Etc

[JMeter] JMeter ์„ค์ • ๊ฐ’, ํ†ต๊ณ„ ๊ฐ’

Ella_K 2022. 11. 24. 17:04

โœ… Thread-Group ์„ค์ • ๊ฐ’

๐Ÿ“Œ Number of Threads(users)

  • Ramp-up Period ๋™์•ˆ ์‹คํ–‰ํ•  user์ˆ˜

 

๐Ÿ“Œ Ramp-up Period(in seconds)

  • ํ•œ๋ฒˆ์˜ ์‹คํ–‰์„ ๋ช‡์ดˆ ๋™์•ˆ ์™„๋ฃŒ ์‹œํ‚ฌ๊ฒƒ์ธ์ง€
Number of Treads๊ฐ€ 5์ด๊ณ , Ramp-up Period๊ฐ€ 1์ด๋ผ๋ฉด, 1์ดˆ๋™์•ˆ 5๋ฒˆ์˜ ์š”์ฒญ์„ ํ•œ๋‹ค.

 

๐Ÿ“Œ Loop Count 

  • ๋ฐ˜๋ณตํ•˜๊ณ ์ž ํ•˜๋Š” ํšŸ์ˆ˜. Infinite๋ฅผ ์ฒดํฌํ•  ๊ฒฝ์šฐ ๋ฌด์ œํ•œ์œผ๋กœ ์‹คํ–‰

 

โœ… Report ํ†ต๊ณ„ ๊ฐ’

๐Ÿ“Œ Label

  • sample๋“ค์˜ label, ์ด๋ฆ„

 

๐Ÿ“Œ # Samples

  • ์‹คํ–‰๋˜๋Š” sample๋“ค์˜ ์ˆ˜
#Samples = Threads(users)์˜ ์ˆ˜ × Loop Count

 

๐Ÿ“Œ Average(millisecond)

  • ๊ฒฐ๊ณผ ์‘๋‹ต ์‹œ๊ฐ„์˜ ํ‰๊ท 
์‘๋‹ต ์‹œ๊ฐ„์€ ์š”์ฒญ์ด ์ „์†ก๋œ ์‹œ์ ๋ถ€ํ„ฐ ํ•ด๋‹น ์š”์ฒญ์˜ ๋งˆ์ง€๋ง‰ ๋ฐ”์ดํŠธ๊ฐ€ ์„œ๋ฒ„์—์„œ ์ˆ˜์‹ ๋  ๋•Œ๊นŒ์ง€ ๊ณ„์‚ฐ๋œ๋‹ค. 
๋”ฐ๋ผ์„œ ๋ณด๋‚ด๊ธฐ ์ „์— ์š”์ฒญ์„ ์ฒ˜๋ฆฌํ•˜๋Š” ์‹œ๊ฐ„(Pre-Processor) ๋˜๋Š” ์ˆ˜์‹  ํ›„ ๋ฐ์ดํ„ฐ๋ฅผ ์ถ”์ถœ/์ฒ˜๋ฆฌํ•˜๋Š” ์‹œ๊ฐ„(Post-Processor)์ด ํฌํ•จ๋ฉ๋‹ˆ๋‹ค.

 

๐Ÿ“Œ Min(millisecond)

  • ๊ฐ™์€ label์˜ sample๋“ค์—์„œ ๊ฐ€์žฅ ์งง์€ ์‘๋‹ต ์‹œ๊ฐ„. ์ตœ์†Œ ์‘๋‹ต ์‹œ๊ฐ„

 

๐Ÿ“Œ Max(millisecond)

  • ๊ฐ™์€ label์˜ sample๋“ค์—์„œ ๊ฐ€์žฅ ๊ธด ์‘๋‹ต ์‹œ๊ฐ„. ์ตœ๋Œ€ ์‘๋‹ต ์‹œ๊ฐ„

 

๐Ÿ“Œ Percentiles(millisecond)

๋ฐฑ๋ถ„์œจ์—์„œ ํ•ด๋‹น ๊ฐ’ ์•„๋ž˜๋กœ ๋–จ์–ด์ง„๋Š” ๊ฐ’์„ ๋‚˜ํƒ€๋‚ด๋Š” ์ธก์ • ๊ฐ’

20th percentile์€ ๊ด€์ฐฐ ๊ทธ๋ฃน(sample๋“ค)์˜ 20%๊ฐ€ ํ•ด๋‹น ๊ฐ’ ์•„๋ž˜์˜ ๊ฐ’์„ ๊ฐ–๋Š”๋‹ค.

Median

  • ๊ฒฐ๊ณผ ์ง‘ํ•ฉ์—์„œ ์ค‘๊ฐ„์— ์žˆ๋Š” ๊ฐ’
  • sample๋“ค์˜ 50%๋Š” ํ•ด๋‹น ๊ฐ’๋ณด๋‹ค ๋‚ฎ์€ ์‘๋‹ต ์‹œ๊ฐ„์„ ๊ฐ€์ง€๊ณ , 50%๋Š” ํ•ด๋‹น ๊ฐ’๋ณด๋‹ค ๋†’์€ ์‘๋‹ต ์‹œ๊ฐ„์„ ๊ฐ–๋Š”๋‹ค.

