryRole::kWriteBuffer>(std::size_t new_mem_used);
template Status CacheReservationManager::UpdateCacheReservation<
CacheEntryRole::kCompressionDictionaryBuildingBuffer>(
std::size_t new_mem_used);
Refactor WriteBufferManager::CacheRep into CacheReservationManager (#8506) Summary: Context: To help cap various memory usage by a single limit of the block cache capacity, we charge the memory usage through inserting/releasing dummy entries in the block cache. CacheReservationManager is such a class (non thread-safe) responsible for inserting/removing dummy entries to reserve cache space for memory used by the class user. - Refactored the inner private class CacheRep of WriteBufferManager into public CacheReservationManager class for reusability such as for https://github.com/facebook/rocksdb/pull/8428 - Encapsulated implementation details of cache key generation and dummy entries insertion/release in cache reservation as discussed in https://github.com/facebook/rocksdb/pull/8506#discussion_r666550838 - Consolidated increase/decrease cache reservation into one API - UpdateCacheReservation. - Adjusted the previous dummy entry release algorithm in decreasing cache reservation to be loop-releasing dummy entries to stay symmetric to dummy entry insertion algorithm - Made the previous dummy entry release algorithm in delayed decrease mode more aggressive for better decreasing cache reservation when memory used is less likely to increase back. Previously, the algorithms only release 1 dummy entries when new_mem_used < 3/4 * cache_allocated_size_ and cache_allocated_size_ - kSizeDummyEntry > new_mem_used. Now, the algorithms loop-releases as many dummy entries as possible when new_mem_used < 3/4 * cache_allocated_size_. - Updated WriteBufferManager's test cases to adapt to changes on the release algorithm mentioned above and left comment for some test cases for clarity - Replaced the previous cache key prefix generation (utilizing object address related to the cache client) with one that utilizes Cache->NewID() to prevent cache-key collision among dummy entry clients sharing the same cache. The specific collision we are preventing happens when the object address is reused for a new cache-key prefix while the old cache-key using that same object address in its prefix still exists in the cache. This could happen due to that, under LRU cache policy, there is a possible delay in releasing a cache entry after the cache client object owning that cache entry get deallocated. In this case, the object address related to the cache client object can get reused for other client object to generate a new cache-key prefix. This prefix generation can be made obsolete after Peter's unification of all the code generating cache key, mentioned in https://github.com/facebook/rocksdb/pull/8506#discussion_r667265255 Pull Request resolved: https://github.com/facebook/rocksdb/pull/8506 Test Plan: - Passing the added unit tests cache_reservation_manager_test.cc - Passing existing and adjusted write_buffer_manager_test.cc Reviewed By: ajkr Differential Revision: D29644135 Pulled By: hx235 fbshipit-source-id: 0fc93fbfe4a40bb41be85c314f8f2bafa8b741f7
2021-08-24 19:42:31 +00:00
// For cache reservation manager unit tests
template Status CacheReservationManager::UpdateCacheReservation<
CacheEntryRole::kMisc>(std::size_t new_mem_used);
RAII support for per cache reservation through handle (#9130) Summary: Note: This PR is the 3rd PR of a bigger PR stack (https://github.com/facebook/rocksdb/issues/9073) and depends on the second PR (https://github.com/facebook/rocksdb/pull/9071). **See changes from this PR only https://github.com/facebook/rocksdb/pull/9130/commits/00447324d082136b0e777d3ab6a3df3a8452c633** Context: pdillinger brought up a good [point](https://github.com/facebook/rocksdb/pull/9073#discussion_r741478309) about lacking RAII support for per cache reservation in `CacheReservationManager` when reviewing https://github.com/facebook/rocksdb/pull/9073. To summarize the discussion, the current API `CacheReservationManager::UpdateCacheReservation()` requires callers to explicitly calculate and pass in a correct`new_mem_used` to release a cache reservation (if they don't want to rely on the clean-up during `CacheReservationManager`'s destruction - such as they want to release it earlier). While this implementation has convenience in some use-case