2011-03-19 06:37:00 +08:00
|
|
|
// 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.
|
|
|
|
|
2011-03-31 02:35:40 +08:00
|
|
|
#include "leveldb/cache.h"
|
2011-03-19 06:37:00 +08:00
|
|
|
|
|
|
|
#include <vector>
|
|
|
|
#include "util/coding.h"
|
|
|
|
#include "util/testharness.h"
|
|
|
|
|
|
|
|
namespace leveldb {
|
|
|
|
|
|
|
|
// Conversions between numeric keys/values and the types expected by Cache.
|
|
|
|
static std::string EncodeKey(int k) {
|
|
|
|
std::string result;
|
|
|
|
PutFixed32(&result, k);
|
|
|
|
return result;
|
|
|
|
}
|
|
|
|
static int DecodeKey(const Slice& k) {
|
|
|
|
assert(k.size() == 4);
|
|
|
|
return DecodeFixed32(k.data());
|
|
|
|
}
|
|
|
|
static void* EncodeValue(uintptr_t v) { return reinterpret_cast<void*>(v); }
|
|
|
|
static int DecodeValue(void* v) { return reinterpret_cast<uintptr_t>(v); }
|
|
|
|
|
|
|
|
class CacheTest {
|
|
|
|
public:
|
|
|
|
static CacheTest* current_;
|
|
|
|
|
|
|
|
static void Deleter(const Slice& key, void* v) {
|
|
|
|
current_->deleted_keys_.push_back(DecodeKey(key));
|
|
|
|
current_->deleted_values_.push_back(DecodeValue(v));
|
|
|
|
}
|
|
|
|
|
2011-08-23 05:08:51 +08:00
|
|
|
static const int kCacheSize = 1000;
|
2011-03-19 06:37:00 +08:00
|
|
|
std::vector<int> deleted_keys_;
|
|
|
|
std::vector<int> deleted_values_;
|
|
|
|
Cache* cache_;
|
|
|
|
|
|
|
|
CacheTest() : cache_(NewLRUCache(kCacheSize)) {
|
|
|
|
current_ = this;
|
|
|
|
}
|
|
|
|
|
|
|
|
~CacheTest() {
|
|
|
|
delete cache_;
|
|
|
|
}
|
|
|
|
|
|
|
|
int Lookup(int key) {
|
|
|
|
Cache::Handle* handle = cache_->Lookup(EncodeKey(key));
|
|
|
|
const int r = (handle == NULL) ? -1 : DecodeValue(cache_->Value(handle));
|
|
|
|
if (handle != NULL) {
|
|
|
|
cache_->Release(handle);
|
|
|
|
}
|
|
|
|
return r;
|
|
|
|
}
|
|
|
|
|
|
|
|
void Insert(int key, int value, int charge = 1) {
|
|
|
|
cache_->Release(cache_->Insert(EncodeKey(key), EncodeValue(value), charge,
|
|
|
|
&CacheTest::Deleter));
|
|
|
|
}
|
|
|
|
|
fix problems in LevelDB's caching code
Background:
LevelDB uses a cache (util/cache.h, util/cache.cc) of (key,value)
pairs for two purposes:
- a cache of (table, file handle) pairs
- a cache of blocks
The cache places the (key,value) pairs in a reference-counted
wrapper. When it returns a value, it returns a reference to this
wrapper. When the client has finished using the reference and
its enclosed (key,value), it calls Release() to decrement the
reference count.
Each (key,value) pair has an associated resource usage. The
cache maintains the sum of the usages of the elements it holds,
and removes values as needed to keep the sum below a capacity
threshold. It maintains an LRU list so that it will remove the
least-recently used elements first.
The max_open_files option to LevelDB sets the size of the cache
of (table, file handle) pairs. The option is not used in any
other way.
The observed behaviour:
If LevelDB at any time used more file handles concurrently than
the cache size set via max_open_files, it attempted to reduce the
number by evicting entries from the table cache. This could
happen most easily during compaction, and if max_open_files was
low. Because the handles were in use, their reference count did
not drop to zero, and so the usage sum in the cache was not
modified by the evictions. Subsequent Insert() calls returned
valid handles, but their entries were immediately evicted from
the cache, which though empty still acted as though full. As a
result, there was effectively no caching, and the number of open
file handles rose []ly until it hit system-imposed limits and
the process died.