90% Line

  • sample๋“ค์˜ 90%๋Š” ํ•ด๋‹น ๊ฐ’๋ณด๋‹ค ๋‚ฎ์€ ์‘๋‹ต ์‹œ๊ฐ„์„ ๊ฐ€์ง€๊ณ , 10%๋Š” ํ•ด๋‹น ๊ฐ’๋ณด๋‹ค ๋†’์€ ์‘๋‹ต ์‹œ๊ฐ„์„ ๊ฐ–๋Š”๋‹ค.

95% Line

  • sample๋“ค์˜ 95%๋Š” ํ•ด๋‹น ๊ฐ’๋ณด๋‹ค ๋‚ฎ์€ ์‘๋‹ต ์‹œ๊ฐ„์„ ๊ฐ€์ง€๊ณ , 5%๋Š” ํ•ด๋‹น ๊ฐ’๋ณด๋‹ค ๋†’์€ ์‘๋‹ต ์‹œ๊ฐ„์„ ๊ฐ–๋Š”๋‹ค.

99% Line

  • sample๋“ค์˜ 99%๋Š” ํ•ด๋‹น ๊ฐ’๋ณด๋‹ค ๋‚ฎ์€ ์‘๋‹ต ์‹œ๊ฐ„์„ ๊ฐ€์ง€๊ณ , 1%๋Š” ํ•ด๋‹น ๊ฐ’๋ณด๋‹ค ๋†’์€ ์‘๋‹ต ์‹œ๊ฐ„์„ ๊ฐ–๋Š”๋‹ค.
percentile values๋Š” jmeter.properties ํŒŒ์ผ์—์„œ ์ˆ˜์ •ํ•  ์ˆ˜ ์žˆ๋‹ค. ํ•˜์ง€๋งŒ 3๊ฐœ๊นŒ์ง€๋งŒ ์„ค์ •ํ•  ์ˆ˜ ์žˆ๋‹ค.
#---------------------------------------------------------------------------
# Aggregate Report and Aggregate Graph - configuration
#---------------------------------------------------------------------------
#
# Percentiles to display in reports
# Can be float value between 0 and 100
# First percentile to display, defaults to 90%
#aggregate_rpt_pct1=90
# Second percentile to display, defaults to 95%
#aggregate_rpt_pct2=95
# Second percentile to display, defaults to 99%
#aggregate_rpt_pct3=99

 

๐Ÿ“Œ Error %

  • ์—๋Ÿฌ๊ฐ€ ๋ฐœ์ƒํ•œ ์š”์ฒญ๋“ค์˜ percent 

 

๐Ÿ“Œ Throughput

  • ์ „์ฒด ์š”์ฒญ์ˆ˜๋ฅผ ์ „์ฒด ์‹œ๊ฐ„์œผ๋กœ ๋‚˜๋ˆˆ ๊ฐ’. ๋‹จ์œ„ ์‹œ๊ฐ„ ๋‹น ์ฒ˜๋ฆฌ๋Ÿ‰
Throughput = (number of requests) / (total time) * conversion
  • number of requests: ์‹คํ–‰๋œ ์š”์ฒญ๋“ค์˜ ์ˆ˜
  • total time: (lastSampleStartTime + lastSampleLoadTime – firstSampleStartTime)
  • conversion: ์‹œ๊ฐ„ ๋‹จ์œ„์— ๋”ฐ๋ฅธ conversion value์˜ ๋‹จ์œ„ 

 

๐Ÿ“Œ KB/sec

  • ์ดˆ๋‹น Kilobytes๋กœ ์ธก์ •๋œ ์ฒ˜๋ฆฌ๋Ÿ‰
Throughput KB/sec = (Throughput * Average Bytes) / 1024
  • Average Bytes: sample ์‘๋‹ต์˜ ํ‰๊ท  ํฌ๊ธฐ (bytes๋‹จ์œ„)

 

๐Ÿ“Œ Total

  • ๋ณด๊ณ ์„œ ๋งˆ์ง€๋ง‰ ํ–‰
  • #Samples๋Š” ๋ผ๋ฒจ๋“ค์˜ #Samples์˜ ํ•ฉ๊ณ„
  • ๋‚˜๋จธ์ง€๋Š” ๋ผ๋ฒจ๋“ค์˜ ๊ฐ ํ†ต๊ณ„๊ฐ’์˜ ํ‰๊ท 

Source

https://jmetervn.com/2016/10/10/analyze-the-aggregate-report-in-jmeter/

 

Analyze the Aggregate Report in JMeter

One of difficulty in Performance Testing is how to read and understanding the report. How to give the accurate conclude with some statistics from the report? In JMeter, we have many kinds of the re…

jmetervn.com

 

'Etc' ์นดํ…Œ๊ณ ๋ฆฌ์˜ ๋‹ค๋ฅธ ๊ธ€

๋ชฐ์ž… ๊ฒฝํ—˜์„ ํ†ตํ•ด ์„ฑ์žฅํ•˜๊ธฐ  (0) 2022.08.12