10 Commits

Author SHA1 Message Date
Buqian Zheng
052fb6c562
feat: add time based eviction to data managed by cachinglayer (#43490)
issue: https://github.com/milvus-io/milvus/issues/41435

also added disk capacity protection

---------

Signed-off-by: Buqian Zheng <zhengbuqian@gmail.com>
2025-07-29 16:17:35 +08:00
Buqian Zheng
b497d3d7a4
fix: call promise->setValue only after released the ListNode mtx (#43547)
issue: #43261

`promise->setValue(folly::Unit());` may run callbacks inline and some of
them may attempt to grab `mtx_`. So we should not call
`promise->setValue(folly::Unit());` while holding the lock.

---------

Signed-off-by: Buqian Zheng <zhengbuqian@gmail.com>
2025-07-26 18:34:55 +08:00
Buqian Zheng
0599113a4b
enhance: add timeout to resource reservation (#43441)
issue: https://github.com/milvus-io/milvus/issues/41435

Signed-off-by: Buqian Zheng <zhengbuqian@gmail.com>
2025-07-22 15:24:53 +08:00
Buqian Zheng
d793def47c
feat: impose a physical memory limit when loading cells (#43222)
issue: #41435 

issue: https://github.com/milvus-io/milvus/issues/43038

This PR also:


1. removed ERROR state from ListNode
2. CacheSlot will do reserveMemory once for all requested cells after
updating the state to LOADING, so now we transit a cell to LOADING
before its resource reservation
3. reject resource reservation directly if size >= max_size

---------

Signed-off-by: Buqian Zheng <zhengbuqian@gmail.com>
2025-07-18 11:18:52 +08:00
sparknack
9b4081e110
enhance: cachinglayer: some performance optimization (#42858)
issue: #41435

We compared the performance using the modified test_sealed.cpp, which
randomly accesses all rows in all chunks and counts the number of runs
within 3s.

## performance data comparison (ops/second)

chunk config: 1x1000

| Field Type | w/o cachinglayer (commit 640f526301) | w/ cachinglayer |
w/ cachinglayer + opt |
|---|---|---|---|
| Bool field | 82428 | -63.6% (29983) | +2.7% (84675) |
| Int8 field | 82228 | -63.3% (30166) | +2.4% (84163) |
| Int16 field | 82572 | -63.8% (29867) | +1.8% (84036) |
| Int32 field | 82797 | -63.7% (30031) | +1.5% (84043) |
| Int64 field | 81077 | -62.9% (30107) | +0.6% (81604) |
| Float field | 82678 | -63.4% (30266) | +1.8% (84146) |
| Double field | 81925 | -63.4% (29974) | +0.2% (82097) |
| Varchar field | 19933 | -19.6% (16027) | +18.9% (23690) |
| JSON field | 16519 | -96.8% (533) | +2.5% (16927) |
| Int array field | 7325 | -13.7% (6321) | -1.4% (7220) |
| Long array field | 6347 | -8.9% (5781) | -0.1% (6344) |
| Bool array field | 8275 | -14.0% (7116) | +0.4% (8311) |
| String array field | 2281 | -5.0% (2168) | +0.2% (2287) |
| Double array field | 6427 | -13.3% (5574) | -2.0% (6301) |
| Float array field | 7291 | -13.0% (6346) | -1.5% (7183) |
| Vector field | 27487 | -40.4% (16371) | -4.7% (26192) |
| Float16 vector field | 49773 | -54.6% (22601) | -5.9% (46834) |
| BFloat16 vector field | 49783 | -53.1% (23350) | -5.7% (46934) |
| Int8 vector field | 63871 | -59.0% (26179) | -6.2% (59926) |

---

chunk config: 10x1000

| Field Type | w/o cachinglayer (commit 640f526301) | w/ cachinglayer |
w/ cachinglayer + opt |
|---|---|---|---|
| Bool field | 3659 | -48.6% (1879) | +110.1% (7686) |
| Int8 field | 3410 | -45.3% (1864) | +123.9% (7636) |
| Int16 field | 3647 | -48.6% (1874) | +110.1% (7661) |
| Int32 field | 3647 | -48.8% (1866) | +109.6% (7645) |
| Int64 field | 3645 | -48.9% (1863) | +107.8% (7573) |
| Float field | 3647 | -49.0% (1861) | +109.5% (7639) |
| Double field | 3640 | -45.1% (1998) | +108.4% (7586) |
| Varchar field | 1594 | -23.9% (1213) | +20.6% (1922) |
| JSON field | 1202 | -26.5% (884) | +16.1% (1396) |
| Int array field | 602 | -12.3% (528) | +12.7% (678) |
| Long array field | 529 | -12.2% (465) | +7.5% (569) |
| Double array field | 537 | -13.0% (467) | +6.4% (571) |
| Vector field | 1520 | -37.9% (943) | -5.5% (1437) |
| Float16 vector field | 2607 | -47.0% (1382) | +6.4% (2774) |
| BFloat16 vector field | 2586 | -46.5% (1383) | +8.8% (2813) |
| Int8 vector field | 3101 | -47.3% (1633) | +41.9% (4400) |

---------

Signed-off-by: Shawn Wang <shawn.wang@zilliz.com>
2025-07-17 11:20:51 +08:00
Buqian Zheng
b0260d8676
feat: manual evict cache after built interim index (#41836)
issue: https://github.com/milvus-io/milvus/issues/41435

this PR also makes HasRawData of ChunkedSegmentSealedImpl to return
based on metadata, without needing to load the cache just to answer this
simple question.

---------

Signed-off-by: Buqian Zheng <zhengbuqian@gmail.com>
2025-05-16 16:34:23 +08:00
Buqian Zheng
ff5c2770e5
feat: cachinglayer: various improvements (#41546)
issue: https://github.com/milvus-io/milvus/issues/41435

this PR is based on https://github.com/milvus-io/milvus/pull/41436. 

Improvements include:

- Lazy Load support for Storage v1
- Use Low/High watermark to control eviction
- Caching Layer related config changes
- Removed ChunkCache related configs and code in golang
- Add `PinAllCells` helper method to CacheSlot class
- Modified ValueAt, RawAt, PrimitiveRawAt to Bulk version, to reduce
caching layer overhead
- Removed some unclear templated bulk_subscript methods
- CachedSearchIterator to store PinWrapper when searching on
ChunkedColumn, and removed unused contrustor.

---------

Signed-off-by: Buqian Zheng <zhengbuqian@gmail.com>
2025-05-10 09:19:16 +08:00
foxspy
e2ddbe4962
feat: add cachinglayer to index (#41653)
issue: #41435

Signed-off-by: xianliang.li <xianliang.li@zilliz.com>
2025-05-08 10:12:54 +08:00
sthuang
e9442f575d
feat: storage v2 seal segment load (#41567)
storage v2 chunked seal segment loading is based on caching layer. A
cell unit in storage v2 is a parquet row group in remote object storage,
containing all fields. Therefore, each field needs a proxy to do related
one field operations.

<img width="965" alt="Screenshot 2025-04-28 at 10 59 30"
src="https://github.com/user-attachments/assets/83e93a10-3b1d-4066-ac17-b996d5650416"
/>

related: #39173

---------

Signed-off-by: shaoting-huang <shaoting.huang@zilliz.com>
2025-04-30 14:22:58 +08:00
Buqian Zheng
3de904c7ea
feat: add cachinglayer to sealed segment (#41436)
issue: https://github.com/milvus-io/milvus/issues/41435

---------

Signed-off-by: Buqian Zheng <zhengbuqian@gmail.com>
2025-04-28 10:52:40 +08:00