Slide 1
Slide 1 text
Holly Cummins
Senior Principal Software Engineer, Quarkus
Code Europe
June 10, 2024
@[email protected]
faster
greener
happier
why quarkus should be
your next tech stack
Slide 2
Slide 2 text
@holly_cummins
#RedHat
Slide 3
Slide 3 text
@holly_cummins
#RedHat
Slide 4
Slide 4 text
@holly_cummins
#RedHat
software development cycle
Slide 5
Slide 5 text
@holly_cummins
#RedHat
software development cycle
Slide 6
Slide 6 text
@holly_cummins
#RedHat
software development cycle
Slide 7
Slide 7 text
@holly_cummins
#RedHat
software development cycle
Slide 8
Slide 8 text
@holly_cummins
#RedHat
software development cycle
Slide 9
Slide 9 text
@holly_cummins
#RedHat
software development cycle
Slide 10
Slide 10 text
@holly_cummins
#RedHat
software development cycle
Slide 11
Slide 11 text
@holly_cummins
#RedHat
software development cycle
Slide 12
Slide 12 text
@holly_cummins
#RedHat
software development cycle
Slide 13
Slide 13 text
@holly_cummins
#RedHat
Slide 14
Slide 14 text
@holly_cummins
#RedHat
the problem:
Slide 15
Slide 15 text
@holly_cummins
#RedHat
tea: slow feedback loop→developers drink a lot of tea
the problem:
Slide 16
Slide 16 text
@holly_cummins
#RedHat
2010s-era java
web frameworks
are verbose
there is a lot of typing
Slide 17
Slide 17 text
@holly_cummins
#RedHat
tea: slow feedback loop→developers drink a lot of tea
the problems:
Slide 18
Slide 18 text
@holly_cummins
#RedHat
tea: slow feedback loop→developers drink a lot of tea
the problems:
typing: verbose programming model → repetitive boilerplate
Slide 19
Slide 19 text
@holly_cummins
#RedHat
tea: slow feedback loop→developers drink a lot of tea
the problems:
typing: verbose programming model → repetitive boilerplate
tedium
Slide 20
Slide 20 text
@holly_cummins #RedHat
Slide 21
Slide 21 text
@holly_cummins #RedHat
“I can’t bring up the
microservices in my Java dev
stack
Slide 22
Slide 22 text
@holly_cummins #RedHat
“I can’t bring up the
microservices in my Java dev
stack
… on my brand new Apple
laptop with a M1 chip and
64GB of RAM.”
- fintech CTO
Slide 23
Slide 23 text
@holly_cummins #RedHat
in production, it’s worse
Slide 24
Slide 24 text
@holly_cummins #RedHat
in production, it’s worse
Slide 25
Slide 25 text
@holly_cummins #RedHat
in production, it’s worse
example microservices maths:
Slide 26
Slide 26 text
@holly_cummins #RedHat
in production, it’s worse
example microservices maths:
airline maintenance scheduling system
Slide 27
Slide 27 text
@holly_cummins #RedHat
in production, it’s worse
example microservices maths:
airline maintenance scheduling system
single service: ½ core + 1 GB RAM
Slide 28
Slide 28 text
@holly_cummins #RedHat
in production, it’s worse
example microservices maths:
airline maintenance scheduling system
single service: ½ core + 1 GB RAM
HA → 3x instances
Slide 29
Slide 29 text
@holly_cummins #RedHat
in production, it’s worse
example microservices maths:
airline maintenance scheduling system
single service: ½ core + 1 GB RAM
HA → 3x instances
~100 microservices
Slide 30
Slide 30 text
@holly_cummins #RedHat
in production, it’s worse
example microservices maths:
airline maintenance scheduling system
single service: ½ core + 1 GB RAM
HA → 3x instances
~100 microservices
= 150 cores + 300 GB RAM
Slide 31
Slide 31 text
@holly_cummins #RedHat
all this
resource usage
is expensive
Slide 32
Slide 32 text
@holly_cummins
#RedHat
tea: slow feedback loop→developers drink a lot of tea
the problems:
typing: verbose programming model → repetitive boilerplate
Slide 33
Slide 33 text
@holly_cummins
#RedHat
tea: slow feedback loop→developers drink a lot of tea
the problems:
typing: verbose programming model → repetitive boilerplate
tolls: high resource usage → paying cloud vendors lots of money
Slide 34
Slide 34 text
@holly_cummins
#RedHat
software development cycle
c’mon, what about the rest of the world?
Slide 35
Slide 35 text
@holly_cummins
#RedHat
software development cycle
c’mon, what about the rest of the world?
