Summary:
`enable_custom_split_merge` is added for enabling the custom split and merge feature, which split the compressed value into chunks so that they may better fit jemalloc bins.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10690
Test Plan:
Unit Tests
Stress Tests
Reviewed By: anand1976
Differential Revision: D39567604
Pulled By: gitbw95
fbshipit-source-id: f6d1d46200f365220055f793514601dcb0edc4b7
Summary:
This change establishes a distinctive name for the experimental new lock-free clock cache (originally developed by guidotag and revamped in PR https://github.com/facebook/rocksdb/issues/10626). A few reasons:
* We want to make it clear that this is a fundamentally different implementation vs. the old clock cache, to avoid people saying "I already tried clock cache."
* We want to highlight the key feature: it's fast (especially under parallel load)
* Because it requires an estimated charge per entry, it is not drop-in API compatible with old clock cache. This estimate might always be required for highest performance, and giving it a distinct name should reduce confusion about the distinct API requirements.
* We might develop a variant requiring the same estimate parameter but with LRU eviction. In that case, using the name HyperLRUCache should make things more clear. (FastLRUCache is just a prototype that might soon be removed.)
Some API detail:
* To reduce copy-pasting parameter lists, etc. as in LRUCache construction, I have a `MakeSharedCache()` function on `HyperClockCacheOptions` instead of `NewHyperClockCache()`.
* Changes -cache_type=clock_cache to -cache_type=hyper_clock_cache for applicable tools. I think this is more consistent / sustainable for reasons already stated.
For performance tests see https://github.com/facebook/rocksdb/pull/10626
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10684
Test Plan: no interesting functional changes; tests updated
Reviewed By: anand1976
Differential Revision: D39547800
Pulled By: pdillinger
fbshipit-source-id: 5c0fe1b5cf3cb680ab369b928c8569682b9795bf
Summary:
* Consolidates most metadata into a single word per slot so that more
can be accomplished with a single atomic update. In the common case,
Lookup was previously about 4 atomic updates, now just 1 atomic update.
Common case Release was previously 1 atomic read + 1 atomic update,
now just 1 atomic update.
* Eliminate spins / waits / yields, which likely threaten some "lock free"
benefits. Compare-exchange loops are only used in explicit Erase, and
strict_capacity_limit=true Insert. Eviction uses opportunistic compare-
exchange.
* Relaxes some aggressiveness and guarantees. For example,
* Duplicate Inserts will sometimes go undetected and the shadow duplicate
will age out with eviction.
* In many cases, the older Inserted value for a given cache key will be kept
(i.e. Insert does not support overwrite).
* Entries explicitly erased (rather than evicted) might not be freed
immediately in some rare cases.
* With strict_capacity_limit=false, capacity limit is not tracked/enforced as
precisely as LRUCache, but is self-correcting and should only deviate by a
very small number of extra or fewer entries.
* Use smaller "computed default" number of cache shards in many cases,
because benefits to larger usage tracking / eviction pools outweigh the small
cost of more lock-free atomic contention. The improvement in CPU and I/O
is dramatic in some limit-memory cases.
* Even without the sharding change, the eviction algorithm is likely more
effective than LRU overall because it's more stateful, even though the
"hot path" state tracking for it is essentially free with ref counting. It
is like a generalized CLOCK with aging (see code comments). I don't have
performance numbers showing a specific improvement, but in theory, for a
Poisson access pattern to each block, keeping some state allows better
estimation of time to next access (Poisson interval) than strict LRU. The
bounded randomness in CLOCK can also reduce "cliff" effect for repeated
range scans approaching and exceeding cache size.
## Hot path algorithm comparison
Rough descriptions, focusing on number and kind of atomic operations:
* Old `Lookup()` (2-5 atomic updates per probe):
```
Loop:
Increment internal ref count at slot
If possible hit:
Check flags atomic (and non-atomic fields)
If cache hit:
Three distinct updates to 'flags' atomic
Increment refs for internal-to-external
Return
Decrement internal ref count
while atomic read 'displacements' > 0
```
* New `Lookup()` (1-2 atomic updates per probe):
```
Loop:
Increment acquire counter in meta word (optimistic)
If visible entry (already read meta word):
If match (read non-atomic fields):
Return
Else:
Decrement acquire counter in meta word
Else if invisible entry (rare, already read meta word):
Decrement acquire counter in meta word
while atomic read 'displacements' > 0
```
* Old `Release()` (1 atomic update, conditional on atomic read, rarely more):
```
Read atomic ref count
If last reference and invisible (rare):
Use CAS etc. to remove
Return
Else:
Decrement ref count
```
* New `Release()` (1 unconditional atomic update, rarely more):
```
Increment release counter in meta word
If last reference and invisible (rare):
Use CAS etc. to remove
Return
```
## Performance test setup
Build DB with
```
TEST_TMPDIR=/dev/shm ./db_bench -benchmarks=fillrandom -num=30000000 -disable_wal=1 -bloom_bits=16
```
Test with
```
TEST_TMPDIR=/dev/shm ./db_bench -benchmarks=readrandom -readonly -num=30000000 -bloom_bits=16 -cache_index_and_filter_blocks=1 -cache_size=${CACHE_MB}000000 -duration 60 -threads=$THREADS -statistics
```
Numbers on a single socket Skylake Xeon system with 48 hardware threads, DEBUG_LEVEL=0 PORTABLE=0. Very similar story on a dual socket system with 80 hardware threads. Using (every 2nd) Fibonacci MB cache sizes to sample the territory between powers of two. Configurations:
base: LRUCache before this change, but with db_bench change to default cache_numshardbits=-1 (instead of fixed at 6)
folly: LRUCache before this change, with folly enabled (distributed mutex) but on an old compiler (sorry)
gt_clock: experimental ClockCache before this change
new_clock: experimental ClockCache with this change
## Performance test results
First test "hot path" read performance, with block cache large enough for whole DB:
4181MB 1thread base -> kops/s: 47.761
4181MB 1thread folly -> kops/s: 45.877
4181MB 1thread gt_clock -> kops/s: 51.092
4181MB 1thread new_clock -> kops/s: 53.944
4181MB 16thread base -> kops/s: 284.567
4181MB 16thread folly -> kops/s: 249.015
4181MB 16thread gt_clock -> kops/s: 743.762
4181MB 16thread new_clock -> kops/s: 861.821
4181MB 24thread base -> kops/s: 303.415
4181MB 24thread folly -> kops/s: 266.548
4181MB 24thread gt_clock -> kops/s: 975.706
4181MB 24thread new_clock -> kops/s: 1205.64 (~= 24 * 53.944)
4181MB 32thread base -> kops/s: 311.251
4181MB 32thread folly -> kops/s: 274.952
4181MB 32thread gt_clock -> kops/s: 1045.98
4181MB 32thread new_clock -> kops/s: 1370.38
4181MB 48thread base -> kops/s: 310.504
4181MB 48thread folly -> kops/s: 268.322
4181MB 48thread gt_clock -> kops/s: 1195.65
4181MB 48thread new_clock -> kops/s: 1604.85 (~= 24 * 1.25 * 53.944)
4181MB 64thread base -> kops/s: 307.839
4181MB 64thread folly -> kops/s: 272.172
4181MB 64thread gt_clock -> kops/s: 1204.47
4181MB 64thread new_clock -> kops/s: 1615.37
4181MB 128thread base -> kops/s: 310.934
4181MB 128thread folly -> kops/s: 267.468
4181MB 128thread gt_clock -> kops/s: 1188.75
4181MB 128thread new_clock -> kops/s: 1595.46
Whether we have just one thread on a quiet system or an overload of threads, the new version wins every time in thousand-ops per second, sometimes dramatically so. Mutex-based implementation quickly becomes contention-limited. New clock cache shows essentially perfect scaling up to number of physical cores (24), and then each hyperthreaded core adding about 1/4 the throughput of an additional physical core (see 48 thread case). Block cache miss rates (omitted above) are negligible across the board. With partitioned instead of full filters, the maximum speed-up vs. base is more like 2.5x rather than 5x.
