SCC: Scalable Concurrent Containers

SCC offers scalable concurrent containers written in the Rust language. The data structures in SCC are meant to be used by database management software running on a server, ane therefore they may not efficiently work with a small set of data.

scc::HashMap
scc::TreeIndex

scc::HashMap

scc::HashMap is a scalable in-memory unique key-value store that is targeted at highly concurrent heavy workloads. It does not distribute data to multiple shards as most concurrent hash maps do, instead only does it have a single array of entries and corresponding metadata cell array. The metadata management strategy is similar to that of Swisstable; a metadata cell which is separated from the key-value array, is a 64-byte data structure for managing consecutive sixteen entries in the key-value array. The metadata cell also has a linked list of entry arrays for hash collision resolution. scc::HashMap automatically enlarges and shrinks the capacity of its internal array, and resizing happens without blocking other operations and threads. In order to keep the predictable latency of each operation, it does not rehash every entry in the container at once when resizing, instead it distributes the resizing workload to future access to the data structure.

The experimental hardware-transactional-memory functions (e.g. https://stdrs.dev/nightly/x8664-pc-windows-gnu/core/corearch/x86/rtm/index.html) allow read operations to be entirely memory-write-free, thereby boosting performance without compromising semantics and memory consumption. scc::HashMap will start using TSX/HTM intrinsics as soon as the functions become stabilized.

Performance

Test setup

OS: SUSE Linux Enterprise Server 15 SP1
CPU: Intel(R) Xeon(R) CPU E7-8880 v4 @ 2.20GHz x 4
RAM: 1TB
Rust compiler version: 1.48.0
SCC version: 0.3.1
The hashmap is generated using the default parameters: the RandomState hasher builder, and 256 preallocated entries.
In order to minimize the cost of page fault handling, all the tests were run twice, and only the best results were taken.

Test data

Each thread is assigned a disjoint range of u64 integers.
The entropy of the test input is very low, however it does not undermine the test result as the key distribution method is agnostic to the input pattern.
The performance test code asserts the expected outcome of each operation, and the post state of the hashmap instance.

Test workload: local

Insert: each thread inserts 128M records.
Read: each thread reads 128M records.
Remove: each thread removes 128M records.
The data for Read/Remove tests is populated by the Insert test.

| | 11 threads | 22 threads | 44 threads | 88 threads | |--------|----------------|----------------|----------------|----------------| | Insert | 156.423361787s | 187.157442477s | 264.075874751s | 463.032489985s | | Read | 81.03393205s | 92.933046817s | 109.303575217s | 137.802145824s | | Remove | 85.563265194s | 102.896206291s | 117.072458551s | 167.450069665s |

Test workload: local-remote

Insert/Remove: each thread additionally tries to perform assigned operations using keys belonging to other threads.
Mixed: each thread performs 128M insert-local -> insert-remote -> read-local -> read-remote -> remove-local -> remove-remote sequences.
The data for Mixed/Remove tests is populated by the Insert test.
The target remote thread is randomly chosen.
The total operation count per Insert/Read thread is 256M, and half of the operations are bound to fail.
The total operation count per Mixed thread is 768M, and about half of the operations are bound to fail.

| | 11 threads | 22 threads | 44 threads | 88 threads | |--------|----------------|----------------|----------------|----------------| | Insert | 272.420310927s | 314.424537182s | 432.493505328s | 772.267595819s | | Mixed | 326.767954659s | 350.603202721s | 375.987412301s | 433.899012681s | | Remove | 164.857461617s | 184.528933216s | 199.187884668s | 250.735616868s |

scc::TreeIndex

Work-in-progress - Known issues with TreeIndex::remove: #21, and #22 - Not fully validated - Not fully optimized

scc::TreeIndex is a B+-tree variant optimized for read operations. Locks are only acquired on structural changes, and read/scan operations are neither blocked nor interrupted by other threads. The semantics of each API is similar to the read-committed isolation level in terms of database management software, as readers may not see the snapshot of data that is newer than the read snapshot.

Changelog

0.3.20

Stabilize TreeIndex::remove: #20 fixed

0.3.19

Stabilize TreeIndex::remove: #19 fixed

0.3.18

Change TreeIndex::export to TreeIndex::print

0.3.17

Stabilize TreeIndex::from and TreeIndex::export

0.3.16

Implement TreeIndex::clear and (WIP)TreeIndex::from / license changed from MIT -> Apache 2.0 to comply with the company policies (SAP)

0.3.15

Fix a data race between TreeIndex::remove and Scanner

0.3.14

Implement TreeIndex::Node memory reclamation

0.3.13

Fix a read-after-free problem

0.3.12

Implement TreeIndex::Leaf memory reclamation

0.3.11

Refine scc::TreeIndex APIs

0.3.10

API change: scc::HashMap::Scanner -> scc::HashMap::Cursor

0.3.9

Fix a data race between scc::TreeIndex::iter and scc::TreeIndex::insert

0.3.8

Implement scc::TreeIndex::remove (work-in-progress), and fix a data race issue

0.3.7

Implement scc::TreeIndex::iter

0.3.6

Make scc::TreeIndex::insert work correctly on aarch64

0.3.5

Fix a data race issue with scc::TreeIndex::insert

0.3.4

Implement scc::TreeIndex::insert and scc::TreeIndex::read

0.3.2

Make scc::HashMap pass Valgrind by deallocating arrays immediately on drop

0.3.1

APIs stabilized