ЕР
Хост в Амстердаме
Size: a a a
2020 July 01
D
лол
ЕР
1,017,744,506,248 bytes allocated in the heap
13,707,342,344 bytes copied during GC
136,441,144 bytes maximum residency (139 sample(s))
2,397,400 bytes maximum slop
130 MB total memory in use (0 MB lost due to fragmentation)Это, кажется, даже в несколько раз больше чем у меня было аллокаций
YS
Ага
YS
4 раза
MK
очень странно
MK
там нет санков, вроде
MK
но я запускал последний раз без профайлинга памяти
YS
В том не было санков толком, да
D
While this library helps in ensuring that only a limited/pre-defined number of actions are evaluated in parallel, it still has one problem (especially with very large input data-sets). If the input data-set has N=2,000,000, this is going to create 2,000,000 asyncs, although 99% of them might not be getting concurrently evaluated. This still results in linear memory growth.
ЕР
Вот с выводом каждого результата в терминал.
Это приложение k6.io , раниться так же на моей машине, рпс в среднем вышел 4к
checks.....................: 99.97% ✓ 249935 ✗ 65В самом конце, правда, сервер действительно подтупил, и 65 запросов из 250к зафейлилось, и кажется в начале рпс был побольше чем к концу
data_received..............: 53 MB 853 kB/s
data_sent..................: 20 MB 324 kB/s
http_req_blocked...........: avg=15.43ms min=0s med=2.36µs max=15.51s p(90)=4.37µs p(95)=5.38µs
http_req_connecting........: avg=15.43ms min=0s med=0s max=15.51s p(90)=0s p(95)=0s
http_req_duration..........: avg=49ms min=0s med=44.86ms max=1.12s p(90)=50.79ms p(95)=59.23ms
http_req_receiving.........: avg=41.94µs min=0s med=32.03µs max=32.51ms p(90)=58.38µs p(95)=77.2µs
http_req_sending...........: avg=15.27µs min=0s med=9.88µs max=16.16ms p(90)=19.07µs p(95)=22.81µs
http_req_tls_handshaking...: avg=0s min=0s med=0s max=0s p(90)=0s p(95)=0s
http_req_waiting...........: avg=48.94ms min=0s med=44.8ms max=1.12s p(90)=50.71ms p(95)=59.15ms
http_reqs..................: 250000 4045.565372/s
iteration_duration.........: avg=69.72ms min=4.24ms med=45.09ms max=30s p(90)=51.5ms p(95)=61.3ms
iterations.................: 250000 4045.565372/s
vus........................: 350 min=350 max=350
vus_max....................: 350 min=350 max=350
Это приложение k6.io , раниться так же на моей машине, рпс в среднем вышел 4к
D
YS
А там не для пула мэп?
MK
для пула
MK
у него нет версии с
_MK
While this library helps in ensuring that only a limited/pre-defined number of actions are evaluated in parallel, it still has one problem (especially with very large input data-sets). If the input data-set has N=2,000,000, this is going to create 2,000,000 asyncs, although 99% of them might not be getting concurrently evaluated. This still results in linear memory growth.
👍