Now test a large block cache with low miss ratio, but some eviction is required:
1597MB 1thread base -> kops/s: 46.603 io_bytes/op: 1584.63 miss_ratio: 0.0201066 max_rss_mb: 1589.23
1597MB 1thread folly -> kops/s: 45.079 io_bytes/op: 1530.03 miss_ratio: 0.019872 max_rss_mb: 1550.43
1597MB 1thread gt_clock -> kops/s: 48.711 io_bytes/op: 1566.63 miss_ratio: 0.0198923 max_rss_mb: 1691.4
1597MB 1thread new_clock -> kops/s: 51.531 io_bytes/op: 1589.07 miss_ratio: 0.0201969 max_rss_mb: 1583.56
1597MB 32thread base -> kops/s: 301.174 io_bytes/op: 1439.52 miss_ratio: 0.0184218 max_rss_mb: 1656.59
1597MB 32thread folly -> kops/s: 273.09 io_bytes/op: 1375.12 miss_ratio: 0.0180002 max_rss_mb: 1586.8
1597MB 32thread gt_clock -> kops/s: 904.497 io_bytes/op: 1411.29 miss_ratio: 0.0179934 max_rss_mb: 1775.89
1597MB 32thread new_clock -> kops/s: 1182.59 io_bytes/op: 1440.77 miss_ratio: 0.0185449 max_rss_mb: 1636.45
1597MB 128thread base -> kops/s: 309.91 io_bytes/op: 1438.25 miss_ratio: 0.018399 max_rss_mb: 1689.98
1597MB 128thread folly -> kops/s: 267.605 io_bytes/op: 1394.16 miss_ratio: 0.0180286 max_rss_mb: 1631.91
1597MB 128thread gt_clock -> kops/s: 691.518 io_bytes/op: 9056.73 miss_ratio: 0.0186572 max_rss_mb: 1982.26
1597MB 128thread new_clock -> kops/s: 1406.12 io_bytes/op: 1440.82 miss_ratio: 0.0185463 max_rss_mb: 1685.63
610MB 1thread base -> kops/s: 45.511 io_bytes/op: 2279.61 miss_ratio: 0.0290528 max_rss_mb: 615.137
610MB 1thread folly -> kops/s: 43.386 io_bytes/op: 2217.29 miss_ratio: 0.0289282 max_rss_mb: 600.996
610MB 1thread gt_clock -> kops/s: 46.207 io_bytes/op: 2275.51 miss_ratio: 0.0290057 max_rss_mb: 637.934
610MB 1thread new_clock -> kops/s: 48.879 io_bytes/op: 2283.1 miss_ratio: 0.0291253 max_rss_mb: 613.5
610MB 32thread base -> kops/s: 306.59 io_bytes/op: 2250 miss_ratio: 0.0288721 max_rss_mb: 683.402
610MB 32thread folly -> kops/s: 269.176 io_bytes/op: 2187.86 miss_ratio: 0.0286938 max_rss_mb: 628.742
610MB 32thread gt_clock -> kops/s: 855.097 io_bytes/op: 2279.26 miss_ratio: 0.0288009 max_rss_mb: 733.062
610MB 32thread new_clock -> kops/s: 1121.47 io_bytes/op: 2244.29 miss_ratio: 0.0289046 max_rss_mb: 666.453
610MB 128thread base -> kops/s: 305.079 io_bytes/op: 2252.43 miss_ratio: 0.0288884 max_rss_mb: 723.457
610MB 128thread folly -> kops/s: 269.583 io_bytes/op: 2204.58 miss_ratio: 0.0287001 max_rss_mb: 676.426
610MB 128thread gt_clock -> kops/s: 53.298 io_bytes/op: 8128.98 miss_ratio: 0.0292452 max_rss_mb: 956.273
610MB 128thread new_clock -> kops/s: 1301.09 io_bytes/op: 2246.04 miss_ratio: 0.0289171 max_rss_mb: 788.812
The new version is still winning every time, sometimes dramatically so, and we can tell from the maximum resident memory numbers (which contain some noise, by the way) that the new cache is not cheating on memory usage. IMPORTANT: The previous generation experimental clock cache appears to hit a serious bottleneck in the higher thread count configurations, presumably due to some of its waiting functionality. (The same bottleneck is not seen with partitioned index+filters.)
Now we consider even smaller cache sizes, with higher miss ratios, eviction work, etc.
233MB 1thread base -> kops/s: 10.557 io_bytes/op: 227040 miss_ratio: 0.0403105 max_rss_mb: 247.371
233MB 1thread folly -> kops/s: 15.348 io_bytes/op: 112007 miss_ratio: 0.0372238 max_rss_mb: 245.293
233MB 1thread gt_clock -> kops/s: 6.365 io_bytes/op: 244854 miss_ratio: 0.0413873 max_rss_mb: 259.844
233MB 1thread new_clock -> kops/s: 47.501 io_bytes/op: 2591.93 miss_ratio: 0.0330989 max_rss_mb: 242.461
233MB 32thread base -> kops/s: 96.498 io_bytes/op: 363379 miss_ratio: 0.0459966 max_rss_mb: 479.227
233MB 32thread folly -> kops/s: 109.95 io_bytes/op: 314799 miss_ratio: 0.0450032 max_rss_mb: 400.738
233MB 32thread gt_clock -> kops/s: 2.353 io_bytes/op: 385397 miss_ratio: 0.048445 max_rss_mb: 500.688
233MB 32thread new_clock -> kops/s: 1088.95 io_bytes/op: 2567.02 miss_ratio: 0.0330593 max_rss_mb: 303.402
233MB 128thread base -> kops/s: 84.302 io_bytes/op: 378020 miss_ratio: 0.0466558 max_rss_mb: 1051.84
233MB 128thread folly -> kops/s: 89.921 io_bytes/op: 338242 miss_ratio: 0.0460309 max_rss_mb: 812.785
233MB 128thread gt_clock -> kops/s: 2.588 io_bytes/op: 462833 miss_ratio: 0.0509158 max_rss_mb: 1109.94
233MB 128thread new_clock -> kops/s: 1299.26 io_bytes/op: 2565.94 miss_ratio: 0.0330531 max_rss_mb: 361.016
89MB 1thread base -> kops/s: 0.574 io_bytes/op: 5.35977e+06 miss_ratio: 0.274427 max_rss_mb: 91.3086
89MB 1thread folly -> kops/s: 0.578 io_bytes/op: 5.16549e+06 miss_ratio: 0.27276 max_rss_mb: 96.8984
89MB 1thread gt_clock -> kops/s: 0.512 io_bytes/op: 4.13111e+06 miss_ratio: 0.242817 max_rss_mb: 119.441
89MB 1thread new_clock -> kops/s: 48.172 io_bytes/op: 2709.76 miss_ratio: 0.0346162 max_rss_mb: 100.754
89MB 32thread base -> kops/s: 5.779 io_bytes/op: 6.14192e+06 miss_ratio: 0.320399 max_rss_mb: 311.812
89MB 32thread folly -> kops/s: 5.601 io_bytes/op: 5.83838e+06 miss_ratio: 0.313123 max_rss_mb: 252.418
89MB 32thread gt_clock -> kops/s: 0.77 io_bytes/op: 3.99236e+06 miss_ratio: 0.236296 max_rss_mb: 396.422
89MB 32thread new_clock -> kops/s: 1064.97 io_bytes/op: 2687.23 miss_ratio: 0.0346134 max_rss_mb: 155.293
89MB 128thread base -> kops/s: 4.959 io_bytes/op: 6.20297e+06 miss_ratio: 0.323945 max_rss_mb: 823.43
89MB 128thread folly -> kops/s: 4.962 io_bytes/op: 5.9601e+06 miss_ratio: 0.319857 max_rss_mb: 626.824
89MB 128thread gt_clock -> kops/s: 1.009 io_bytes/op: 4.1083e+06 miss_ratio: 0.242512 max_rss_mb: 1095.32
89MB 128thread new_clock -> kops/s: 1224.39 io_bytes/op: 2688.2 miss_ratio: 0.0346207 max_rss_mb: 218.223
^ Now something interesting has happened: the new clock cache has gained a dramatic lead in the single-threaded case, and this is because the cache is so small, and full filters are so big, that dividing the cache into 64 shards leads to significant (random) imbalances in cache shards and excessive churn in imbalanced shards. This new clock cache only uses two shards for this configuration, and that helps to ensure that entries are part of a sufficiently big pool that their eviction order resembles the single-shard order. (This effect is not seen with partitioned index+filters.)
