[DOMStorage] Rate limiting writes to disk.

content/browser/dom_storage/dom_storage_area.cc

 // Copyright 2013 The Chromium Authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file.
 
 #include "content/browser/dom_storage/dom_storage_area.h"
 
+#include <algorithm>
+
 #include "base/bind.h"
 #include "base/location.h"
 #include "base/logging.h"
 #include "base/metrics/histogram.h"
 #include "base/strings/utf_string_conversions.h"
+#include "base/sys_info.h"
 #include "base/time/time.h"
 #include "content/browser/dom_storage/dom_storage_namespace.h"
 #include "content/browser/dom_storage/dom_storage_task_runner.h"
 #include "content/browser/dom_storage/local_storage_database_adapter.h"
 #include "content/browser/dom_storage/session_storage_database.h"
 #include "content/browser/dom_storage/session_storage_database_adapter.h"
 #include "content/common/dom_storage/dom_storage_map.h"
 #include "content/common/dom_storage/dom_storage_types.h"
 #include "storage/browser/database/database_util.h"
 #include "storage/common/database/database_identifier.h"
 #include "storage/common/fileapi/file_system_util.h"
 
 using storage::DatabaseUtil;
 
 namespace content {
 
-static const int kCommitTimerSeconds = 1;
+// Delay for a moment after a value is set in anticipation
+// of other values being set, so changes are batched.
+static const int kCommitDefaultDelay = 5;

[cmumford, 2015/02/11 22:52:27]

What is best practice regarding units? Like kCommitDefaultDelaySecs?

[michaeln, 2015/02/11 23:30:27]

yup, done

 
-DOMStorageArea::CommitBatch::CommitBatch()
-  : clear_all_first(false) {
+// The delay used if a commit is scheduled during startup, also the
+// process uptime duration used as a proxy for 'during startup'.
+static const int kCommitDuringStartupDelay = 60;
+
+// To avoid excessive IO we apply limits to the amount of data being written
+// and the frequency of writes. The specific values used are somewhat arbitrary.
+static const int kMaxDataPerDay = kPerStorageAreaQuota * 2;

[cmumford, 2015/02/11 22:52:27]

kMaxBytesPerDay?

[michaeln, 2015/02/11 23:30:28]

Done.

+static const int kMaxCommitsPerHour = 6;
+
+DOMStorageArea::RateLimiter::RateLimiter(size_t desired_rate,
+                                         base::TimeDelta time_quantum)
+    : rate_(desired_rate), samples_(0), time_quantum_(time_quantum) {
+  DCHECK_GT(desired_rate, 0ul);
+}
+
+void DOMStorageArea::RateLimiter::add_samples(size_t samples) {
+  samples_ += samples;
+}
+
+// Computes the total time needed to process the total samples seen
+// at the desired rate.
+base::TimeDelta DOMStorageArea::RateLimiter::ComputeTimeNeeded() const {
+  return base::TimeDelta::FromInternalValue((samples_ / rate_) *
+                                            time_quantum_.ToInternalValue());
+}
+
+// Given the elapsed time since the start of the rate limiting session,
+// computes the delay needed to mimic having processed the total samples
+// seen at the desired rate.
+base::TimeDelta DOMStorageArea::RateLimiter::ComputeDelayNeeded(
+    const base::TimeDelta elapsed_time) const {
+  base::TimeDelta time_needed = ComputeTimeNeeded();
+  if (time_needed > elapsed_time)
+    return time_needed - elapsed_time;
+  return base::TimeDelta();
+}
+
+DOMStorageArea::CommitBatch::CommitBatch() : clear_all_first(false) {
 }
 DOMStorageArea::CommitBatch::~CommitBatch() {}
 
+size_t DOMStorageArea::CommitBatch::GetDataSize() const {
+  return DOMStorageMap::CountBytes(changed_values);
+}
 
 // static
 const base::FilePath::CharType DOMStorageArea::kDatabaseFileExtension[] =
     FILE_PATH_LITERAL(".localstorage");
 
 // static
 base::FilePath DOMStorageArea::DatabaseFileNameFromOrigin(const GURL& origin) {
   std::string filename = storage::GetIdentifierFromOrigin(origin);
   // There is no base::FilePath.AppendExtension() method, so start with just the
   // extension as the filename, and then InsertBeforeExtension the desired
   // name.
   return base::FilePath().Append(kDatabaseFileExtension).
       InsertBeforeExtensionASCII(filename);
 }
 
 // static
 GURL DOMStorageArea::OriginFromDatabaseFileName(const base::FilePath& name) {
   DCHECK(name.MatchesExtension(kDatabaseFileExtension));
   std::string origin_id =
       name.BaseName().RemoveExtension().MaybeAsASCII();
   return storage::GetOriginFromIdentifier(origin_id);
 }
 