such as `WriteBufferManager`, where [reservation](https://github.com/facebook/rocksdb/blob/main/memtable/write_buffer_manager.cc#L69-L91) and [release](https://github.com/facebook/rocksdb/blob/main/memtable/write_buffer_manager.cc#L109-L129) amounts do not necessarily correspond symmetrically and thus a flexible `new_mem_used` inputing is needed, it can be prone to caller's calculation error as well as cause a mass of codes in releasing cache in other use-case such as filter construction, where reservation and release amounts do correspond symmetrically and many code paths requiring a cache release, as [pointed](https://github.com/facebook/rocksdb/pull/9073#discussion_r741478309) out by pdillinger. Therefore we decided to provide a new API in `CacheReservationManager` to update reservation with better RAII support for per cache reservation, using a handle to manage the life time of that particular cache reservation. - Added a new class `CacheReservationHandle` - Added a new API `CacheReservationManager::MakeCacheReservation()` that outputs a `CacheReservationHandle` for managing the reservation - Updated class comments to clarify two different cache reservation methods Tests: - Passing new tests Pull Request resolved: https://github.com/facebook/rocksdb/pull/9130 Reviewed By: pdillinger Differential Revision: D32199446 Pulled By: hx235 fbshipit-source-id: 1cba7c636e5ecfb55b0c1e0c2d218cc9b5b30b4e
2021-11-09 20:04:51 +00:00
template <CacheEntryRole R>
Status CacheReservationManager::MakeCacheReservation(
std::size_t incremental_memory_used,
std::unique_ptr<CacheReservationHandle<R>>* handle) {
assert(handle != nullptr);
Status s =
UpdateCacheReservation<R>(GetTotalMemoryUsed() + incremental_memory_used);
(*handle).reset(new CacheReservationHandle<R>(incremental_memory_used,
shared_from_this()));
return s;
}
template Status
CacheReservationManager::MakeCacheReservation<CacheEntryRole::kMisc>(
std::size_t incremental_memory_used,
std::unique_ptr<CacheReservationHandle<CacheEntryRole::kMisc>>* handle);
Refactor WriteBufferManager::CacheRep into CacheReservationManager (#8506) Summary: Context: To help cap various memory usage by a single limit of the block cache capacity, we charge the memory usage through inserting/releasing dummy entries in the block cache. CacheReservationManager is such a class (non thread-safe) responsible for inserting/removing dummy entries to reserve cache space for memory used by the class user. - Refactored the inner private class CacheRep of WriteBufferManager into public CacheReservationManager class for reusability such as for https://github.com/facebook/rocksdb/pull/8428 - Encapsulated implementation details of cache key generation and dummy entries insertion/release in cache reservation as discussed in https://github.com/facebook/rocksdb/pull/8506#discussion_r666550838 - Consolidated increase/decrease cache reservation into one API - UpdateCacheReservation. - Adjusted the previous dummy entry release algorithm in decreasing cache reservation to be loop-releasing dummy entries to stay symmetric to dummy entry insertion algorithm - Made the previous dummy entry release algorithm in delayed decrease mode more aggressive for better decreasing cache reservation when memory used is less likely to increase back. Previously, the algorithms only release 1 dummy entries when new_mem_used < 3/4 * cache_allocated_size_ and cache_allocated_size_ - kSizeDummyEntry > new_mem_used. Now, the algorithms loop-releases as many dummy entries as possible when new_mem_used < 3/4 * cache_allocated_size_. - Updated WriteBufferManager's test cases to adapt to changes on the release algorithm mentioned above and left comment for some test cases for clarity - Replaced the previous cache key prefix generation (utilizing object address related to the cache client) with one that utilizes Cache->NewID() to prevent cache-key collision among dummy entry clients sharing the same cache. The specific collision we are preventing happens when the object address is reused for a new cache-key prefix while the old cache-key using that same object address in its prefix still exists in the cache. This could happen due to that, under LRU cache policy, there is a possible delay in releasing a cache entry after the cache client object owning that cache entry get deallocated. In this case, the object address related to the cache client object can get reused for other client object to generate a new cache-key prefix. This prefix generation can be made obsolete after Peter's unification of all the code generating cache key, mentioned in https://github.com/facebook/rocksdb/pull/8506#discussion_r667265255 Pull Request resolved: https://github.com/facebook/rocksdb/pull/8506 Test Plan: - Passing the added unit tests cache_reservation_manager_test.cc - Passing existing and adjusted write_buffer_manager_test.cc Reviewed By: ajkr Differential Revision: D29644135 Pulled By: hx235 fbshipit-source-id: 0fc93fbfe4a40bb41be85c314f8f2bafa8b741f7