Even smaller cache size:
34MB 1thread base -> kops/s: 0.198 io_bytes/op: 1.65342e+07 miss_ratio: 0.939466 max_rss_mb: 48.6914
34MB 1thread folly -> kops/s: 0.201 io_bytes/op: 1.63416e+07 miss_ratio: 0.939081 max_rss_mb: 45.3281
34MB 1thread gt_clock -> kops/s: 0.448 io_bytes/op: 4.43957e+06 miss_ratio: 0.266749 max_rss_mb: 100.523
34MB 1thread new_clock -> kops/s: 1.055 io_bytes/op: 1.85439e+06 miss_ratio: 0.107512 max_rss_mb: 75.3125
34MB 32thread base -> kops/s: 3.346 io_bytes/op: 1.64852e+07 miss_ratio: 0.93596 max_rss_mb: 180.48
34MB 32thread folly -> kops/s: 3.431 io_bytes/op: 1.62857e+07 miss_ratio: 0.935693 max_rss_mb: 137.531
34MB 32thread gt_clock -> kops/s: 1.47 io_bytes/op: 4.89704e+06 miss_ratio: 0.295081 max_rss_mb: 392.465
34MB 32thread new_clock -> kops/s: 8.19 io_bytes/op: 3.70456e+06 miss_ratio: 0.20826 max_rss_mb: 519.793
34MB 128thread base -> kops/s: 2.293 io_bytes/op: 1.64351e+07 miss_ratio: 0.931866 max_rss_mb: 449.484
34MB 128thread folly -> kops/s: 2.34 io_bytes/op: 1.6219e+07 miss_ratio: 0.932023 max_rss_mb: 396.457
34MB 128thread gt_clock -> kops/s: 1.798 io_bytes/op: 5.4241e+06 miss_ratio: 0.324881 max_rss_mb: 1104.41
34MB 128thread new_clock -> kops/s: 10.519 io_bytes/op: 2.39354e+06 miss_ratio: 0.136147 max_rss_mb: 1050.52
As the miss ratio gets higher (say, above 10%), the CPU time spent in eviction starts to erode the advantage of using fewer shards (13% miss rate much lower than 94%). LRU's O(1) eviction time can eventually pay off when there's enough block cache churn:
13MB 1thread base -> kops/s: 0.195 io_bytes/op: 1.65732e+07 miss_ratio: 0.946604 max_rss_mb: 45.6328
13MB 1thread folly -> kops/s: 0.197 io_bytes/op: 1.63793e+07 miss_ratio: 0.94661 max_rss_mb: 33.8633
13MB 1thread gt_clock -> kops/s: 0.519 io_bytes/op: 4.43316e+06 miss_ratio: 0.269379 max_rss_mb: 100.684
13MB 1thread new_clock -> kops/s: 0.176 io_bytes/op: 1.54148e+07 miss_ratio: 0.91545 max_rss_mb: 66.2383
13MB 32thread base -> kops/s: 3.266 io_bytes/op: 1.65544e+07 miss_ratio: 0.943386 max_rss_mb: 132.492
13MB 32thread folly -> kops/s: 3.396 io_bytes/op: 1.63142e+07 miss_ratio: 0.943243 max_rss_mb: 101.863
13MB 32thread gt_clock -> kops/s: 2.758 io_bytes/op: 5.13714e+06 miss_ratio: 0.310652 max_rss_mb: 396.121
13MB 32thread new_clock -> kops/s: 3.11 io_bytes/op: 1.23419e+07 miss_ratio: 0.708425 max_rss_mb: 321.758
13MB 128thread base -> kops/s: 2.31 io_bytes/op: 1.64823e+07 miss_ratio: 0.939543 max_rss_mb: 425.539
13MB 128thread folly -> kops/s: 2.339 io_bytes/op: 1.6242e+07 miss_ratio: 0.939966 max_rss_mb: 346.098
13MB 128thread gt_clock -> kops/s: 3.223 io_bytes/op: 5.76928e+06 miss_ratio: 0.345899 max_rss_mb: 1087.77
13MB 128thread new_clock -> kops/s: 2.984 io_bytes/op: 1.05341e+07 miss_ratio: 0.606198 max_rss_mb: 898.27
gt_clock is clearly blowing way past its memory budget for lower miss rates and best throughput. new_clock also seems to be exceeding budgets, and this warrants more investigation but is not the use case we are targeting with the new cache. With partitioned index+filter, the miss ratio is much better, and although still high enough that the eviction CPU time is definitely offsetting mutex contention:
13MB 1thread base -> kops/s: 16.326 io_bytes/op: 23743.9 miss_ratio: 0.205362 max_rss_mb: 65.2852
13MB 1thread folly -> kops/s: 15.574 io_bytes/op: 19415 miss_ratio: 0.184157 max_rss_mb: 56.3516
13MB 1thread gt_clock -> kops/s: 14.459 io_bytes/op: 22873 miss_ratio: 0.198355 max_rss_mb: 63.9688
13MB 1thread new_clock -> kops/s: 16.34 io_bytes/op: 24386.5 miss_ratio: 0.210512 max_rss_mb: 61.707
13MB 128thread base -> kops/s: 289.786 io_bytes/op: 23710.9 miss_ratio: 0.205056 max_rss_mb: 103.57
13MB 128thread folly -> kops/s: 185.282 io_bytes/op: 19433.1 miss_ratio: 0.184275 max_rss_mb: 116.219
13MB 128thread gt_clock -> kops/s: 354.451 io_bytes/op: 23150.6 miss_ratio: 0.200495 max_rss_mb: 102.871
13MB 128thread new_clock -> kops/s: 295.359 io_bytes/op: 24626.4 miss_ratio: 0.212452 max_rss_mb: 121.109
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10626
Test Plan: updated unit tests, stress/crash test runs including with TSAN, ASAN, UBSAN
Reviewed By: anand1976
Differential Revision: D39368406
Pulled By: pdillinger
fbshipit-source-id: 5afc44da4c656f8f751b44552bbf27bd3ca6fef9
Summary:
**Summary:**
When a block is firstly `Lookup` from the secondary cache, we just insert a dummy block in the primary cache (charging the actual size of the block) and don’t erase the block from the secondary cache. A standalone handle is returned from `Lookup`. Only if the block is hit again, we erase it from the secondary cache and add it into the primary cache.
When a block is firstly evicted from the primary cache to the secondary cache, we just insert a dummy block (size 0) in the secondary cache. When the block is evicted again, it is treated as a hot block and is inserted into the secondary cache.
**Implementation Details**
Add a new state of LRUHandle: The handle is never inserted into the LRUCache (both hash table and LRU list) and it doesn't experience the above three states. The entry can be freed when refs becomes 0. (refs >= 1 && in_cache == false && IS_STANDALONE == true)
The behaviors of `LRUCacheShard::Lookup()` are updated if the secondary_cache is CompressedSecondaryCache:
1. If a handle is found in primary cache:
1.1. If the handle's value is not nullptr, it is returned immediately.
1.2. If the handle's value is nullptr, this means the handle is a dummy one. For a dummy handle, if it was retrieved from secondary cache, it may still exist in secondary cache.
- 1.2.1. If no valid handle can be `Lookup` from secondary cache, return nullptr.
- 1.2.2. If the handle from secondary cache is valid, erase it from the secondary cache and add it into the primary cache.
2. If a handle is not found in primary cache:
2.1. If no valid handle can be `Lookup` from secondary cache, return nullptr.
2.2. If the handle from secondary cache is valid, insert a dummy block in the primary cache (charging the actual size of the block) and return a standalone handle.
The behaviors of `LRUCacheShard::Promote()` are updated as follows:
1. If `e->sec_handle` has value, one of the following steps can happen:
1.1. Insert a dummy handle and return a standalone handle to caller when `secondary_cache_` is `CompressedSecondaryCache` and e is a standalone handle.
1.2. Insert the item into the primary cache and return the handle to caller.
1.3. Exception handling.
3. If `e->sec_handle` has no value, mark the item as not in cache and charge the cache as its only metadata that'll shortly be released.
The behavior of `CompressedSecondaryCache::Insert()` is updated:
1. If a block is evicted from the primary cache for the first time, a dummy item is inserted.
4. If a dummy item is found for a block, the block is inserted into the secondary cache.
The behavior of `CompressedSecondaryCache:::Lookup()` is updated:
1. If a handle is not found or it is a dummy item, a nullptr is returned.
2. If `erase_handle` is true, the handle is erased.
The behaviors of `LRUCacheShard::Release()` are adjusted for the standalone handles.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10527
Test Plan:
1. stress tests.
5. unit tests.
6. CPU profiling for db_bench.
Reviewed By: siying
Differential Revision: D38747613
Pulled By: gitbw95
fbshipit-source-id: 74a1eba7e1957c9affb2bd2ae3e0194584fa6eca
Summary:
The patch adds a dedicated cache entry role for blob values and switches
to a registered deleter so that blobs show up as a separate bucket
(as opposed to "Misc") in the cache occupancy statistics, e.g.
```
Block cache entry stats(count,size,portion): DataBlock(133515,531.73 MB,13.6866%) BlobValue(1824855,3.10 GB,81.7071%) Misc(1,0.00 KB,0%)
```
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10601
Test Plan: Ran `make check` and tested the cache occupancy statistics using `db_bench`.
