Memory and speed benchmarks: kysely vs kysely + kysely-params
Results may vary depending on environment, so please share yours, and I'll update readme.
It takes about an hour to complete, so be patient.
(It's hard to measure memory usage in node due to GC)
Reduce the *_ROUNDS in src/main.ts if you want to see the result quickly.
See Diagram to check how it works.
Average of memory usage in kilobyte
| query | kysely | kysely-params |
|---|---|---|
| 0 | 0.99 | 0.23 |
| 1 | 2.05 | 0.48 |
| 2 | 14.64 | 0.38 |
| 3 | 12.25 | 0.39 |
| 4 | 10.92 | 0.56 |
Average of time spend in milliseconds
| query | kysely | kysely-params |
|---|---|---|
| 0 | 0.001 | 0.000 |
| 1 | 0.002 | 0.000 |
| 2 | 0.013 | 0.000 |
| 3 | 0.026 | 0.000 |
| 4 | 0.038 | 0.000 |
os linux Ubuntu 22.04.2 LTS x64
cpu Intel Gen Intel® Core™ i7-12700K
memory 46.85GB
WARMUP_ROUNDS 4000
QUERY_ROUND 1000
TEST_ROUNDS 10000
no-op
-----------------------------------------------
min 0.00kb
max 0.22kb
avg 0.00kb
min 0.000ms
max 0.000ms
avg 0.000ms
kysely-0
-----------------------------------------------
min 0.92kb
max 1.39kb
avg 0.99kb
min 0.001ms
max 0.002ms
avg 0.001ms
kysely-params-0
-----------------------------------------------
min 0.23kb
max 0.49kb
avg 0.23kb
min 0.000ms
max 0.000ms
avg 0.000ms
kysely-1
-----------------------------------------------
min 1.87kb
max 2.36kb
avg 2.05kb
min 0.002ms
max 0.005ms
avg 0.002ms
kysely-params-1
-----------------------------------------------
min 0.25kb
max 0.75kb
avg 0.48kb
min 0.000ms
max 0.000ms
avg 0.000ms
kysely-2
-----------------------------------------------
min 14.40kb
max 14.93kb
avg 14.64kb
min 0.012ms
max 0.018ms
avg 0.013ms
kysely-params-2
-----------------------------------------------
min 0.27kb
max 0.52kb
avg 0.38kb
min 0.000ms
max 0.000ms
avg 0.000ms
kysely-3
-----------------------------------------------
min 11.89kb
max 12.52kb
avg 12.25kb
min 0.024ms
max 0.043ms
avg 0.026ms
kysely-params-3
-----------------------------------------------
min 0.28kb
max 0.53kb
avg 0.39kb
min 0.000ms
max 0.000ms
avg 0.000ms
kysely-4
-----------------------------------------------
min 10.42kb
max 11.13kb
avg 10.92kb
min 0.036ms
max 0.063ms
avg 0.038ms
kysely-params-4
-----------------------------------------------
min 0.41kb
max 0.66kb
avg 0.56kb
min 0.000ms
max 0.001ms
avg 0.000ms
Done in 2739.24s.
-
setup node_modules
yarn installornpm install -
start
yarn startornpm start
- To see which query is used, pass
-q:
yarn start -qornpm start -q
query-0
-----------------------------------------------
select
*
from
"person"
query-1
-----------------------------------------------
select
*
from
"person"
where
"id" = $1
query-2
-----------------------------------------------
select
"user_following"."id",
"user_following"."from_user_id",
"user_following"."to_user_id",
"user_following"."accepted",
"user_profile"."nickname",
"user_profile"."search_id",
"user_profile"."profile_image_url"
from
"user_following"
inner join "user_handle" on "user_handle"."id" = "user_following"."to_user_id"
inner join "user_profile" on "user_profile"."user_id" = "user_handle"."id"
where
"user_handle"."type" = $1
and "user_following"."to_user_id" = $2
order by
"user_following"."accepted" asc,
"user_following"."id" desc
query-3
-----------------------------------------------
select
"t1"."Trade_Stock",
"t1"."TRADE_ID",
"t2"."TRADE_ID",
"t1"."Trade_Timestamp",
"t2"."Trade_Timestamp",
"t1"."Price",
"t2"."Price",
abs("t1"."Price" - "t2"."Price") * 1.0 / "t1"."Price" * $1 as "price_diff_pct"
from
"Trade_tbl" as "t1"
inner join "Trade_tbl" as "t2" on "t1"."Trade_Stock" = "t2"."Trade_Stock"
where
(
"t1"."Trade_Timestamp" < "t2"."Trade_Timestamp"
and datediff (
second,
"t1"."Trade_Timestamp",
"t2"."Trade_Timestamp"
) < $2
and abs("t1"."Price" - "t2"."Price") * 1.0 / "t1"."Price" * $3 > $4
)
order by
"t1"."TRADE_ID"
query-4
-----------------------------------------------
select
"user"."id",
"user"."first_name",
"user"."last_name",
"user"."age",
"marriage"."married",
"marriage"."married_at",
"pet"."name",
"pet"."kind",
"job"."kind",
"employment"."employed_at",
"company"."name",
"company"."created_at",
"country"."id",
"country"."code",
"country"."population",
"country"."population_updated_at",
"planet"."id",
"planet"."kind"
from
"user"
inner join "employment" on "employment"."user_id" = "user"."id"
inner join "job" on "job"."id" = "employment"."job_id"
inner join "company" on "company"."id" = "employment"."company_id"
inner join "country" on "country"."id" = "company"."country_id"
inner join "planet" on "planet"."id" = "country"."planet_id"
left join "marriage" on "marriage"."user_id" = "user"."id"
left join "pet" on "pet"."owner_id" = "user"."id"
where
(
"user"."name" like $1
and "user"."age" >= $2
and "user"."age" <= $3
and "marriage"."married" = $4
and "pet"."kind" = $5
and "job"."kind" = $6
and "country"."population" >= $7
and "planet"."kind" = $8
)
order by
"marriage"."married" desc,
"country"."population" asc
stateDiagram-v2
[*] --> _build_query
_build_query --> _build_query : (WARMUP_ROUNDS)
_build_query --> gc
gc --> record_before
record_before --> build_query
build_query --> build_query : (QUERY_ROUNDS)
build_query --> record_after
record_after --> save_diff
save_diff --> gc : (TEST_ROUNDS)
save_diff --> print_results : finish