2021-08-24 19:42:31 +00:00
template <CacheEntryRole R>
Status CacheReservationManager::IncreaseCacheReservation(
std::size_t new_mem_used) {
Status return_status = Status::OK();
while (new_mem_used > cache_allocated_size_.load(std::memory_order_relaxed)) {
Cache::Handle* handle = nullptr;
return_status = cache_->Insert(GetNextCacheKey(), nullptr, kSizeDummyEntry,
GetNoopDeleterForRole<R>(), &handle);
if (return_status != Status::OK()) {
return return_status;
}
dummy_handles_.push_back(handle);
cache_allocated_size_ += kSizeDummyEntry;
}
return return_status;
}
Status CacheReservationManager::DecreaseCacheReservation(
std::size_t new_mem_used) {
Status return_status = Status::OK();
// Decrease to the smallest multiple of kSizeDummyEntry that is greater than
// or equal to new_mem_used We do addition instead of new_mem_used <=
// cache_allocated_size_.load(std::memory_order_relaxed) - kSizeDummyEntry to
// avoid underflow of size_t when cache_allocated_size_ = 0
while (new_mem_used + kSizeDummyEntry <=
cache_allocated_size_.load(std::memory_order_relaxed)) {
assert(!dummy_handles_.empty());
auto* handle = dummy_handles_.back();
cache_->Release(handle, true);
dummy_handles_.pop_back();
cache_allocated_size_ -= kSizeDummyEntry;
}
return return_status;
}
std::size_t CacheReservationManager::GetTotalReservedCacheSize() {
return cache_allocated_size_.load(std::memory_order_relaxed);
}
Add new API CacheReservationManager::GetTotalMemoryUsage() (#9071) Summary: Note: This PR is the 2nd PR of a bigger PR stack (https://github.com/facebook/rocksdb/pull/9073). Context: `CacheReservationManager::UpdateCacheReservation(std::size_t new_memory_used)` accepts an accumulated total memory used (e.g, used 10MB so far) instead of usage change (e.g, increase by 5 MB, decrease by 5 MB). It has benefits including consolidating API for increase and decrease as described in https://github.com/facebook/rocksdb/pull/8506. However, not every `CacheReservationManager` user keeps track of this accumulated total memory usage. For example, Bloom/Ribbon Filter construction (e.g, [here](https://github.com/facebook/rocksdb/blob/822d729fcd9f7af9f371ca7168e52dbdab898e41/table/block_based/filter_policy.cc#L587) in https://github.com/facebook/rocksdb/pull/9073) does not while WriteBufferManager and compression dictionary buffering do. Considering future users might or might not keep track of this counter and implementing this counter within `CacheReservationManager` is easy due to the passed-in `std::size_t new_memory_used` in calling `CacheReservationManager::UpdateCacheReservation(std::size_t new_memory_used)`, it is proposed to add a new API `CacheReservationManager::GetTotalMemoryUsage()`. As noted in the API comments, since `CacheReservationManager` is NOT thread-safe, external synchronization is needed in calling `UpdateCacheReservation()` if you want `GetTotalMemoryUsed()` returns the indeed latest memory used. - Added and updated private counter `memory_used_` every time `CacheReservationManager::UpdateCacheReservation(std::size_t new_memory_used)` is called regardless if the call returns non-okay status - Added `CacheReservationManager::GetTotalMemoryUsage()` to return `memory_used_` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9071 Test Plan: - Passing new tests - Passing existing tests Reviewed By: ajkr Differential Revision: D31887813 Pulled By: hx235 fbshipit-source-id: 9a09f0c8683822673260362894c878b61ee60ceb
2021-11-09 16:15:29 +00:00
std::size_t CacheReservationManager::GetTotalMemoryUsed() {
return memory_used_;
}