Reviewed By: riversand963
Differential Revision: D39107915
Pulled By: ltamasi
fbshipit-source-id: 8446c3b190a41a144030df73f318eeda4398c125
Summary:
RocksDB's `Cache` abstraction currently supports two priority levels for items: high (used for frequently accessed/highly valuable SST metablocks like index/filter blocks) and low (used for SST data blocks). Blobs are typically lower-value targets for caching than data blocks, since 1) with BlobDB, data blocks containing blob references conceptually form an index structure which has to be consulted before we can read the blob value, and 2) cached blobs represent only a single key-value, while cached data blocks generally contain multiple KVs. Since we would like to make it possible to use the same backing cache for the block cache and the blob cache, it would make sense to add a new, lower-than-low cache priority level (bottom level) for blobs so data blocks are prioritized over them.
This task is a part of https://github.com/facebook/rocksdb/issues/10156
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10461
Reviewed By: siying
Differential Revision: D38672823
Pulled By: ltamasi
fbshipit-source-id: 90cf7362036563d79891f47be2cc24b827482743
Summary:
... so that cache keys can be derived from DB manifest data
before reading the file from storage--so that every part of the file
can potentially go in a persistent cache.
See updated comments in cache_key.cc for technical details. Importantly,
the new cache key encoding uses some fancy but efficient math to pack
data into the cache key without depending on the sizes of the various
pieces. This simplifies some existing code creating cache keys, like
cache warming before the file size is known.
This should provide us an essentially permanent mapping between SST
unique IDs and base cache keys, with the ability to "upgrade" SST
unique IDs (and thus cache keys) with new SST format_versions.
These cache keys are of similar, perhaps indistinguishable quality to
the previous generation. Before this change (see "corrected" days
between collision):
```
./cache_bench -stress_cache_key -sck_keep_bits=43
18 collisions after 2 x 90 days, est 10 days between (1.15292e+19 corrected)
```
After this change (keep 43 bits, up through 50, to validate "trajectory"
is ok on "corrected" days between collision):
```
19 collisions after 3 x 90 days, est 14.2105 days between (1.63836e+19 corrected)
16 collisions after 5 x 90 days, est 28.125 days between (1.6213e+19 corrected)
15 collisions after 7 x 90 days, est 42 days between (1.21057e+19 corrected)
15 collisions after 17 x 90 days, est 102 days between (1.46997e+19 corrected)
15 collisions after 49 x 90 days, est 294 days between (2.11849e+19 corrected)
15 collisions after 62 x 90 days, est 372 days between (1.34027e+19 corrected)
15 collisions after 53 x 90 days, est 318 days between (5.72858e+18 corrected)
15 collisions after 309 x 90 days, est 1854 days between (1.66994e+19 corrected)
```
However, the change does modify (probably weaken) the "guaranteed unique" promise from this
> SST files generated in a single process are guaranteed to have unique cache keys, unless/until number session ids * max file number = 2**86
to this (see https://github.com/facebook/rocksdb/issues/10388)
> With the DB id limitation, we only have nice guaranteed unique cache keys for files generated in a single process until biggest session_id_counter and offset_in_file reach combined 64 bits
I don't think this is a practical concern, though.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10394
Test Plan: unit tests updated, see simulation results above
Reviewed By: jay-zhuang
Differential Revision: D38667529
Pulled By: pdillinger
fbshipit-source-id: 49af3fe7f47e5b61162809a78b76c769fd519fba
Summary:
A test in db_block_cache_test.cc was skipping ClockCache due to the 16-byte key length requirement. We fixed this. Along the way, we fixed a bug in ApplyToSomeEntries, which assumed the function being applied could modify handle metadata, and thus took an exclusive reference. This is incompatible with calls that need to inspect every element (including externally referenced ones) to gather stats.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10482
Test Plan: ``make -j24 check``
Reviewed By: anand1976
Differential Revision: D38553073
Pulled By: guidotag
fbshipit-source-id: 0ed63fed4d3b89e5056b35b7091fce579f5647ae
Summary:
This fix is to replace `AllocateBlock()` with `new`. Once I figure out why `AllocateBlock()` might cause the segfault, I will update the implementation.
Fix the bug that causes ./compressed_secondary_cache_test output following test failures:
```
Note: Google Test filter = CompressedSecondaryCacheTest.MergeChunksIntoValueTest
[==========] Running 1 test from 1 test case.
[----------] Global test environment set-up.
[----------] 1 test from CompressedSecondaryCacheTest
[ RUN ] CompressedSecondaryCacheTest.MergeChunksIntoValueTest
[ OK ] CompressedSecondaryCacheTest.MergeChunksIntoValueTest (1 ms)
[----------] 1 test from CompressedSecondaryCacheTest (1 ms total)
[----------] Global test environment tear-down
[==========] 1 test from 1 test case ran. (9 ms total)
[ PASSED ] 1 test.
t/run-compressed_secondary_cache_test-CompressedSecondaryCacheTest.MergeChunksIntoValueTest: line 4: 1091086 Segmentation fault (core dumped) TEST_TMPDIR=$d ./compressed_secondary_cache_test --gtest_filter=CompressedSecondaryCacheTest.MergeChunksIntoValueTest
Note: Google Test filter = CompressedSecondaryCacheTest.BasicTestWithMemoryAllocatorAndCompression
[==========] Running 1 test from 1 test case.
[----------] Global test environment set-up.
[----------] 1 test from CompressedSecondaryCacheTest
[ RUN ] CompressedSecondaryCacheTest.BasicTestWithMemoryAllocatorAndCompression
[ OK ] CompressedSecondaryCacheTest.BasicTestWithMemoryAllocatorAndCompression (1 ms)
[----------] 1 test from CompressedSecondaryCacheTest (1 ms total)
[----------] Global test environment tear-down
[==========] 1 test from 1 test case ran. (2 ms total)
[ PASSED ] 1 test.
t/run-compressed_secondary_cache_test-CompressedSecondaryCacheTest.BasicTestWithMemoryAllocatorAndCompression: line 4: 1090883 Segmentation fault (core dumped) TEST_TMPDIR=$d ./compressed_secondary_cache_test --gtest_filter=CompressedSecondaryCacheTest.BasicTestWithMemoryAllocatorAndCompression
```
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10507
Test Plan:
Test 1:
```
$make -j 24
$./compressed_secondary_cache_test
```
Test 2:
```
$COMPILE_WITH_ASAN=1 make -j 24
$./compressed_secondary_cache_test
```
Test 3:
```
$COMPILE_WITH_TSAN=1 make -j 24
$./compressed_secondary_cache_test
```
Reviewed By: anand1976
Differential Revision: D38529885
Pulled By: gitbw95
fbshipit-source-id: d903fa3fadbd4d29f9528728c63a4f61c4396890
Summary:
Currently, `SetIsInSecondaryCache` is after `Promote`. After `Promote`, a handle can be accessed and its flags can be set. This causes data race.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10472
Test Plan:
unit tests
stress tests
Reviewed By: pdillinger
Differential Revision: D38403991
Pulled By: gitbw95
fbshipit-source-id: 0aaa2d2edeaf5bc799fcce605648fe49eb7119c2
Summary:
### **Summary:**
To minimize the internal fragmentation caused by the variable size of the compressed blocks, the original block is split according to the jemalloc bin size in `Insert()` and then merged back in `Lookup()`. Based on the analysis of the results of the following tests, from the overall internal fragmentation perspective, this PR does mitigate the internal fragmentation issue.
_Do more myshadow tests with the latest commit. I finished several myshadow AB Testing and the results are promising. For the config of 4GB primary cache and 3GB secondary cache, Jemalloc resident stats shows consistently ~0.15GB memory saving; the allocated and active stats show similar memory savings. The CPU usage is almost the same before and after this PR._
To evaluate the issue of memory fragmentations and the benefits of this PR, I conducted two sets of local tests as follows.
**T1**
Keys: 16 bytes each (+ 0 bytes user-defined timestamp)
Values: 100 bytes each (50 bytes after compression)
Entries: 90000000
RawSize: 9956.4 MB (estimated)
FileSize: 5664.8 MB (estimated)
| Test Name | Primary Cache Size (MB) | Compressed Secondary Cache Size (MB) |
| - | - | - |
| T1_3 | 4000 | 4000 |
| T1_4 | 2000 | 3000 |
Populate the DB:
./db_bench --benchmarks=fillrandom --num=90000000 -db=/mem_fragmentation/db_bench_1
Overwrite it to a stable state:
./db_bench --benchmarks=overwrite --num=90000000 -use_existing_db -db=/mem_fragmentation/db_bench_1
Run read tests with differnt cache setting:
T1_3:
MALLOC_CONF="prof:true,prof_stats:true" ../rocksdb/db_bench --benchmarks=seekrandom --threads=16 --num=90000000 -use_existing_db --benchmark_write_rate_limit=52000000 -use_direct_reads --cache_size=4000000000 -compressed_secondary_cache_size=4000000000 -use_compressed_secondary_cache -db=/mem_fragmentation/db_bench_1 --print_malloc_stats=true > ~/temp/mem_frag/20220710/jemalloc_stats_json_T1_3_20220710 -duration=1800 &
T1_4:
MALLOC_CONF="prof:true,prof_stats:true" ../rocksdb/db_bench --benchmarks=seekrandom --threads=16 --num=90000000 -use_existing_db --benchmark_write_rate_limit=52000000 -use_direct_reads --cache_size=2000000000 -compressed_secondary_cache_size=3000000000 -use_compressed_secondary_cache -db=/mem_fragmentation/db_bench_1 --print_malloc_stats=true > ~/temp/mem_frag/20220710/jemalloc_stats_json_T1_4_20220710 -duration=1800 &
For T1_3 and T1_4, I also conducted the tests before and after this PR. The following table show the important jemalloc stats.
| Test Name | T1_3 | T1_3 after mem defrag | T1_4 | T1_4 after mem defrag |
| - | - | - | - | - |
| allocated (MB) | 8728 | 8076 | 5518 | 5043 |
| available (MB) | 8753 | 8092 | 5536 | 5051 |
| external fragmentation rate | 0.003 | 0.002 | 0.003 | 0.0016 |
| resident (MB) | 8956 | 8365 | 5655 | 5235 |
**T2**
Keys: 32 bytes each (+ 0 bytes user-defined timestamp)
Values: 256 bytes each (128 bytes after compression)
Entries: 40000000
RawSize: 10986.3 MB (estimated)
FileSize: 6103.5 MB (estimated)
| Test Name | Primary Cache Size (MB) | Compressed Secondary Cache Size (MB) |
| - | - | - |
| T2_3 | 4000 | 4000 |
| T2_4 | 2000 | 3000 |
Create DB (10GB):
./db_bench -benchmarks=fillrandom -use_direct_reads=true -num=40000000 -key_size=32 -value_size=256 -db=/mem_fragmentation/db_bench_2
Overwrite it to a stable state:
./db_bench --benchmarks=overwrite --num=40000000 -use_existing_db -key_size=32 -value_size=256 -db=/mem_fragmentation/db_bench_2
Run read tests with differnt cache setting:
T2_3:
MALLOC_CONF="prof:true,prof_stats:true" ./db_bench --benchmarks="mixgraph" -use_direct_io_for_flush_and_compaction=true -use_direct_reads=true -cache_size=4000000000 -compressed_secondary_cache_size=4000000000 -use_compressed_secondary_cache -keyrange_dist_a=14.18 -keyrange_dist_b=-2.917 -keyrange_dist_c=0.0164 -keyrange_dist_d=-0.08082 -keyrange_num=30 -value_k=0.2615 -value_sigma=25.45 -iter_k=2.517 -iter_sigma=14.236 -mix_get_ratio=0.85 -mix_put_ratio=0.14 -mix_seek_ratio=0.01 -sine_mix_rate_interval_milliseconds=5000 -sine_a=1000 -sine_b=0.000073 -sine_d=400000 -reads=80000000 -num=40000000 -key_size=32 -value_size=256 -use_existing_db=true -db=/mem_fragmentation/db_bench_2 --print_malloc_stats=true > ~/temp/mem_frag/jemalloc_stats_T2_3 -duration=1800 &
T2_4:
MALLOC_CONF="prof:true,prof_stats:true" ./db_bench --benchmarks="mixgraph" -use_direct_io_for_flush_and_compaction=true -use_direct_reads=true -cache_size=2000000000 -compressed_secondary_cache_size=3000000000 -use_compressed_secondary_cache -keyrange_dist_a=14.18 -keyrange_dist_b=-2.917 -keyrange_dist_c=0.0164 -keyrange_dist_d=-0.08082 -keyrange_num=30 -value_k=0.2615 -value_sigma=25.45 -iter_k=2.517 -iter_sigma=14.236 -mix_get_ratio=0.85 -mix_put_ratio=0.14 -mix_seek_ratio=0.01 -sine_mix_rate_interval_milliseconds=5000 -sine_a=1000 -sine_b=0.000073 -sine_d=400000 -reads=80000000 -num=40000000 -key_size=32 -value_size=256 -use_existing_db=true -db=/mem_fragmentation/db_bench_2 --print_malloc_stats=true > ~/temp/mem_frag/jemalloc_stats_T2_4 -duration=1800 &
For T2_3 and T2_4, I also conducted the tests before and after this PR. The following table show the important jemalloc stats.
| Test Name | T2_3 | T2_3 after mem defrag | T2_4 | T2_4 after mem defrag |
| - | - | - | - | - |
| allocated (MB) | 8425 | 8093 | 5426 | 5149 |
| available (MB) | 8489 | 8138 | 5435 | 5158 |
| external fragmentation rate | 0.008 | 0.0055 | 0.0017 | 0.0017 |
| resident (MB) | 8676 | 8392 | 5541 | 5321 |
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10287
Test Plan: Unit tests.
Reviewed By: anand1976
Differential Revision: D37743362
Pulled By: gitbw95
fbshipit-source-id: 0010c5af08addeacc5ebbc4ffe5be882fb1d38ad
Summary:
If a secondary cache is configured, its possible that a cache lookup will get a hit in the secondary cache. In that case, the ```LRUCacheShard::Lookup``` doesn't immediately update the ```total_charge``` for the item handle if the ```wait``` parameter is false (i.e caller will call later to check the completeness). However, ```BlockBasedTable::GetEntryFromCache``` assumes the handle is complete and calls ```UpdateCacheHitMetrics```, which checks the usage of the cache item and fails the assert in https://github.com/facebook/rocksdb/blob/main/cache/lru_cache.h#L237 (```assert(total_charge >= meta_charge)```).
To fix this, we call ```UpdateCacheHitMetrics``` later in ```MultiGet```, after waiting for all cache lookup completions.
Test plan -
Run crash test with changes from https://github.com/facebook/rocksdb/issues/10160
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10440
Reviewed By: gitbw95
Differential Revision: D38283968
Pulled By: anand1976
fbshipit-source-id: 31c54ef43517726c6e5fdda81899b364241dd7e1
Summary:
RocksDB's `Cache` abstraction currently supports two priority levels for items: high (used for frequently accessed/highly valuable SST metablocks like index/filter blocks) and low (used for SST data blocks). Blobs are typically lower-value targets for caching than data blocks, since 1) with BlobDB, data blocks containing blob references conceptually form an index structure which has to be consulted before we can read the blob value, and 2) cached blobs represent only a single key-value, while cached data blocks generally contain multiple KVs. Since we would like to make it possible to use the same backing cache for the block cache and the blob cache, it would make sense to add a new, lower-than-low cache priority level (bottom level) for blobs so data blocks are prioritized over them.
This task is a part of https://github.com/facebook/rocksdb/issues/10156
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10309
Reviewed By: ltamasi
Differential Revision: D38211655
Pulled By: gangliao
fbshipit-source-id: 65ef33337db4d85277cc6f9782d67c421ad71dd5
Summary:
This fixes two issues:
- [T127355728](https://www.internalfb.com/intern/tasks/?t=127355728): In the stress tests, when the ClockCache is operating close to full capacity and a burst of inserts are concurrently executed, every slot in the hash table may become occupied. This contradicts an assertion in the code, which is no longer valid in the lock-free setting. We are removing that assertion and handling the case of an insertion into a full table.
- [T127427659](https://www.internalfb.com/intern/tasks/?t=127427659): There was a memory leak when an insertion is performed over capacity, but no handle is provided. In that case, a handle was dynamically allocated, but the pointer wasn't stored anywhere.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10430
Test Plan:
- ``make -j24 check``
- ``make -j24 USE_CLANG=1 COMPILE_WITH_ASAN=1 COMPILE_WITH_UBSAN=1 CRASH_TEST_EXT_ARGS="--duration=960 --cache_type=clock_cache" blackbox_crash_test_with_atomic_flush``
- ``make -j24 USE_CLANG=1 COMPILE_WITH_TSAN=1 CRASH_TEST_EXT_ARGS="--duration=960 --cache_type=clock_cache" blackbox_crash_test_with_atomic_flush``
Reviewed By: pdillinger
Differential Revision: D38226114
Pulled By: guidotag
fbshipit-source-id: 18f6ab7e6214e11e9721d5ff289db1bf795d0008
Summary:
In this PR we bring ClockCache closer to production quality. We implement the following changes:
1. Fixed a few bugs in ClockCache.
2. ClockCache now fully supports ``strict_capacity_limit == false``: When an insertion over capacity is commanded, we allocate a handle separately from the hash table.