If one set max_open_files lower, the cache was more likely to
exhibit this beahviour, and cause the process to run out of file
descriptors. That is, max_open_files acted in almost exactly the
opposite manner from what was intended.
The problems:
1. The cache kept all elements on its LRU list eligible for capacity
eviction---even those with outstanding references from clients. This was
ineffective in reducing resource consumption because there was an
outstanding reference, guaranteeing that the items remained. A secondary
issue was that there is no guarantee that these in-use items will be the
last things reached in the LRU chain, which actually recorded
"least-recently requested" rather than "least-recently used".
2. The sum of usages was decremented not when a (key,value) was evicted from
the cache, but when its reference count went to zero. Thus, when things
were removed from the cache, either by garbage collection or via Erase(),
the usage sum was not necessarily decreased. This allowed the cache to act
as though full when it was in fact not, reducing caching effectiveness, and
leading to more resources being consumed---the opposite of what the
evictions were intended to achieve.
3. (minor) The cache's clients insert items into it by first looking up the
key, and inserting only if no value is found. Although the cache has an
internal lock, the clients use no locking to ensure atomicity of the
Lookup/Insert pair. (see table/table.cc: block_cache->Insert() and
db/table_cache.cc: cache_->Insert()). Thus, if two threads Insert() at
about the same time, they can both Lookup(), find nothing, and both
Insert(). The second Insert() would evict the first value, leaving each
thread with a handle on its own version of the data, and with the second
version in the cache. It would be better if both threads ended up with a
handle on the same (key,value) pair, which implies it must be the first item
inserted. This suggests that Insert() should not replace an existing value.
This can be made safe with current usage inside LeveDB itself, but this is
not easy to change first because Cache is a public interface, so to change
the semantics of an existing call might break things, second because Cache
is an abstract virtual class, so adding a new abstract virtual method may
break other implementations, and third, the new method "insert without
replacing" cannot be implemented in terms of the existing methods, so cannot
be implemented with a non-abstract default. But fortunately, the effects
of this issue are minor, so this issue is not fixed by this change.
The changes:
The assumption in the fixes is that it is always better to cache
entries unless removal from the cache would lead to deallocation.
Cache entries now have an "in_cache" boolean indicating whether
the cache has a reference on the entry. The only ways that this can
become false without the entry being passed to its "deleter" are via
Erase(), via Insert() when an element with a duplicate key is inserted,
or on destruction of the cache.
The cache now keeps two linked lists instead of one. All items
in the cache are in one list or the other, and never both. Items
still referenced by clients but erased from the cache are in
neither list. The lists are:
- in-use: contains the items currently referenced by clients, in no particular
order. (This list is used for invariant checking. If we removed the check,
elements that would otherwise be on this list could be left as disconnected
singleton lists.)
- LRU: contains the items not currently referenced by clients, in LRU order
A new internal Ref() method increments the reference count. If
incrementing from 1 to 2 for an item in the cache, it is moved
from the LRU list to the in-use list.
The Unref() call now moves things from the in-use list to the LRU
list if the reference count falls to 1, and the item is in the
cache. It no longer adjusts the usage sum. The usage sum now
reflects only what is in the cache, rather than including
still-referenced items that have been evicted.
The LRU_Append() now takes a "list" parameter so that it can be
used to append either to the LRU list or the in-use list.
Lookup() is modified to use the new Ref() call, rather than
adjusting the reference count and LRU chain directly.
Insert() eviction code is also modified to adjust the usage sum and the
in_cache boolean of the evicted elements. Some LevelDB tests assume that there
will be no caching whatsoever if the cache size is set to zero, so this is
handled as a special case.
A new private method FinishErase() is factored out
with the common code from where items are removed from the cache.
Erase() is modified to adjust the usage sum and the in_cache
boolean of the erased elements, and to use FinishErase().
Prune() is modified to use FinishErase() also, and to make use of the fact that
the lru_ list now contains only items with reference count 1.
- EvictionPolicy is modified to test that an entry with an
outstanding handle is not evicted. This test fails with the old cache.cc.
