DB Bench

Attention! The goal of the tests was to discover and measure the specific use-cases. It's not an extensive (and accurate) benchmark.

Tests were performed on an empty database with default config using a computer containing 40 cores and 125G of RAM. The document and edge structure defined for ArangoDB:

type arangoArtifact struct {

	// Mandatory `key` field.
	Key string `json:"_key,omitempty"`

	// Other fields.
	Name        string    `json:"name"`
	Description string    `json:"description"`
	CreateTime  time.Time `json:"create_time"`
	Item        int       `json:"item"`
}

type arangoEdge struct {

	// Mandatory `key` field.
	Key string `json:"_key,omitempty"`

	// Edge fields.
	From string `json:"_from"`
	To   string `json:"_to"`

	// Other random fields.
	Body string `json:"body"`
}

The corresponding tables to simulate on using PostgreSQL:

CREATE TABLE IF NOT EXISTS artifacts
(
    id           UUID PRIMARY KEY,
    "name"       TEXT NOT NULL,
    description  TEXT,
    item         INTEGER DEFAULT 1,
    create_time  TIMESTAMP NOT NULL DEFAULT CLOCK_TIMESTAMP()
);

CREATE TABLE IF NOT EXISTS edges
(
    id      UUID PRIMARY KEY,
    "from"  UUID REFERENCES artifacts,
    "to"    UUID REFERENCES artifacts,
    body    TEXT
)

Num	Test	ArangoDB	PostgreSQL	Neo4j
1	Create 10 entries	10 ms	15 ms	195 ms
2	Create 100 entries	84 ms	41 ms	715 ms
3	Create 1000 entries	690 ms	334 ms	1.414 s
4	Create 1000 entries (bulk)	45 ms	93ms	91 ms
5	Create 10000 entries (bulk)	401 ms	4.2 s	63 ms
6	↪ Read 10000 entries (not a query)	10 s	skipped	skipped
7	↪ Read 10000 entries (not a query, bulk)	257 ms	skipped	skipped
8	↪ Update 10000 entries	12.3 s	84.4 s	64.1 s
9	↪ Update 10000 entries (bulk)	848 ms	4.9 s	25.6 s
10	↪ Read 10000 entries (using query)	328 ms	3 s	318 ms
11	Create 10 connected pairs	7 ms	9 ms	437 ms
12	Create 100 connected pairs	19 ms	43 ms	247 ms
13	Create 10000 connected pairs	1 s	68.9 s	13.7 s
14	↪ Query all neighbours in pair	554 ms	23 ms	394 ms
15	↪ Query all neighbours in pair (within one year)	40 ms	10 ms	77 ms
16	Create chain with 10000 artifacts	773 ms	23 s	N/A
17	↪ Query 10th artifact in chain	1 ms	15 ms	N/A
18	↪ Query 100th artifact in chain	2 ms	127 ms	N/A
19	↪ Query 1000th artifact in chain	316 ms	1.1 s	N/A
20	↪ Query 2000th artifact in chain	2.5 s	2.2 s	N/A
21	↪ Query 5000th artifact in chain	40.7 s	5.7 s	N/A
22	↪ Query 7000th artifact in chain	110.7 s	7.8 s	N/A
23	↪ Sum 5000 `item`s in chain	39.9 s	5.6 s	N/A
24	Create 100 direct neighbours	18 ms	27 ms	N/A
25	Create 1000 direct neighbours	94 ms	328 ms	N/A
26	Create 10000 direct neighbours	785 ms	28.9 s	N/A
27	↪ Query all neighbours (sorted by name)	354 ms	45 ms	N/A

The following tests were performed on a pre-populated database of 1 million entries:

Num	Test	ArangoDB	PostgreSQL
1	Create 10 entries	9 ms	89 ms
2	Create 100 entries	77 ms	81 ms
3	Create 1000 entries	805 ms	417 ms
4	Create 1000 entries (bulk)	49 ms	133 ms
5	Create 10000 entries (bulk)	421 ms	4.1 s
6	↪ Read 10000 entries (not a query)	9.5 s	skipped
7	↪ Read 10000 entries (not a query, bulk)	266 ms	skipped
8	↪ Update 10000 entries	12.2 s	90 s
9	↪ Update 10000 entries (bulk)	983 ms	4.8 s
10	↪ Read 10000 entries (using query)	331 ms	3 s
11	Create 10 connected pairs	5 ms	45 ms
12	Create 100 connected pairs	16 ms	72 ms
13	Create 10000 connected pairs	1 s	50.2 s
14	↪ Query all neighbours in pair	4.1 s	51 ms
15	↪ Query all neighbours in pair (within one year)	4.4 s	411 ms
16	Create chain with 10000 artifacts	770 ms	22.9 s
17	↪ Query 10th artifact in chain	1 ms	12 ms
18	↪ Query 100th artifact in chain	2 ms	115 ms
19	↪ Query 1000th artifact in chain	387 ms	1.1 s
20	↪ Query 2000th artifact in chain	2.5 s	2.2 s
21	↪ Query 5000th artifact in chain	39.2 s	5.9 s
22	↪ Query 7000th artifact in chain	108.1 s	8.1 s
23	↪ Sum 5000 `item`s in chain	40.1 s	5.9 s
24	Create 100 direct neighbours	19 ms	53 ms
25	Create 1000 direct neighbours	96 ms	327 ms
26	Create 10000 direct neighbours	825 ms	21.5 s
27	↪ Query all neighbours (sorted by name)	338 ms	71 ms

Tests indented by ↪ depend on previous not-indented test.

Pair


|````|     |````|
| N1 | --> | N2 |
|____|     |____|

N  = Node (document/artifact).
N2 = Neighbour node to N1.

Chain


|````|     |````|           |````|
| N1 | --> | N2 | -> ... -> | Nn |
|____|     |____|           |____|

N  = Node (document/artifact).
Nn = Nth node in a chain.

Direct neighbours


|````|     |````|     |````|
| N2 | <-- | N1 | --> | N3 |
|____|     |____|     |____|
             |
             v
           |````|
           | Nn |
           |____|

N  = Node (document/artifact).
Nn = Nth direct neighbour.

geomodular / db-bench Goto Github PK

db-bench's Introduction

DB Bench

Pair

Chain

Direct neighbours

db-bench's People

Contributors

Watchers

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent