Closed andrewsignori-aot closed 1 week ago
Below is an article with some additional information on the default Nodejs max memory.
The same article has the suggested fix.
Source: https://medium.com/geekculture/node-js-default-memory-settings-3c0fe8a9ba1
Other references:
/\ |‾‾| /‾‾/ /‾‾/
/\ / \ | |/ / / /
/ \/ \ | ( / ‾‾\
/ \ | |\ \ | (‾) |
/ __________ \ |__| \__\ \_____/ .io
execution: local
script: ./dist/institution-get-profile.test.js
output: -
scenarios: (100.00%) 1 scenario, 240 max VUs, 11m0s max duration (incl. graceful stop):
* default: Up to 240 looping VUs for 10m30s over 2 stages (gracefulRampDown: 30s, gracefulStop: 30s)
✓ Retrieved with success
█ setup
checks.........................: 100.00% ✓ 129514 ✗ 0
data_received..................: 281 MB 445 kB/s
data_sent......................: 217 MB 343 kB/s
http_req_blocked...............: avg=3.01ms min=0s med=0s max=1.04s p(90)=0s p(95)=33.35ms
http_req_connecting............: avg=1.3ms min=0s med=0s max=1.03s p(90)=0s p(95)=14.4ms
http_req_duration..............: avg=132.35ms min=15.68ms med=25.69ms max=3.49s p(90)=223.76ms p(95)=892.07ms
{ expected_response:true }...: avg=132.35ms min=15.68ms med=25.69ms max=3.49s p(90)=223.76ms p(95)=892.07ms
http_req_failed................: 0.00% ✓ 0 ✗ 130707
http_req_receiving.............: avg=118.88µs min=0s med=0s max=7.52ms p(90)=530.9µs p(95)=807.73µs
http_req_sending...............: avg=9.19µs min=0s med=0s max=51.71ms p(90)=0s p(95)=0s
http_req_tls_handshaking.......: avg=1.7ms min=0s med=0s max=346.35ms p(90)=0s p(95)=18.28ms
http_req_waiting...............: avg=132.22ms min=15.46ms med=25.56ms max=3.49s p(90)=223.24ms p(95)=892.04ms
http_reqs......................: 130707 207.101887/s
iteration_duration.............: avg=1.14s min=0s med=1.03s max=4.63s p(90)=1.26s p(95)=1.89s
iterations.....................: 129514 205.211609/s
vus............................: 16 min=8 max=240
vus_max........................: 240 min=240 max=240
running (10m31.1s), 000/240 VUs, 129514 complete and 0 interrupted iterations
default ✓ [======================================] 000/240 VUs 10m30s
Parallelism allowed for bulk insert: 1 Liveness Probe: 3000 Readiness Probed: 1000 CPU Limit: 0.2 CPU Requested: 0.1 Memory Limit: 768M Memory Requested: 512M max-old-space-size=550 Process started at 12:46:53 PM Finished at 12:53:47 PM (7 min)
Parallelism allowed for bulk insert: 4 Liveness Probe: 3000 Readiness Probed: 1000 CPU Limit: 0.2 CPU Requested: 0.1 Memory Limit: 768M Memory Requested: 512M max-old-space-size=550 Process started at 2:16:07 PM Finished at 2:22:06 PM (6 min)
Parallelism allowed for bulk insert: 4 Liveness Probe: 3000 Readiness Probed: 1000 CPU Limit: 0.3 CPU Requested: 0.2 Memory Limit: 1024M Memory Requested: 768M max-old-space-size=700 Process started at 2:51:19 PM Finished at 2:56:07 PM (5min)
Parallelism allowed for bulk insert: 4 Liveness Probe: 3000 Readiness Probed: 2500 CPU Limit: 0.2 CPU Requested: 0.1 Memory Limit: 768M Memory Requested: 512M max-old-space-size=500 Process started at 4:40:18 PM Finished at 4:46:41 PM (6min) maxStalledCount: 0, lockDuration: 60000, lockRenewTime: 5000,
The SIMS Api demonstrated some degradation on TEST where a preliminary analysis indicated that it may be caused by memory being consumed and not well handled leading the POD to respond with a major delay causing timeouts in the Web application request (e.g. related ticket #3224).
Technical
SIMS-Api
,queue-consumers
, andworkers
.HorizontalPodAutoscaler
memory forSIMS-Api
,queue-consumers
, andworkers
.AverageValue
from 230M to 350M--max-old-space-size=250
to all prod scripts.Additional context
Current memory consumption on PROD on May 15th, 2024.
Current SIMS-Api memory on PROD on May 15th, 2024.