- A new test case UseExceedsCacheSize verifies that even when the
cache is overfull of entries with outstanding handles, none are
evicted. This test fails with the old cache.cc, and is the key
issue that causes file descriptors to run out when the cache
size is set too small.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=123247237
2016-05-26 04:32:03 +08:00
|
|
|
Cache::Handle* InsertAndReturnHandle(int key, int value, int charge = 1) {
|
|
|
|
return cache_->Insert(EncodeKey(key), EncodeValue(value), charge,
|
|
|
|
&CacheTest::Deleter);
|
|
|
|
}
|
|
|
|
|
2011-03-19 06:37:00 +08:00
|
|
|
void Erase(int key) {
|
|
|
|
cache_->Erase(EncodeKey(key));
|
|
|
|
}
|
|
|
|
};
|
|
|
|
CacheTest* CacheTest::current_;
|
|
|
|
|
|
|
|
TEST(CacheTest, HitAndMiss) {
|
|
|
|
ASSERT_EQ(-1, Lookup(100));
|
|
|
|
|
|
|
|
Insert(100, 101);
|
|
|
|
ASSERT_EQ(101, Lookup(100));
|
|
|
|
ASSERT_EQ(-1, Lookup(200));
|
|
|
|
ASSERT_EQ(-1, Lookup(300));
|
|
|
|
|
|
|
|
Insert(200, 201);
|
|
|
|
ASSERT_EQ(101, Lookup(100));
|
|
|
|
ASSERT_EQ(201, Lookup(200));
|
|
|
|
ASSERT_EQ(-1, Lookup(300));
|
|
|
|
|
|
|
|
Insert(100, 102);
|
|
|
|
ASSERT_EQ(102, Lookup(100));
|
|
|
|
ASSERT_EQ(201, Lookup(200));
|
|
|
|
ASSERT_EQ(-1, Lookup(300));
|
|
|
|
|
|
|
|
ASSERT_EQ(1, deleted_keys_.size());
|
|
|
|
ASSERT_EQ(100, deleted_keys_[0]);
|
|
|
|
ASSERT_EQ(101, deleted_values_[0]);
|
|
|
|
}
|
|
|
|
|
|
|
|
TEST(CacheTest, Erase) {
|
|
|
|
Erase(200);
|
|
|
|
ASSERT_EQ(0, deleted_keys_.size());
|
|
|
|
|
|
|
|
Insert(100, 101);
|
|
|
|
Insert(200, 201);
|
|
|
|
Erase(100);
|
|
|
|
ASSERT_EQ(-1, Lookup(100));
|
|
|
|
ASSERT_EQ(201, Lookup(200));
|
|
|
|
ASSERT_EQ(1, deleted_keys_.size());
|
|
|
|
ASSERT_EQ(100, deleted_keys_[0]);
|
|
|
|
ASSERT_EQ(101, deleted_values_[0]);
|
|
|
|
|
|
|
|
Erase(100);
|
|
|
|
ASSERT_EQ(-1, Lookup(100));
|
|
|
|
ASSERT_EQ(201, Lookup(200));
|
|
|
|
ASSERT_EQ(1, deleted_keys_.size());
|
|
|
|
}
|
|
|
|
|
|
|
|
TEST(CacheTest, EntriesArePinned) {
|
|
|
|
Insert(100, 101);
|
|
|
|
Cache::Handle* h1 = cache_->Lookup(EncodeKey(100));
|
|
|
|
ASSERT_EQ(101, DecodeValue(cache_->Value(h1)));
|
|
|
|
|
|
|
|
Insert(100, 102);
|
|
|
|
Cache::Handle* h2 = cache_->Lookup(EncodeKey(100));
|
|
|
|
ASSERT_EQ(102, DecodeValue(cache_->Value(h2)));
|
|
|
|
ASSERT_EQ(0, deleted_keys_.size());
|
|
|
|
|
|
|
|
cache_->Release(h1);
|
|
|
|
ASSERT_EQ(1, deleted_keys_.size());
|
|
|
|
ASSERT_EQ(100, deleted_keys_[0]);
|
|
|
|
ASSERT_EQ(101, deleted_values_[0]);
|
|
|
|
|
|
|
|
Erase(100);
|
|
|
|
ASSERT_EQ(-1, Lookup(100));
|
|
|
|
ASSERT_EQ(1, deleted_keys_.size());
|
|
|
|
|
|
|
|
cache_->Release(h2);
|
|
|
|
ASSERT_EQ(2, deleted_keys_.size());
|
|
|
|
ASSERT_EQ(100, deleted_keys_[1]);
|
|
|
|
ASSERT_EQ(102, deleted_values_[1]);
|
|
|
|
}
|
|
|
|
|
|
|
|
TEST(CacheTest, EvictionPolicy) {
|
|
|
|
Insert(100, 101);
|
|
|
|
Insert(200, 201);
|
fix problems in LevelDB's caching code
Background:
LevelDB uses a cache (util/cache.h, util/cache.cc) of (key,value)
pairs for two purposes:
- a cache of (table, file handle) pairs
- a cache of blocks
The cache places the (key,value) pairs in a reference-counted
wrapper. When it returns a value, it returns a reference to this
wrapper. When the client has finished using the reference and
its enclosed (key,value), it calls Release() to decrement the
reference count.
Each (key,value) pair has an associated resource usage. The
cache maintains the sum of the usages of the elements it holds,
and removes values as needed to keep the sum below a capacity
threshold. It maintains an LRU list so that it will remove the
least-recently used elements first.
The max_open_files option to LevelDB sets the size of the cache
of (table, file handle) pairs. The option is not used in any
other way.
The observed behaviour:
If LevelDB at any time used more file handles concurrently than
the cache size set via max_open_files, it attempted to reduce the
number by evicting entries from the table cache. This could
happen most easily during compaction, and if max_open_files was
low. Because the handles were in use, their reference count did
not drop to zero, and so the usage sum in the cache was not
modified by the evictions. Subsequent Insert() calls returned
valid handles, but their entries were immediately evicted from
the cache, which though empty still acted as though full. As a
result, there was effectively no caching, and the number of open
file handles rose []ly until it hit system-imposed limits and
the process died.
If one set max_open_files lower, the cache was more likely to
exhibit this beahviour, and cause the process to run out of file
descriptors. That is, max_open_files acted in almost exactly the
opposite manner from what was intended.
The problems:
1. The cache kept all elements on its LRU list eligible for capacity
eviction---even those with outstanding references from clients. This was
ineffective in reducing resource consumption because there was an
outstanding reference, guaranteeing that the items remained. A secondary
issue was that there is no guarantee that these in-use items will be the
last things reached in the LRU chain, which actually recorded
"least-recently requested" rather than "least-recently used".
2. The sum of usages was decremented not when a (key,value) was evicted from
the cache, but when its reference count went to zero. Thus, when things
were removed from the cache, either by garbage collection or via Erase(),
the usage sum was not necessarily decreased. This allowed the cache to act
as though full when it was in fact not, reducing caching effectiveness, and
leading to more resources being consumed---the opposite of what the
evictions were intended to achieve.
3. (minor) The cache's clients insert items into it by first looking up the
key, and inserting only if no value is found. Although the cache has an
internal lock, the clients use no locking to ensure atomicity of the
Lookup/Insert pair. (see table/table.cc: block_cache->Insert() and
db/table_cache.cc: cache_->Insert()). Thus, if two threads Insert() at
about the same time, they can both Lookup(), find nothing, and both
Insert(). The second Insert() would evict the first value, leaving each
thread with a handle on its own version of the data, and with the second
version in the cache. It would be better if both threads ended up with a
handle on the same (key,value) pair, which implies it must be the first item
inserted. This suggests that Insert() should not replace an existing value.
This can be made safe with current usage inside LeveDB itself, but this is
not easy to change first because Cache is a public interface, so to change
the semantics of an existing call might break things, second because Cache
is an abstract virtual class, so adding a new abstract virtual method may
break other implementations, and third, the new method "insert without
replacing" cannot be implemented in terms of the existing methods, so cannot
be implemented with a non-abstract default. But fortunately, the effects
of this issue are minor, so this issue is not fixed by this change.
The changes:
The assumption in the fixes is that it is always better to cache
entries unless removal from the cache would lead to deallocation.
Cache entries now have an "in_cache" boolean indicating whether
the cache has a reference on the entry. The only ways that this can
become false without the entry being passed to its "deleter" are via
Erase(), via Insert() when an element with a duplicate key is inserted,
or on destruction of the cache.
The cache now keeps two linked lists instead of one. All items
in the cache are in one list or the other, and never both. Items
still referenced by clients but erased from the cache are in
neither list. The lists are:
- in-use: contains the items currently referenced by clients, in no particular
order. (This list is used for invariant checking. If we removed the check,
elements that would otherwise be on this list could be left as disconnected
singleton lists.)
- LRU: contains the items not currently referenced by clients, in LRU order
A new internal Ref() method increments the reference count. If
incrementing from 1 to 2 for an item in the cache, it is moved
from the LRU list to the in-use list.
The Unref() call now moves things from the in-use list to the LRU
list if the reference count falls to 1, and the item is in the
cache. It no longer adjusts the usage sum. The usage sum now
reflects only what is in the cache, rather than including
still-referenced items that have been evicted.
The LRU_Append() now takes a "list" parameter so that it can be
used to append either to the LRU list or the in-use list.
Lookup() is modified to use the new Ref() call, rather than
adjusting the reference count and LRU chain directly.
Insert() eviction code is also modified to adjust the usage sum and the
in_cache boolean of the evicted elements. Some LevelDB tests assume that there
will be no caching whatsoever if the cache size is set to zero, so this is
handled as a special case.
A new private method FinishErase() is factored out
with the common code from where items are removed from the cache.
Erase() is modified to adjust the usage sum and the in_cache
boolean of the erased elements, and to use FinishErase().
Prune() is modified to use FinishErase() also, and to make use of the fact that
the lru_ list now contains only items with reference count 1.
- EvictionPolicy is modified to test that an entry with an
outstanding handle is not evicted. This test fails with the old cache.cc.
- A new test case UseExceedsCacheSize verifies that even when the
cache is overfull of entries with outstanding handles, none are
evicted. This test fails with the old cache.cc, and is the key
issue that causes file descriptors to run out when the cache
size is set too small.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=123247237
2016-05-26 04:32:03 +08:00
|
|
|
Insert(300, 301);
|
|
|
|
Cache::Handle* h = cache_->Lookup(EncodeKey(300));
|
2011-03-19 06:37:00 +08:00
|
|
|
|
fix problems in LevelDB's caching code
Background:
LevelDB uses a cache (util/cache.h, util/cache.cc) of (key,value)
pairs for two purposes:
- a cache of (table, file handle) pairs
- a cache of blocks
The cache places the (key,value) pairs in a reference-counted
wrapper. When it returns a value, it returns a reference to this
wrapper. When the client has finished using the reference and
its enclosed (key,value), it calls Release() to decrement the
reference count.
Each (key,value) pair has an associated resource usage. The
cache maintains the sum of the usages of the elements it holds,
and removes values as needed to keep the sum below a capacity
threshold. It maintains an LRU list so that it will remove the
least-recently used elements first.
The max_open_files option to LevelDB sets the size of the cache
of (table, file handle) pairs. The option is not used in any
other way.
The observed behaviour:
If LevelDB at any time used more file handles concurrently than
the cache size set via max_open_files, it attempted to reduce the
number by evicting entries from the table cache. This could
happen most easily during compaction, and if max_open_files was
low. Because the handles were in use, their reference count did
not drop to zero, and so the usage sum in the cache was not
modified by the evictions. Subsequent Insert() calls returned
valid handles, but their entries were immediately evicted from
the cache, which though empty still acted as though full. As a
result, there was effectively no caching, and the number of open
file handles rose []ly until it hit system-imposed limits and
the process died.
If one set max_open_files lower, the cache was more likely to
exhibit this beahviour, and cause the process to run out of file
descriptors. That is, max_open_files acted in almost exactly the
opposite manner from what was intended.
The problems:
1. The cache kept all elements on its LRU list eligible for capacity
eviction---even those with outstanding references from clients. This was
ineffective in reducing resource consumption because there was an
outstanding reference, guaranteeing that the items remained. A secondary
issue was that there is no guarantee that these in-use items will be the
last things reached in the LRU chain, which actually recorded
"least-recently requested" rather than "least-recently used".
2. The sum of usages was decremented not when a (key,value) was evicted from
the cache, but when its reference count went to zero. Thus, when things
were removed from the cache, either by garbage collection or via Erase(),
the usage sum was not necessarily decreased. This allowed the cache to act
as though full when it was in fact not, reducing caching effectiveness, and
leading to more resources being consumed---the opposite of what the
evictions were intended to achieve.
3. (minor) The cache's clients insert items into it by first looking up the
key, and inserting only if no value is found. Although the cache has an
internal lock, the clients use no locking to ensure atomicity of the
Lookup/Insert pair. (see table/table.cc: block_cache->Insert() and
db/table_cache.cc: cache_->Insert()). Thus, if two threads Insert() at
about the same time, they can both Lookup(), find nothing, and both
Insert(). The second Insert() would evict the first value, leaving each
thread with a handle on its own version of the data, and with the second
version in the cache. It would be better if both threads ended up with a
handle on the same (key,value) pair, which implies it must be the first item
inserted. This suggests that Insert() should not replace an existing value.
This can be made safe with current usage inside LeveDB itself, but this is
not easy to change first because Cache is a public interface, so to change
the semantics of an existing call might break things, second because Cache
is an abstract virtual class, so adding a new abstract virtual method may
break other implementations, and third, the new method "insert without
replacing" cannot be implemented in terms of the existing methods, so cannot
be implemented with a non-abstract default. But fortunately, the effects
of this issue are minor, so this issue is not fixed by this change.
The changes:
The assumption in the fixes is that it is always better to cache
entries unless removal from the cache would lead to deallocation.
Cache entries now have an "in_cache" boolean indicating whether
the cache has a reference on the entry. The only ways that this can
become false without the entry being passed to its "deleter" are via
Erase(), via Insert() when an element with a duplicate key is inserted,
or on destruction of the cache.
The cache now keeps two linked lists instead of one. All items
in the cache are in one list or the other, and never both. Items
still referenced by clients but erased from the cache are in
neither list. The lists are:
- in-use: contains the items currently referenced by clients, in no particular
order. (This list is used for invariant checking. If we removed the check,
elements that would otherwise be on this list could be left as disconnected
singleton lists.)
- LRU: contains the items not currently referenced by clients, in LRU order
A new internal Ref() method increments the reference count. If
incrementing from 1 to 2 for an item in the cache, it is moved
from the LRU list to the in-use list.
The Unref() call now moves things from the in-use list to the LRU
list if the reference count falls to 1, and the item is in the
cache. It no longer adjusts the usage sum. The usage sum now
reflects only what is in the cache, rather than including
still-referenced items that have been evicted.
The LRU_Append() now takes a "list" parameter so that it can be
used to append either to the LRU list or the in-use list.
Lookup() is modified to use the new Ref() call, rather than
adjusting the reference count and LRU chain directly.
Insert() eviction code is also modified to adjust the usage sum and the
in_cache boolean of the evicted elements. Some LevelDB tests assume that there
will be no caching whatsoever if the cache size is set to zero, so this is
handled as a special case.
A new private method FinishErase() is factored out
with the common code from where items are removed from the cache.
Erase() is modified to adjust the usage sum and the in_cache
boolean of the erased elements, and to use FinishErase().
Prune() is modified to use FinishErase() also, and to make use of the fact that
the lru_ list now contains only items with reference count 1.
- EvictionPolicy is modified to test that an entry with an
outstanding handle is not evicted. This test fails with the old cache.cc.
- A new test case UseExceedsCacheSize verifies that even when the
cache is overfull of entries with outstanding handles, none are
evicted. This test fails with the old cache.cc, and is the key
issue that causes file descriptors to run out when the cache
size is set too small.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=123247237
2016-05-26 04:32:03 +08:00
|
|
|
// Frequently used entry must be kept around,
|
|
|
|
// as must things that are still in use.
|
2011-08-23 05:08:51 +08:00
|
|
|
for (int i = 0; i < kCacheSize + 100; i++) {
|
2011-03-19 06:37:00 +08:00
|
|
|
Insert(1000+i, 2000+i);
|
|
|
|
ASSERT_EQ(2000+i, Lookup(1000+i));
|
|
|
|
ASSERT_EQ(101, Lookup(100));
|
|
|
|
}
|
|
|
|
ASSERT_EQ(101, Lookup(100));
|
2011-08-23 05:08:51 +08:00
|
|
|
ASSERT_EQ(-1, Lookup(200));
|
fix problems in LevelDB's caching code
Background:
LevelDB uses a cache (util/cache.h, util/cache.cc) of (key,value)
pairs for two purposes:
- a cache of (table, file handle) pairs
- a cache of blocks
The cache places the (key,value) pairs in a reference-counted
wrapper. When it returns a value, it returns a reference to this
wrapper. When the client has finished using the reference and
its enclosed (key,value), it calls Release() to decrement the
reference count.
Each (key,value) pair has an associated resource usage. The
cache maintains the sum of the usages of the elements it holds,
and removes values as needed to keep the sum below a capacity
threshold. It maintains an LRU list so that it will remove the
least-recently used elements first.
The max_open_files option to LevelDB sets the size of the cache
of (table, file handle) pairs. The option is not used in any
other way.
The observed behaviour:
If LevelDB at any time used more file handles concurrently than
the cache size set via max_open_files, it attempted to reduce the
number by evicting entries from the table cache. This could
happen most easily during compaction, and if max_open_files was
low. Because the handles were in use, their reference count did
not drop to zero, and so the usage sum in the cache was not
modified by the evictions. Subsequent Insert() calls returned
valid handles, but their entries were immediately evicted from
the cache, which though empty still acted as though full. As a
result, there was effectively no caching, and the number of open
file handles rose []ly until it hit system-imposed limits and
the process died.
If one set max_open_files lower, the cache was more likely to
exhibit this beahviour, and cause the process to run out of file
descriptors. That is, max_open_files acted in almost exactly the
opposite manner from what was intended.
The problems:
1. The cache kept all elements on its LRU list eligible for capacity
eviction---even those with outstanding references from clients. This was
ineffective in reducing resource consumption because there was an
outstanding reference, guaranteeing that the items remained. A secondary
issue was that there is no guarantee that these in-use items will be the
last things reached in the LRU chain, which actually recorded
"least-recently requested" rather than "least-recently used".
2. The sum of usages was decremented not when a (key,value) was evicted from
the cache, but when its reference count went to zero. Thus, when things
were removed from the cache, either by garbage collection or via Erase(),
the usage sum was not necessarily decreased. This allowed the cache to act
as though full when it was in fact not, reducing caching effectiveness, and
leading to more resources being consumed---the opposite of what the
evictions were intended to achieve.
3. (minor) The cache's clients insert items into it by first looking up the
key, and inserting only if no value is found. Although the cache has an
internal lock, the clients use no locking to ensure atomicity of the
Lookup/Insert pair. (see table/table.cc: block_cache->Insert() and
db/table_cache.cc: cache_->Insert()). Thus, if two threads Insert() at
about the same time, they can both Lookup(), find nothing, and both
Insert(). The second Insert() would evict the first value, leaving each
thread with a handle on its own version of the data, and with the second
version in the cache. It would be better if both threads ended up with a
handle on the same (key,value) pair, which implies it must be the first item
inserted. This suggests that Insert() should not replace an existing value.
This can be made safe with current usage inside LeveDB itself, but this is
not easy to change first because Cache is a public interface, so to change
the semantics of an existing call might break things, second because Cache
is an abstract virtual class, so adding a new abstract virtual method may
break other implementations, and third, the new method "insert without
replacing" cannot be implemented in terms of the existing methods, so cannot
be implemented with a non-abstract default. But fortunately, the effects
of this issue are minor, so this issue is not fixed by this change.
The changes:
The assumption in the fixes is that it is always better to cache
entries unless removal from the cache would lead to deallocation.
Cache entries now have an "in_cache" boolean indicating whether
the cache has a reference on the entry. The only ways that this can
become false without the entry being passed to its "deleter" are via
Erase(), via Insert() when an element with a duplicate key is inserted,
or on destruction of the cache.
The cache now keeps two linked lists instead of one. All items
in the cache are in one list or the other, and never both. Items
still referenced by clients but erased from the cache are in
neither list. The lists are:
- in-use: contains the items currently referenced by clients, in no particular
order. (This list is used for invariant checking. If we removed the check,
elements that would otherwise be on this list could be left as disconnected
singleton lists.)
- LRU: contains the items not currently referenced by clients, in LRU order
A new internal Ref() method increments the reference count. If
incrementing from 1 to 2 for an item in the cache, it is moved
from the LRU list to the in-use list.
The Unref() call now moves things from the in-use list to the LRU
list if the reference count falls to 1, and the item is in the
cache. It no longer adjusts the usage sum. The usage sum now
reflects only what is in the cache, rather than including
still-referenced items that have been evicted.
The LRU_Append() now takes a "list" parameter so that it can be
used to append either to the LRU list or the in-use list.
Lookup() is modified to use the new Ref() call, rather than
adjusting the reference count and LRU chain directly.
Insert() eviction code is also modified to adjust the usage sum and the
in_cache boolean of the evicted elements. Some LevelDB tests assume that there
will be no caching whatsoever if the cache size is set to zero, so this is
handled as a special case.
A new private method FinishErase() is factored out
with the common code from where items are removed from the cache.
Erase() is modified to adjust the usage sum and the in_cache
boolean of the erased elements, and to use FinishErase().
Prune() is modified to use FinishErase() also, and to make use of the fact that
the lru_ list now contains only items with reference count 1.