Refactor WriteBufferManager::CacheRep into CacheReservationManager (#8506) Summary: Context: To help cap various memory usage by a single limit of the block cache capacity, we charge the memory usage through inserting/releasing dummy entries in the block cache. CacheReservationManager is such a class (non thread-safe) responsible for inserting/removing dummy entries to reserve cache space for memory used by the class user. - Refactored the inner private class CacheRep of WriteBufferManager into public CacheReservationManager class for reusability such as for https://github.com/facebook/rocksdb/pull/8428 - Encapsulated implementation details of cache key generation and dummy entries insertion/release in cache reservation as discussed in https://github.com/facebook/rocksdb/pull/8506#discussion_r666550838 - Consolidated increase/decrease cache reservation into one API - UpdateCacheReservation. - Adjusted the previous dummy entry release algorithm in decreasing cache reservation to be loop-releasing dummy entries to stay symmetric to dummy entry insertion algorithm - Made the previous dummy entry release algorithm in delayed decrease mode more aggressive for better decreasing cache reservation when memory used is less likely to increase back. Previously, the algorithms only release 1 dummy entries when new_mem_used < 3/4 * cache_allocated_size_ and cache_allocated_size_ - kSizeDummyEntry > new_mem_used. Now, the algorithms loop-releases as many dummy entries as possible when new_mem_used < 3/4 * cache_allocated_size_. - Updated WriteBufferManager's test cases to adapt to changes on the release algorithm mentioned above and left comment for some test cases for clarity - Replaced the previous cache key prefix generation (utilizing object address related to the cache client) with one that utilizes Cache->NewID() to prevent cache-key collision among dummy entry clients sharing the same cache. The specific collision we are preventing happens when the object address is reused for a new cache-key prefix while the old cache-key using that same object address in its prefix still exists in the cache. This could happen due to that, under LRU cache policy, there is a possible delay in releasing a cache entry after the cache client object owning that cache entry get deallocated. In this case, the object address related to the cache client object can get reused for other client object to generate a new cache-key prefix. This prefix generation can be made obsolete after Peter's unification of all the code generating cache key, mentioned in https://github.com/facebook/rocksdb/pull/8506#discussion_r667265255 Pull Request resolved: https://github.com/facebook/rocksdb/pull/8506 Test Plan: - Passing the added unit tests cache_reservation_manager_test.cc - Passing existing and adjusted write_buffer_manager_test.cc Reviewed By: ajkr Differential Revision: D29644135 Pulled By: hx235 fbshipit-source-id: 0fc93fbfe4a40bb41be85c314f8f2bafa8b741f7
2021-08-24 19:42:31 +00:00
Slice CacheReservationManager::GetNextCacheKey() {
// Calling this function will have the side-effect of changing the
// underlying cache_key_ that is shared among other keys generated from this
// fucntion. Therefore please make sure the previous keys are saved/copied
// before calling this function.
std::memset(cache_key_ + kCacheKeyPrefixSize, 0, kMaxVarint64Length);
char* end =
EncodeVarint64(cache_key_ + kCacheKeyPrefixSize, next_cache_key_id_++);
return Slice(cache_key_, static_cast<std::size_t>(end - cache_key_));
}
RAII support for per cache reservation through handle (#9130) Summary: Note: This PR is the 3rd PR of a bigger PR stack (https://github.com/facebook/rocksdb/issues/9073) and depends on the second PR (https://github.com/facebook/rocksdb/pull/9071). **See changes from this PR only https://github.com/facebook/rocksdb/pull/9130/commits/00447324d082136b0e777d3ab6a3df3a8452c633** Context: pdillinger brought up a good [point](https://github.com/facebook/rocksdb/pull/9073#discussion_r741478309) about lacking RAII support for per cache reservation in `CacheReservationManager` when reviewing https://github.com/facebook/rocksdb/pull/9073. To summarize the discussion, the current API `CacheReservationManager::UpdateCacheReservation()` requires callers to explicitly calculate and pass in a correct`new_mem_used` to release a cache reservation (if they don't want to rely on the clean-up during `CacheReservationManager`'s destruction - such as they want to release it earlier). While this implementation has convenience in some use-case such as `WriteBufferManager`, where [reservation](https://github.com/facebook/rocksdb/blob/main/memtable/write_buffer_manager.cc#L69-L91) and [release](https://github.com/facebook/rocksdb/blob/main/memtable/write_buffer_manager.cc#L109-L129) amounts do not necessarily correspond symmetrically and thus a flexible `new_mem_used` inputing is needed, it can be prone to caller's calculation error as well as cause a mass of codes in releasing cache in other use-case such as filter construction, where reservation and release amounts do correspond symmetrically and many code paths requiring a cache release, as [pointed](https://github.com/facebook/rocksdb/pull/9073#discussion_r741478309) out by pdillinger. Therefore we decided to provide a new API in `CacheReservationManager` to update reservation with better RAII support for per cache reservation, using a handle to manage the life time of that particular cache reservation. - Added a new class `CacheReservationHandle` - Added a new API `CacheReservationManager::MakeCacheReservation()` that outputs a `CacheReservationHandle` for managing the reservation - Updated class comments to clarify two different cache reservation methods Tests: - Passing new tests Pull Request resolved: https://github.com/facebook/rocksdb/pull/9130 Reviewed By: pdillinger Differential Revision: D32199446 Pulled By: hx235 fbshipit-source-id: 1cba7c636e5ecfb55b0c1e0c2d218cc9b5b30b4e
2021-11-09 20:04:51 +00:00
template <CacheEntryRole R>
CacheReservationHandle<R>::CacheReservationHandle(
std::size_t incremental_memory_used,
std::shared_ptr<CacheReservationManager> cache_res_mgr)
: incremental_memory_used_(incremental_memory_used) {
assert(cache_res_mgr != nullptr);
cache_res_mgr_ = cache_res_mgr;
}
template <CacheEntryRole R>
CacheReservationHandle<R>::~CacheReservationHandle() {
assert(cache_res_mgr_ != nullptr);
assert(cache_res_mgr_->GetTotalMemoryUsed() >= incremental_memory_used_);
Status s = cache_res_mgr_->UpdateCacheReservation<R>(
cache_res_mgr_->GetTotalMemoryUsed() - incremental_memory_used_);
s.PermitUncheckedError();
}
// Explicitly instantiate templates for "CacheEntryRole" values we use.
// This makes it possible to keep the template definitions in the .cc file.
template class CacheReservationHandle<CacheEntryRole::kMisc>;
Add new API CacheReservationManager::GetTotalMemoryUsage() (#9071) Summary: Note: This PR is the 2nd PR of a bigger PR stack (https://github.com/facebook/rocksdb/pull/9073). Context: `CacheReservationManager::UpdateCacheReservation(std::size_t new_memory_used)` accepts an accumulated total memory used (e.g, used 10MB so far) instead of usage change (e.g, increase by 5 MB, decrease by 5 MB). It has benefits including consolidating API for increase and decrease as described in https://github.com/facebook/rocksdb/pull/8506. However, not every `CacheReservationManager` user keeps track of this accumulated total memory usage. For example, Bloom/Ribbon Filter construction (e.g, [here](https://github.com/facebook/rocksdb/blob/822d729fcd9f7af9f371ca7168e52dbdab898e41/table/block_based/filter_policy.cc#L587) in https://github.com/facebook/rocksdb/pull/9073) does not while WriteBufferManager and compression dictionary buffering do. Considering future users might or might not keep track of this counter and implementing this counter within `CacheReservationManager` is easy due to the passed-in `std::size_t new_memory_used` in calling `CacheReservationManager::UpdateCacheReservation(std::size_t new_memory_used)`, it is proposed to add a new API `CacheReservationManager::GetTotalMemoryUsage()`. As noted in the API comments, since `CacheReservationManager` is NOT thread-safe, external synchronization is needed in calling `UpdateCacheReservation()` if you want `GetTotalMemoryUsed()` returns the indeed latest memory used. - Added and updated private counter `memory_used_` every time `CacheReservationManager::UpdateCacheReservation(std::size_t new_memory_used)` is called regardless if the call returns non-okay status - Added `CacheReservationManager::GetTotalMemoryUsage()` to return `memory_used_` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9071 Test Plan: - Passing new tests - Passing existing tests Reviewed By: ajkr Differential Revision: D31887813 Pulled By: hx235 fbshipit-source-id: 9a09f0c8683822673260362894c878b61ee60ceb
2021-11-09 16:15:29 +00:00
} // namespace ROCKSDB_NAMESPACE