3. ClockCache now runs on almost every test in cache_test. The only exceptions are a test where either the LRU policy is required, and a test that dynamically increases the table capacity.
4. ClockCache now supports dynamically decreasing capacity via SetCapacity. (This is easy: we shrink the capacity upper bound and run the clock algorithm.)
5. Old FastLRUCache tests in lru_cache_test.cc are now also used on ClockCache.
As a byproduct of 1. and 2. we are able to turn on ClockCache in the stress tests.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10418
Test Plan:
- ``make -j24 USE_CLANG=1 COMPILE_WITH_ASAN=1 COMPILE_WITH_UBSAN=1 check``
- ``make -j24 USE_CLANG=1 COMPILE_WITH_TSAN=1 check``
- ``make -j24 USE_CLANG=1 COMPILE_WITH_ASAN=1 COMPILE_WITH_UBSAN=1 CRASH_TEST_EXT_ARGS="--duration=960 --cache_type=clock_cache" blackbox_crash_test_with_atomic_flush``
- ``make -j24 USE_CLANG=1 COMPILE_WITH_TSAN=1 CRASH_TEST_EXT_ARGS="--duration=960 --cache_type=clock_cache" blackbox_crash_test_with_atomic_flush``
Reviewed By: pdillinger
Differential Revision: D38170673
Pulled By: guidotag
fbshipit-source-id: 508987b9dc9d9d68f1a03eefac769820b680340a
Summary:
I recently discovered that block cache keys are slightly lower
quality than previously thought, because my stress testing tool failed
to simulate the effect of DB ID differences. This change updates the
tool and gives us data to guide future developments. (No changes to
production code here and now.)
Nevertheless, the following promise still holds
```
// In fact, if our SST files are all < 4TB (see
// BlockBasedTable::kMaxFileSizeStandardEncoding), then SST files generated
// in a single process are guaranteed to have unique cache keys, unless/until
// number session ids * max file number = 2**86 ...
```
because although different DB IDs could cause collision in file number
and offset data, that would have to be using the same DB session (lower)
to cause a block cache key collision, which is not possible in the same
process. (A session is associated with only one DB ID.)
This change fixes cache_bench -stress_cache_key to set and reset DB IDs in
a parameterized way to evaluate the effect. Previous results assumed to
be representative (using -sck_keep_bits=43):
```
15 collisions after 15 x 90 days, est 90 days between (1.03763e+20 corrected)
```
or expected collision on a single machine every 104 billion billion
days (see "corrected" value).
After accounting for DB IDs, test never really changing, intermediate, and very
frequently changing (using default -sck_db_count=100):
```
-sck_newdb_nreopen=1000000000:
15 collisions after 2 x 90 days, est 12 days between (1.38351e+19 corrected)
-sck_newdb_nreopen=10000:
17 collisions after 2 x 90 days, est 10.5882 days between (1.22074e+19 corrected)
-sck_newdb_nreopen=100:
19 collisions after 2 x 90 days, est 9.47368 days between (1.09224e+19 corrected)
```
or roughly 10x more often than previously thought (still extremely if
not impossibly rare), and better than random base cache keys
(with -sck_randomize), though < 10x better than random:
```
31 collisions after 1 x 90 days, est 2.90323 days between (3.34719e+18 corrected)
```
If we simply fixed this by ignoring DB ID for cache keys, we would
potentially have a shortage of entropy for some cases, such as small
file numbers and offsets (e.g. many short-lived processes each using
SstFileWriter to create a small file), because existing DB session IDs
only provide ~103 bits of entropy. We could upgrade the entropy in DB
session IDs to accommodate, but it's not known what all would be
affected by changing from 20 digit session IDs to something larger.
Instead, my plan is to
1) Move to block cache keys derived from SST unique IDs (so that we can
derive block cache keys from manifest data without reading file on
storage), and show no significant regression in expected collision
rate.
2) Generate better SST unique IDs in format_version=6 (https://github.com/facebook/rocksdb/issues/9058),
which should have ~100x lower expected/predicted collision rate based
on simulations with this stress test:
```
./cache_bench -stress_cache_key -sck_keep_bits=39 -sck_newdb_nreopen=100 -sck_footer_unique_id
...
15 collisions after 19 x 90 days, est 114 days between (2.10293e+21 corrected)
```
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10388
Test Plan: no production changes
Reviewed By: jay-zhuang
Differential Revision: D37986714
Pulled By: pdillinger
fbshipit-source-id: e759b2469e3365cb01c6661a69e0ab849ef4c3df
Summary:
ClockCache completely free of locks. As part of this PR we have also pushed clock algorithm functionality out of ClockCacheShard into ClockHandleTable, so that ClockCacheShard acts more as an interface and less as an actual data structure.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10390
Test Plan:
- ``make -j24 check``
- ``make -j24 CRASH_TEST_EXT_ARGS="--duration=960 --cache_type=clock_cache --cache_size=1073741824 --block_size=16384" blackbox_crash_test_with_atomic_flush``
Reviewed By: pdillinger
Differential Revision: D38106945
Pulled By: guidotag
fbshipit-source-id: 6cbf6bd2397dc9f582809ccff5118a8a33ea6cb1
Summary:
To help service owners to manage their memory budget effectively, we have been working towards counting all major memory users inside RocksDB towards a single global memory limit (see e.g. https://github.com/facebook/rocksdb/wiki/Write-Buffer-Manager#cost-memory-used-in-memtable-to-block-cache). The global limit is specified by the capacity of the block-based table's block cache, and is technically implemented by inserting dummy entries ("reservations") into the block cache. The goal of this task is to support charging the memory usage of the new blob cache against this global memory limit when the backing cache of the blob cache and the block cache are different.
This PR is a part of https://github.com/facebook/rocksdb/issues/10156
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10321
Reviewed By: ltamasi
Differential Revision: D37913590
Pulled By: gangliao
fbshipit-source-id: eaacf23907f82dc7d18964a3f24d7039a2937a72
Summary:
RocksDB supports a two-level cache hierarchy (see https://rocksdb.org/blog/2021/05/27/rocksdb-secondary-cache.html), where items evicted from the primary cache can be spilled over to the secondary cache, or items from the secondary cache can be promoted to the primary one. We have a CacheLib-based non-volatile secondary cache implementation that can be used to improve read latencies and reduce the amount of network bandwidth when using distributed file systems. In addition, we have recently implemented a compressed secondary cache that can be used as a replacement for the OS page cache when e.g. direct I/O is used. The goals of this task are to add support for using a secondary cache with the blob cache and to measure the potential performance gains using `db_bench`.
This task is a part of https://github.com/facebook/rocksdb/issues/10156
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10349
Reviewed By: ltamasi
Differential Revision: D37896773
Pulled By: gangliao
fbshipit-source-id: 7804619ce4a44b73d9e11ad606640f9385969c84
Summary:
This is a prototype of a partially lock-free version of ClockCache. Roughly speaking, reads are lock-free and writes are lock-based:
- Lookup is lock-free.
- Release is lock-free, unless (i) no references to the element are left and (ii) it was marked for deletion or ``erase_if_last_ref`` is set.
- Insert and Erase still use a per-shard lock.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10347
Test Plan:
- ``make -j24 check``
- ``make -j24 CRASH_TEST_EXT_ARGS="--duration=960 --cache_type=clock_cache --cache_size=1073741824 --block_size=16384" blackbox_crash_test_with_atomic_flush``
Reviewed By: pdillinger
Differential Revision: D37898776
Pulled By: guidotag
fbshipit-source-id: 6418fd980f786d69b871bf2fe959398e44cd3d80
Summary:
Sometimes we may not want to include extra computation in our cache_bench experiments. Here we add a flag to avoid any extra work. We also moved the timer start after the key generation.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10363
Test Plan: Run cache_bench with and without the new flag and check that the appropriate code is being executed.
Reviewed By: pdillinger
Differential Revision: D37870416
Pulled By: guidotag
fbshipit-source-id: f853207b6643b9328e774251c3f679b1fd78a11a
Summary:
This complements https://github.com/facebook/rocksdb/issues/10351. This PR reverts NewClockCache's signature to an older version, expected by the users of the old (buggy) ClockCache.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10358
Test Plan: ``make -j24 check`` and re-run the pre-release tests.
