Tot

al
User Flow Inv
ocat
ion
Login 1

IST 1
CMS 1

Transfer to
own account 2

logout 1
Experience API URL

https://perffab02.ussaas.t24-jumpstart.com/authService/100000002/login?provider=DbxUserLogin

https://perffab02.ussaas.t24-jumpstart.com/services/IST
https://perffab02.ussaas.t24-jumpstart.com/services/CMS

https://perffab02.ussaas.t24-jumpstart.com/authService/100000002/login?provider=KonyBankingAdminConsoleIdentityServi

https://perffab02.ussaas.t24-jumpstart.com/authService/100000002/logout?provider=DbxUserLogin
Fabric DB RT Comments tables involved in fabric Infinity db

Quantumidconfigdb/ 512
fabricidconfigdbDB identity dbo. Session dbxdb

middleware_session if
quantummfreportsdb/ 662 something is wrong -
fabricmfreportsdb invalid_messages
662 integration middleware_request
402
integration +
(admin
Quantumkpnsdb/ engagement(sms
console)
notification - send
fabrickpnsdbDB 656
activation code) +
(activatio
identity
n code)
sms_request dbxdb
Quantumidconfigdb/ 62
fabricidconfigdbDB identity dbo. Session dbxdb

console pod (app import/publish)

Identity + integration
 Hyperscale standard
series (Gen 5) 2
DB Database tier vcores

Standard_D8s_v3

app node size (general purpose)

namespace cpu (request)

kony-fabric-apiportal-84b5d4f95b-fjhkq 300M
kony-fabric-console-557f4c95b4-hzkvp 500m

kony-fabric-engagement-745b86c8b5-x4f2p 1
kony-fabric-identity-6fc8bb9d94-9szmv 1500m
pod names kony-fabric-integration-76d5d795d5-rskpr 1200m
kubernets version 1.25.5
2200mb*65%=1430mb is the max heap of jvm
(8 vcpu, 32 GB, 16
Data disc, 12800
max IOPS, Temp
storage - 64 GB)
Memory
CPU Utilizatio
memory (request) memory (limit) utilization n Comments
1G 1400 M
4200M 4200 M

Initially memory allocated for

engagement pod is 2.2gb.
2.2gb*65%=1.4gb reserved to jvm

out of 1.4gb apprx. 75% would be

reserved by application for
maintaining its objects
1.4*75%=1.05gb

so only 350mb is available for

creating runtime objects while
processing the load. looking at the
above 350mb would be very small
to scale

so if we increase the ram to 4gb for

engagement pod, we will have

4gb*65%=2.6 for jvm heap

2.6-1.4= 1.2gb for runtime objects

creations (so more 3.5 times
memory). Hence we Increased to
4000 M 4000 M 2.18% 47% 4GB
3900M 3900 M 5% 83%
8G 8G 58% 60%
userflow API details Local time VM Jmeter

CPU
https://perffab02.ussaas.t24-jumpstart.com/
authService/100000002/login?
Login provider=DbxUserLogin 12.57 to 1.13 13%

DB
TPS
VM Jmeter APP pod metrics DB metrics Comments

Samples
Memory processed TPS RT CPU Memory cpu memory
No long running
queries or bottle
67% 1604 2 639 3% 57% 0.00% 21.23% necks during the run
Response time over time
 VM Jmeter
userflow API details Local time CPU Memory

admin console identity

service 2:25 pm to 2.40 pm 20% 85%
Transfer to own account

send activation code 3:05 pm to 3:20 pm 16% 85%

admin console identity

service
TPS
Integration

Engagement
DB
 APP pod metrics DB metrics
Samples processed TPS RT (95%) RT (avg) CPU Memory cpu

2413 2.7/SEC 396 238 22% 42% 0.11%

16% (INTEGRATION) 43%

1586 1.8/SEC 706 567 0.8% (engagement) 78% 1.50%

RT
DB metrics Comments
memory

No long running
21.23% queries or bottle
necks during the
run. SMS provider
not configured for
this environment
because of which we
are not seeing any
sms request
populating in fabric
engagement db
21.40%
userflow API details Local time VM Jmeter APP pod metrics
CPU Memory samples pr TPS RT CPU

3:35 pm
to 3:55
Logout https://perffab02.ussaas.t24-jumpstart.com/authService/100000002/logout?provider=DbxUserLogin
pm 41% 86% 15794 17.5/sec 68 3%

TPS
APP pod metrics DB metrics Comments
Memory cpu memory

No long running queries or bottle

57% 2.07% 24.29% necks during the run
userflow API details Local time VM Jmeter
CPU Memory samples processed TPS

https://
perffab02.ussaas.t24-
jumpstart.com/services/ 12.35 PM TO
IST IST 12.50 PM 96% 89% 40407 45

TPS

Integration server
DB
 APP pod metrics DB metrics Comments
RT CPU Memory cpu memory

No long
running
queries or
bottle
necks
during
21 13% 45% 2.00% 23.21% the run

RT
userflow API details Local time VM Jmeter APP pod metrics
CPU Memory samples processed TPS RT CPU

https://
perffab02
.ussaas.t2
4-
jumpstart
.com/ 1:14 PM
services/ to 1.24
CMS CMS PM 33% 79% 40406 67 20 15%

TPS

Integration server - CPU

DB
APP pod metrics DB metrics Comments
Memory cpu memory

No long running
queries or bottle
necks during the
45% 6.10% 24.00% run

RT
memory
userflow API details Local time VM Jmeter APP pod metrics
CPU Memory samples processed TPS RT CPU

https://
perffab02
.ussaas.t2
6:04 PM
4-
IST TO 6:19 45292 50 20
jumpstart
PM
.com/
services/
IST
80% 84% 15%
https://
perffab02
.ussaas.t2
6:00 PM
4-
CMS TO 6:15 45291 50 7
jumpstart
PM
.com/
services/
CMS

IST
CMS
DB
PP pod metrics DB metrics
Memory cpu memory Comments

1.No long running queries or

bottle necks during the run. 2.
45% 3.85% 26.28% Rather than file based system in
user parameters we are getting
better results
 userflow API details Local time VM Jmeter
CPU Memory samples processed TPS

Transfer to own 2:04 PM to 2:11

accounts PM 16% 85% 502 1

Integration - cpu
Engagement - cpu

Identity pod CPU

DB - CPU
 APP pod metrics DB metrics Comments
RT CPU Memory cpu memory

1. fabric db not
updating after 500
records 2.
Integration - 17% Integration - 45% Memory consumption
engagement - 0.5% Engagement - 94% of engagement pod
958 Identity - 0.06% Identity- 57% 1.75% 21.99% reaches 93%

memory
Memory

Memory
Memory