-DOMStorageArea::DOMStorageArea(
-    const GURL& origin, const base::FilePath& directory,
-    DOMStorageTaskRunner* task_runner)
-    : namespace_id_(kLocalStorageNamespaceId), origin_(origin),
+DOMStorageArea::DOMStorageArea(const GURL& origin,
+                               const base::FilePath& directory,
+                               DOMStorageTaskRunner* task_runner)
+    : namespace_id_(kLocalStorageNamespaceId),
+      origin_(origin),
       directory_(directory),
       task_runner_(task_runner),
       map_(new DOMStorageMap(kPerStorageAreaQuota +
                              kPerStorageAreaOverQuotaAllowance)),
       is_initial_import_done_(true),
       is_shutdown_(false),
-      commit_batches_in_flight_(0) {
+      commit_batches_in_flight_(0),
+      start_time_(base::TimeTicks::Now()),
+      data_rate_limiter_(kMaxDataPerDay, base::TimeDelta::FromHours(24)),
+      commit_rate_limiter_(kMaxCommitsPerHour, base::TimeDelta::FromHours(1)) {
   if (!directory.empty()) {
     base::FilePath path = directory.Append(DatabaseFileNameFromOrigin(origin_));
     backing_.reset(new LocalStorageDatabaseAdapter(path));
     is_initial_import_done_ = false;
   }
 }
 
-DOMStorageArea::DOMStorageArea(
-    int64 namespace_id,
-    const std::string& persistent_namespace_id,
-    const GURL& origin,
-    SessionStorageDatabase* session_storage_backing,
-    DOMStorageTaskRunner* task_runner)
+DOMStorageArea::DOMStorageArea(int64 namespace_id,
+                               const std::string& persistent_namespace_id,
+                               const GURL& origin,
+                               SessionStorageDatabase* session_storage_backing,
+                               DOMStorageTaskRunner* task_runner)
     : namespace_id_(namespace_id),
       persistent_namespace_id_(persistent_namespace_id),
       origin_(origin),
       task_runner_(task_runner),
       map_(new DOMStorageMap(kPerStorageAreaQuota +
                              kPerStorageAreaOverQuotaAllowance)),
       session_storage_backing_(session_storage_backing),
       is_initial_import_done_(true),
       is_shutdown_(false),
-      commit_batches_in_flight_(0) {
+      commit_batches_in_flight_(0),
+      start_time_(base::TimeTicks::Now()),
+      data_rate_limiter_(kMaxDataPerDay, base::TimeDelta::FromHours(24)),
+      commit_rate_limiter_(kMaxCommitsPerHour, base::TimeDelta::FromHours(1)) {
   DCHECK(namespace_id != kLocalStorageNamespaceId);
   if (session_storage_backing) {
     backing_.reset(new SessionStorageDatabaseAdapter(
         session_storage_backing, persistent_namespace_id, origin));
     is_initial_import_done_ = false;
   }
 }
 
 DOMStorageArea::~DOMStorageArea() {
 }
@@ skipping 28 matching lines @@
 
 bool DOMStorageArea::SetItem(const base::string16& key,
                              const base::string16& value,
                              base::NullableString16* old_value) {
   if (is_shutdown_)
     return false;
   InitialImportIfNeeded();
   if (!map_->HasOneRef())
     map_ = map_->DeepCopy();
   bool success = map_->SetItem(key, value, old_value);
-  if (success && backing_) {
+  if (success && backing_ &&
+      (old_value->is_null() || old_value->string() != value)) {
     CommitBatch* commit_batch = CreateCommitBatchIfNeeded();
     commit_batch->changed_values[key] = base::NullableString16(value, false);
   }
   return success;
 }
 
 bool DOMStorageArea::RemoveItem(const base::string16& key,
                                 base::string16* old_value) {
   if (is_shutdown_)
     return false;
@@ skipping 54 matching lines @@
   DCHECK_NE(kLocalStorageNamespaceId, destination_namespace_id);
 
   DOMStorageArea* copy = new DOMStorageArea(
       destination_namespace_id, destination_persistent_namespace_id, origin_,
       session_storage_backing_.get(), task_runner_.get());
   copy->map_ = map_;
   copy->is_shutdown_ = is_shutdown_;
   copy->is_initial_import_done_ = true;
 
   // All the uncommitted changes to this area need to happen before the actual
-  // shallow copy is made (scheduled by the upper layer). Another OnCommitTimer
-  // call might be in the event queue at this point, but it's handled gracefully
-  // when it fires.
+  // shallow copy is made (scheduled by the upper layer sometime after return).
   if (commit_batch_)
-    OnCommitTimer();
+    ScheduleImmediateCommit();
   return copy;
 }
 
 bool DOMStorageArea::HasUncommittedChanges() const {
-  DCHECK(!is_shutdown_);
   return commit_batch_.get() || commit_batches_in_flight_;
 }
 