Reviewed By: siying
Differential Revision: D37832601
Pulled By: guidotag
fbshipit-source-id: 32a91d3da4119be187935003b7b897272ceb1950
Summary:
If the primary cache is LRU cache and there is a secondary cache, add Secondary Cache printable options into LRUCache::GetPrintableOptions.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10346
Test Plan:
1. Current Unit Tests should pass.
2. Use db_bench (with compressed_secondary_cache ) and the LOG should includes the new printable options from Seoncdary Cache.
Reviewed By: jay-zhuang
Differential Revision: D37779310
Pulled By: gitbw95
fbshipit-source-id: 88ce1f7df6b5f25740e598d9e7fa91e4c414cb8f
Summary:
ClockCache is still in experimental stage, and currently fails some pre-release fbcode tests. See https://www.internalfb.com/diff/D37772011. API calls to construct ClockCache are done via the function NewClockCache. For now, NewClockCache calls will return an LRUCache (with appropriate arguments), which is stable.
The idea that NewClockCache returns nullptr was also floated, but this would be interpreted as unsupported cache, and a default LRUCache would be constructed instead, potentially causing a performance regression that is harder to identify.
A new version of the NewClockCache function was created for our internal tests.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10351
Test Plan: ``make -j24 check`` and re-run the pre-release tests.
Reviewed By: pdillinger
Differential Revision: D37802685
Pulled By: guidotag
fbshipit-source-id: 0a8d10612ff21e576f7360cb13e20bc36e244972
Summary:
The blob cache enables an optimization on the read path: when a blob is found in the cache, we can avoid copying it into the buffer provided by the application. Instead, we can simply transfer ownership of the cache handle to the target `PinnableSlice`. (Note: this relies on the `Cleanable` interface, which is implemented by `PinnableSlice`.)
This has the potential to save a lot of CPU, especially with large blob values.
This task is a part of https://github.com/facebook/rocksdb/issues/10156
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10297
Reviewed By: riversand963
Differential Revision: D37640311
Pulled By: gangliao
fbshipit-source-id: 92de0e35cc703d06c87c5c1861cc2899ec52234a
Summary:
When an element is first inserted into the ClockCache, it is now assigned either medium or high clock priority, depending on whether its cache priority is low or high, respectively. This is a variant of LRUCache's midpoint insertions. The main difference is that LRUCache can specify the allocated capacity for high-priority elements via the ``high_pri_pool_ratio`` parameter. Contrarily, in ClockCache, low- and high-priority elements compete for all cache slots, and one group can take over the other (of course, it takes more low-priority insertions to push out high-priority elements). However, just as LRUCache, ClockCache provides the following guarantee: a high-priority element will not be evicted before a low-priority element that was inserted earlier in time.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10305
Test Plan: ``make -j24 check``
Reviewed By: pdillinger
Differential Revision: D37607787
Pulled By: guidotag
fbshipit-source-id: 24d9f2523d2f4e6415e7f0029cc061fa275c2040
Summary:
I noticed it would clean up some things to have Cache::Insert()
return our MemoryLimit Status instead of Incomplete for the case in
which the capacity limit is reached. I suspect this fixes some existing but
unknown bugs where this Incomplete could be confused with other uses
of Incomplete, especially no_io cases. This is the most suspicious case I
noticed, but was not able to reproduce a bug, in part because the existing
code is not covered by unit tests (FIXME added): 57adbf0e91/table/get_context.cc (L397)
I audited all the existing uses of IsIncomplete and updated those that
seemed relevant.
HISTORY updated with a clear warning to users of strict_capacity_limit=true
to update uses of `IsIncomplete()`
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10262
Test Plan: updated unit tests
Reviewed By: hx235
Differential Revision: D37473155
Pulled By: pdillinger
fbshipit-source-id: 4bd9d9353ccddfe286b03ebd0652df8ce20f99cb
Summary:
We fix two bugs in CalcHashBits. The first one is an off-by-one error: the desired number of table slots is the real number ``capacity / (kLoadFactor * handle_charge)``, which should not be rounded down. The second one is that we should disallow inputs that set the element charge to 0, namely ``estimated_value_size == 0 && metadata_charge_policy == kDontChargeCacheMetadata``.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10295
Test Plan: CalcHashBits is tested by CalcHashBitsTest (in lru_cache_test.cc). The test now iterates over many more inputs; it covers, in particular, the rounding error edge case. Overall, the test is now more robust. Run ``make -j24 check``.
Reviewed By: pdillinger
Differential Revision: D37573797
Pulled By: guidotag
fbshipit-source-id: ea4f4439f7196ab1c1afb88f566fe92850537262
Summary:
This is the initial step in the development of a lock-free clock cache. This PR includes the base hash table design (which we mostly ported over from FastLRUCache) and the clock eviction algorithm. Importantly, it's still _not_ lock-free---all operations use a shard lock. Besides the locking, there are other features left as future work:
- Remove keys from the handles. Instead, use 128-bit bijective hashes of them for handle comparisons, probing (we need two 32-bit hashes of the key for double hashing) and sharding (we need one 6-bit hash).
- Remove the clock_usage_ field, which is updated on every lookup. Even if it were atomically updated, it could cause memory invalidations across cores.
- Middle insertions into the clock list.
- A test that exercises the clock eviction policy.
- Update the Java API of ClockCache and Java calls to C++.
Along the way, we improved the code and comments quality of FastLRUCache. These changes are relatively minor.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10273
Test Plan: ``make -j24 check``
Reviewed By: pdillinger
Differential Revision: D37522461
Pulled By: guidotag
fbshipit-source-id: 3d70b737dbb70dcf662f00cef8c609750f083943
Summary:
cache_bench wasn't generating 16B keys, which are necessary for FastLRUCache. Also:
- Added asserts in cache_bench, which is assuming that inserts never fail. When they fail (for example, if we used keys of the wrong size), memory allocated to the values will becomes leaked, and eventually the program crashes.
- Move kCacheKeySize to the right spot.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10234
Test Plan:
``make -j24 check``. Also, run cache_bench with FastLRUCache and check that memory usage doesn't blow up:
``./cache_bench -cache_type=fast_lru_cache -num_shard_bits=6 -skewed=true \
-lookup_insert_percent=100 -lookup_percent=0 -insert_percent=0 -erase_percent=0 \
-populate_cache=true -cache_size=1073741824 -ops_per_thread=10000000 \
-value_bytes=8192 -resident_ratio=1 -threads=16``
Reviewed By: pdillinger
Differential Revision: D37382949
Pulled By: guidotag
fbshipit-source-id: b697a942ebb215de5d341f98dc8566763436ba9b
Summary:
In FastLRUCache, we replace the current chained per-shard hash table by an open-addressing hash table. In particular, this allows us to preallocate all handles.
Because all handles are preallocated, this implementation doesn't support strict_capacity_limit = false (i.e., allowing insertions beyond the predefined capacity). This clashes with current assumptions of some tests, namely two tests in cache_test and the crash tests. We have disabled these for now.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10194
Test Plan: ``make -j24 check``
Reviewed By: pdillinger
Differential Revision: D37296770
Pulled By: guidotag
fbshipit-source-id: 232ff1b8260331d868ebf4e3e5d8ad709390b0ad
Summary:
folly DistributedMutex is faster than standard mutexes though
imposes some static obligations on usage. See
https://github.com/facebook/folly/blob/main/folly/synchronization/DistributedMutex.h
for details. Here we use this alternative for our Cache implementations
(especially LRUCache) for better locking performance, when RocksDB is
compiled with folly.
Also added information about which distributed mutex implementation is
being used to cache_bench output and to DB LOG.
Intended follow-up:
* Use DMutex in more places, perhaps improving API to support non-scoped
locking
* Fix linking with fbcode compiler (needs ROCKSDB_NO_FBCODE=1 currently)
Credit: Thanks Siying for reminding me about this line of work that was previously
left unfinished.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10179
Test Plan:
for correctness, existing tests. CircleCI config updated.
Also Meta-internal buck build updated.
For performance, ran simultaneous before & after cache_bench. Out of three
comparison runs, the middle improvement to ops/sec was +21%:
Baseline: USE_CLANG=1 DEBUG_LEVEL=0 make -j24 cache_bench (fbcode
compiler)
```
Complete in 20.201 s; Rough parallel ops/sec = 1584062
Thread ops/sec = 107176
Operation latency (ns):
Count: 32000000 Average: 9257.9421 StdDev: 122412.04
Min: 134 Median: 3623.0493 Max: 56918500
Percentiles: P50: 3623.05 P75: 10288.02 P99: 30219.35 P99.9: 683522.04 P99.99: 7302791.63
```
New: (add USE_FOLLY=1)
```
Complete in 16.674 s; Rough parallel ops/sec = 1919135 (+21%)
Thread ops/sec = 135487
Operation latency (ns):
Count: 32000000 Average: 7304.9294 StdDev: 108530.28
Min: 132 Median: 3777.6012 Max: 91030902
Percentiles: P50: 3777.60 P75: 10169.89 P99: 24504.51 P99.9: 59721.59 P99.99: 1861151.83
```
Reviewed By: anand1976
Differential Revision: D37182983
Pulled By: pdillinger
fbshipit-source-id: a17eb05f25b832b6a2c1356f5c657e831a5af8d1
Summary:
We make the size of the per-shard hash table fixed. The base level of the hash table is now preallocated with the required capacity. The user must provide an estimate of the size of the values.