- EvictionPolicy is modified to test that an entry with an
outstanding handle is not evicted. This test fails with the old cache.cc.
- A new test case UseExceedsCacheSize verifies that even when the
cache is overfull of entries with outstanding handles, none are
evicted. This test fails with the old cache.cc, and is the key
issue that causes file descriptors to run out when the cache
size is set too small.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=123247237
2016-05-26 04:32:03 +08:00
|
|
|
ASSERT_EQ(301, Lookup(300));
|
|
|
|
cache_->Release(h);
|
|
|
|
}
|
|
|
|
|
|
|
|
TEST(CacheTest, UseExceedsCacheSize) {
|
|
|
|
// Overfill the cache, keeping handles on all inserted entries.
|
|
|
|
std::vector<Cache::Handle*> h;
|
|
|
|
for (int i = 0; i < kCacheSize + 100; i++) {
|
|
|
|
h.push_back(InsertAndReturnHandle(1000+i, 2000+i));
|
|
|
|
}
|
|
|
|
|
|
|
|
// Check that all the entries can be found in the cache.
|
|
|
|
for (int i = 0; i < h.size(); i++) {
|
|
|
|
ASSERT_EQ(2000+i, Lookup(1000+i));
|
|
|
|
}
|
|
|
|
|
|
|
|
for (int i = 0; i < h.size(); i++) {
|
|
|
|
cache_->Release(h[i]);
|
|
|
|
}
|
2011-03-19 06:37:00 +08:00
|
|
|
}
|
|
|
|
|
2011-08-23 05:08:51 +08:00
|
|
|
TEST(CacheTest, HeavyEntries) {
|
|
|
|
// Add a bunch of light and heavy entries and then count the combined
|
|
|
|
// size of items still in the cache, which must be approximately the
|
|
|
|
// same as the total capacity.
|
|
|
|
const int kLight = 1;
|
|
|
|
const int kHeavy = 10;
|
|
|
|
int added = 0;
|
|
|
|
int index = 0;
|
|
|
|
while (added < 2*kCacheSize) {
|
|
|
|
const int weight = (index & 1) ? kLight : kHeavy;
|
|
|
|
Insert(index, 1000+index, weight);
|
|
|
|
added += weight;
|
|
|
|
index++;
|
|
|
|
}
|
|
|
|
|
|
|
|
int cached_weight = 0;
|
|
|
|
for (int i = 0; i < index; i++) {
|
|
|
|
const int weight = (i & 1 ? kLight : kHeavy);
|
|
|
|
int r = Lookup(i);
|
|
|
|
if (r >= 0) {
|
|
|
|
cached_weight += weight;
|
|
|
|
ASSERT_EQ(1000+i, r);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
ASSERT_LE(cached_weight, kCacheSize + kCacheSize/10);
|
2011-03-19 06:37:00 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
TEST(CacheTest, NewId) {
|
|
|
|
uint64_t a = cache_->NewId();
|
|
|
|
uint64_t b = cache_->NewId();
|
|
|
|
ASSERT_NE(a, b);
|
|
|
|
}
|
|
|
|
|
2015-08-24 11:35:01 +08:00
|
|
|
TEST(CacheTest, Prune) {
|
|
|
|
Insert(1, 100);
|
|
|
|
Insert(2, 200);
|
|
|
|
|
|
|
|
Cache::Handle* handle = cache_->Lookup(EncodeKey(1));
|
|
|
|
ASSERT_TRUE(handle);
|
|
|
|
cache_->Prune();
|
|
|
|
cache_->Release(handle);
|
|
|
|
|
|
|
|
ASSERT_EQ(100, Lookup(1));
|
|
|
|
ASSERT_EQ(-1, Lookup(2));
|
|
|
|
}
|
|
|
|
|
2017-10-03 04:57:41 +08:00
|
|
|
TEST(CacheTest, ZeroSizeCache) {
|
|
|
|
delete cache_;
|
|
|
|
cache_ = NewLRUCache(0);
|
|
|
|
|
|
|
|
Insert(1, 100);
|
|
|
|
ASSERT_EQ(-1, Lookup(1));
|
|
|
|
}
|
|
|
|
|
2011-11-01 01:22:06 +08:00
|
|
|
} // namespace leveldb
|
2011-03-19 06:37:00 +08:00
|
|
|
|
|
|
|
int main(int argc, char** argv) {
|
|
|
|
return leveldb::test::RunAllTests();
|
|
|
|
}
|