+void DOMStorageArea::ScheduleImmediateCommit() {
+  DCHECK(HasUncommittedChanges());
+  OnCommitTimer();
+}
+
 void DOMStorageArea::DeleteOrigin() {
   DCHECK(!is_shutdown_);
   // This function shouldn't be called for sessionStorage.
   DCHECK(!session_storage_backing_.get());
   if (HasUncommittedChanges()) {
     // TODO(michaeln): This logically deletes the data immediately,
     // and in a matter of a second, deletes the rows from the backing
     // database file, but the file itself will linger until shutdown
     // or purge time. Ideally, this should delete the file more
     // quickly.
@@ skipping 82 matching lines @@
 DOMStorageArea::CommitBatch* DOMStorageArea::CreateCommitBatchIfNeeded() {
   DCHECK(!is_shutdown_);
   if (!commit_batch_) {
     commit_batch_.reset(new CommitBatch());
 
     // Start a timer to commit any changes that accrue in the batch, but only if
     // no commits are currently in flight. In that case the timer will be
     // started after the commits have happened.
     if (!commit_batches_in_flight_) {
       task_runner_->PostDelayedTask(
-          FROM_HERE,
-          base::Bind(&DOMStorageArea::OnCommitTimer, this),
-          base::TimeDelta::FromSeconds(kCommitTimerSeconds));
+          FROM_HERE, base::Bind(&DOMStorageArea::OnCommitTimer, this),
+          ComputeCommitDelay());
     }
   }
   return commit_batch_.get();
 }
 
+base::TimeDelta DOMStorageArea::ComputeCommitDelay() const {
+  base::TimeDelta elapsed_time = base::TimeTicks::Now() - start_time_;
+  base::TimeDelta delay;
+  if (base::SysInfo::Uptime() < kCommitDuringStartupDelay * 1000)
+    base::TimeDelta::FromSeconds(kCommitDuringStartupDelay);

[cmumford, 2015/02/11 22:52:27]

Did you want to assign these values to delay?

[michaeln, 2015/02/11 23:30:28]

that would be a huge improvement!

+  else
+    base::TimeDelta::FromSeconds(kCommitDefaultDelay);
+  delay = std::max(
+      delay,
+      std::max(commit_rate_limiter_.ComputeDelayNeeded(elapsed_time),
+               data_rate_limiter_.ComputeDelayNeeded(elapsed_time)));
+  UMA_HISTOGRAM_LONG_TIMES("LocalStorage.CommitDelay", delay);
+  return delay;
+}
+
 void DOMStorageArea::OnCommitTimer() {
   if (is_shutdown_)
     return;
 
   DCHECK(backing_.get());
 
-  // It's possible that there is nothing to commit, since a shallow copy occured
-  // before the timer fired.
+  // It's possible that there is nothing to commit if an immediate
+  // commit occured after the timer was scheduled but before it fired.
   if (!commit_batch_)
     return;
 
+  commit_rate_limiter_.add_samples(1);
+  data_rate_limiter_.add_samples(commit_batch_->GetDataSize());
+
   // This method executes on the primary sequence, we schedule
   // a task for immediate execution on the commit sequence.
   DCHECK(task_runner_->IsRunningOnPrimarySequence());
   bool success = task_runner_->PostShutdownBlockingTask(
       FROM_HERE,
       DOMStorageTaskRunner::COMMIT_SEQUENCE,
       base::Bind(&DOMStorageArea::CommitChanges, this,
                  base::Owned(commit_batch_.release())));
   ++commit_batches_in_flight_;
   DCHECK(success);
@@ skipping 13 matching lines @@
 
 void DOMStorageArea::OnCommitComplete() {
   // We're back on the primary sequence in this method.
   DCHECK(task_runner_->IsRunningOnPrimarySequence());
   --commit_batches_in_flight_;
   if (is_shutdown_)
     return;
   if (commit_batch_.get() && !commit_batches_in_flight_) {
     // More changes have accrued, restart the timer.
     task_runner_->PostDelayedTask(
-        FROM_HERE,
-        base::Bind(&DOMStorageArea::OnCommitTimer, this),
-        base::TimeDelta::FromSeconds(kCommitTimerSeconds));
+        FROM_HERE, base::Bind(&DOMStorageArea::OnCommitTimer, this),
+        ComputeCommitDelay());
   }
 }
 
 void DOMStorageArea::ShutdownInCommitSequence() {
   // This method executes on the commit sequence.
   DCHECK(task_runner_->IsRunningOnCommitSequence());
   DCHECK(backing_.get());
   if (commit_batch_) {
     // Commit any changes that accrued prior to the timer firing.
     bool success = backing_->CommitChanges(
         commit_batch_->clear_all_first,
         commit_batch_->changed_values);
     DCHECK(success);
   }
   commit_batch_.reset();
   backing_.reset();
   session_storage_backing_ = NULL;
 }
 
 }  // namespace content