Notice that even though the base level becomes fixed, the chains are still dynamic. Overall, the shard capacity mechanisms haven't changed, so we don't need to test this.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10154
Test Plan: `make -j24 check`
Reviewed By: pdillinger
Differential Revision: D37124451
Pulled By: guidotag
fbshipit-source-id: cba6ac76052fe0ec60b8ff4211b3de7650e80d0c
Summary:
FastLRUCache now only supports 16B keys. The tests have changed to reflect this.
Because the unit tests were designed for caches that accept any string as keys, some tests are no longer compatible with FastLRUCache. We have disabled those for runs with FastLRUCache. (We could potentially change all tests to use 16B keys, but we don't because the cache public API does not require this.)
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10137
Test Plan: make -j24 check
Reviewed By: gitbw95
Differential Revision: D37083934
Pulled By: guidotag
fbshipit-source-id: be1719cf5f8364a9a32bc4555bce1a0de3833b0d
Summary:
Update SecondaryCache::CreateFromString and enable it to create sec cache based on the uri for CompressedSecondaryCache.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10132
Test Plan: Add unit tests.
Reviewed By: anand1976
Differential Revision: D36996997
Pulled By: gitbw95
fbshipit-source-id: 882ad563cff6d38b306a53426ad7e47273f34edc
Summary:
As seen in https://github.com/facebook/rocksdb/issues/10137, simply churning the cache key hashes (e.g.
by changing the raw cache keys) could trigger failure in this test, due
to possibility of some cache shard exceeding its portion of capacity
and evicting entries. Updated the test to be less fragile by using
greater margins, and added a pre-check for evictions, which doesn't
manifest as a race condition, before the main check that can race.
Also added stack trace handler to cache_test for debugging.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10145
Test Plan:
test thousands of iterations with gtest-parallel, including
with changes in https://github.com/facebook/rocksdb/issues/10137 that were surfacing the problem. Pre-check
without the fix would always fail with https://github.com/facebook/rocksdb/issues/10137
Reviewed By: guidotag
Differential Revision: D37058771
Pulled By: pdillinger
fbshipit-source-id: a7cf137967aef49c07ae9602d8523c63e7388fab
Summary:
cache_bench can now run with FastLRUCache.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10095
Test Plan:
- Temporarily add an ``assert(false)`` in the execution path that sets up the FastLRUCache. Run ``make -j24 cache_bench``. Then test the appropriate code is used by running ``./cache_bench -cache_type=fast_lru_cache`` and checking that the assert is called. Repeat for LRUCache.
- Verify that FastLRUCache (currently a clone of LRUCache) has similar latency distribution than LRUCache, by comparing the outputs of ``./cache_bench -cache_type=fast_lru_cache`` and ``./cache_bench -cache_type=lru_cache``.
Reviewed By: pdillinger
Differential Revision: D36875834
Pulled By: guidotag
fbshipit-source-id: eb2ad0bb32c2717a258a6ac66ed736e06f826cd8
Summary:
In LRU Cache mutex, we sometimes call malloc_usable_size() to calculate memory used by the metadata object. We prevent it by saving the charge + metadata size, rather than charge, inside the metadata itself. Within the mutex, usually only total charge is needed so we don't need to repeat.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10026
Test Plan: Run existing tests.
Reviewed By: pdillinger
Differential Revision: D36556253
fbshipit-source-id: f60c96d13cde3af77732e5548e4eac4182fa9801
Summary:
ToString() is created as some platform doesn't support std::to_string(). However, we've already used std::to_string() by mistake for 16 months (in db/db_info_dumper.cc). This commit just remove ToString().
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9955
Test Plan: Watch CI tests
Reviewed By: riversand963
Differential Revision: D36176799
fbshipit-source-id: bdb6dcd0e3a3ab96a1ac810f5d0188f684064471
Summary:
To support a project to prototype and evaluate algorithmic
enhancments and alternatives to LRUCache, here I have separated out
LRUCache into internal-only "FastLRUCache" and cut it down to
essentials, so that details like secondary cache handling and
priorities do not interfere with prototyping. These can be
re-integrated later as needed, along with refactoring to minimize code
duplication (which would slow down prototyping for now).
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9917
Test Plan:
unit tests updated to ensure basic functionality has (likely)
been preserved
Reviewed By: anand1976
Differential Revision: D35995554
Pulled By: pdillinger
fbshipit-source-id: d67b20b7ada3b5d3bfe56d897a73885894a1d9db
Summary:
This gives users the ability to examine the map populated by `GetMapProperty()` with property `kBlockCacheEntryStats`. It also sets us up for a possible future where cache reservations are configured according to `CacheEntryRole`s rather than flags coupled to roles.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9838
Test Plan:
- migrated test DBBlockCacheTest.CacheEntryRoleStats to use this API. That test verifies some of the contents are as expected
- added a DBPropertiesTest to verify the public map keys are present, and nothing else
Reviewed By: hx235
Differential Revision: D35629493
Pulled By: ajkr
fbshipit-source-id: 5c4356b8560e85d1f881fd32c44c15960b02fc68
Summary:
Especially after updating to C++17, I don't see a compelling case for
*requiring* any folly components in RocksDB. I was able to purge the existing
hard dependencies, and it can be quite difficult to strip out non-trivial components
from folly for use in RocksDB. (The prospect of doing that on F14 has changed
my mind on the best approach here.)
But this change creates an optional integration where we can plug in
components from folly at compile time, starting here with F14FastMap to replace
std::unordered_map when possible (probably no public APIs for example). I have
replaced the biggest CPU users of std::unordered_map with compile-time
pluggable UnorderedMap which will use F14FastMap when USE_FOLLY is set.
USE_FOLLY is always set in the Meta-internal buck build, and a simulation of
that is in the Makefile for public CI testing. A full folly build is not needed, but
checking out the full folly repo is much simpler for getting the dependency,
and anything else we might want to optionally integrate in the future.
Some picky details:
* I don't think the distributed mutex stuff is actually used, so it was easy to remove.
* I implemented an alternative to `folly::constexpr_log2` (which is much easier
in C++17 than C++11) so that I could pull out the hard dependencies on
`ConstexprMath.h`
* I had to add noexcept move constructors/operators to some types to make
F14's complainUnlessNothrowMoveAndDestroy check happy, and I added a
macro to make that easier in some common cases.
* Updated Meta-internal buck build to use folly F14Map (always)
No updates to HISTORY.md nor INSTALL.md as this is not (yet?) considered a
production integration for open source users.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9546
Test Plan:
CircleCI tests updated so that a couple of them use folly.
Most internal unit & stress/crash tests updated to use Meta-internal latest folly.
(Note: they should probably use buck but they currently use Makefile.)
Example performance improvement: when filter partitions are pinned in cache,
they are tracked by PartitionedFilterBlockReader::filter_map_ and we can build
a test that exercises that heavily. Build DB with
```
TEST_TMPDIR=/dev/shm/rocksdb ./db_bench -benchmarks=fillrandom -num=10000000 -disable_wal=1 -write_buffer_size=30000000 -bloom_bits=16 -compaction_style=2 -fifo_compaction_max_table_files_size_mb=10000 -fifo_compaction_allow_compaction=0 -partition_index_and_filters
```
and test with (simultaneous runs with & without folly, ~20 times each to see
convergence)
```
TEST_TMPDIR=/dev/shm/rocksdb ./db_bench_folly -readonly -use_existing_db -benchmarks=readrandom -num=10000000 -bloom_bits=16 -compaction_style=2 -fifo_compaction_max_table_files_size_mb=10000 -fifo_compaction_allow_compaction=0 -partition_index_and_filters -duration=40 -pin_l0_filter_and_index_blocks_in_cache
```
Average ops/s no folly: 26229.2
Average ops/s with folly: 26853.3 (+2.4%)
Reviewed By: ajkr
Differential Revision: D34181736
Pulled By: pdillinger
fbshipit-source-id: ffa6ad5104c2880321d8a1aa7187e00ab0d02e94