Slide 36
Slide 36 text
@holly_cummins
#RedHat
c’mon, what about the rest of the world? coffee, not tea
Slide 37
Slide 37 text
@holly_cummins
#RedHat
the problems:
c’mon, what about the rest of the world?
internationalised version
coffee, not tea
Slide 38
Slide 38 text
@holly_cummins
#RedHat
coffee: slow feedback loop→developers drink too much coffee
the problems:
c’mon, what about the rest of the world?
internationalised version
coffee, not tea
Slide 39
Slide 39 text
@holly_cummins
#RedHat
coffee: slow feedback loop→developers drink too much coffee
the problems:
copious code: verbose programming model → repetitive boilerplate
c’mon, what about the rest of the world?
internationalised version
coffee, not tea
Slide 40
Slide 40 text
@holly_cummins
#RedHat
coffee: slow feedback loop→developers drink too much coffee
the problems:
copious code: verbose programming model → repetitive boilerplate
cost: high resource usage → paying cloud vendors lots of money
c’mon, what about the rest of the world?
internationalised version
coffee, not tea
Slide 41
Slide 41 text
@holly_cummins #RedHat
why is this happening?
mismatch between what we need
and what the platform is optimised for
Slide 42
Slide 42 text
@holly_cummins
#RedHat
java was not designed
for the cloud
Slide 43
Slide 43 text
@holly_cummins
#RedHat
old Java frameworks were
optimised for …
Slide 44
Slide 44 text
@holly_cummins
#RedHat
long-lived processes
old Java frameworks were
optimised for …
Slide 45
Slide 45 text
@holly_cummins
#RedHat
long-lived processes
annual (!) deployments
old Java frameworks were
optimised for …
Slide 46
Slide 46 text
@holly_cummins
#RedHat
long-lived processes
annual (!) deployments
late-binding
old Java frameworks were
optimised for …
Slide 47
Slide 47 text
@holly_cummins
#RedHat
long-lived processes
annual (!) deployments
late-binding
re-configurable without restart
old Java frameworks were
optimised for …
Slide 48
Slide 48 text
@holly_cummins #RedHat
Container platform
machine
go go
go go
go go
go go
go go
go go
go go
Slide 49
Slide 49 text
@holly_cummins #RedHat
Container platform
machine
node.js
node.js
node.js
node.js
node.js
node.js
node.js
machine
go go
go go
go go
go go
go go
go go
go go
Slide 50
Slide 50 text
@holly_cummins #RedHat
Container platform
machine
HotSpot Heap
HotSpot Heap
HotSpot Heap
HotSpot Heap
machine
node.js
node.js
node.js
node.js
node.js
node.js
node.js
machine
go go
go go
go go
go go
go go
go go
go go
Slide 51
Slide 51 text
can we do better than that?
Slide 52
Slide 52 text
No content
Slide 53
Slide 53 text
enter … quarkus
a Java framework that gets you going faster, faster.
Slide 54
Slide 54 text
@holly_cummins #RedHat
traditional cloud-native
java stack
traditional cloud-native
java stack
traditional cloud-native
java stack
traditional cloud-native
java stack
node.js
node.js
node.js
node.js
node.js
node.js
node.js go
go
machine
go go go
go go go
go go go
go go go
go go go
go go
go go
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
spoiler: we made stuff better :)
container orchestration
machine machine machine
https:/
/developers.redhat.com/blog/2017/03/14/java-inside-docker/
Slide 55
Slide 55 text
@holly_cummins #RedHat
machine
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
container orchestration
machine
traditional cloud-native java stack
traditional cloud-native java stack
traditional cloud-native java stack
traditional cloud-native java stack
… a lot better quarkus native
Slide 56
Slide 56 text
@holly_cummins #RedHat
machine
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
quarkus
container orchestration
machine
traditional cloud-native java stack
traditional cloud-native java stack
traditional cloud-native java stack
traditional cloud-native java stack
… a lot better quarkus native
(but quarkus on jvm
is also way smaller
than traditional java)
Slide 57
Slide 57 text
quarkus applications start fast
quarkus + graalvm 0.014 Seconds
rest application
quarkus + open jdk 0.75 Seconds
traditional cloud-native stack 4.3 Seconds
https://quarkus.io/blog/runtime-performance/
Slide 58
Slide 58 text
@holly_cummins
to the code!
demo!
Slide 59
Slide 59 text
@holly_cummins #RedHat
mvn quarkus:build -Pnative
target/code-with-quarkus-1.0.0-SNAPSHOT-runner
Slide 60
Slide 60 text
faster than a lightbulb
Slide 61
Slide 61 text
ok but does startup time matter?
Slide 62
Slide 62 text
ok but does startup time matter?
fast boot time means auto-scaling works better
more resilience to load spikes
Slide 63
Slide 63 text
quarkus + GraalVM
13 MB
quarkus + OpenJDK
74 MB
Traditional Cloud-Native Stack
140 MB
rest application
https://quarkus.io/blog/runtime-performance/
quarkus improves memory utilization
Slide 64
Slide 64 text
#Quarkus @holly_cummins
Slide 65
Slide 65 text
#Quarkus @holly_cummins
hey, wanna see
quarkus?
hey, wanna see
quarkus?
hey, wanna see
quarkus?
hey, wanna see
quarkus?
hey, wanna see
quarkus?
Slide 66
Slide 66 text
#Quarkus @holly_cummins
hey, wanna see
quarkus?
hey, wanna see
quarkus?
hey, wanna see
quarkus?
hey, wanna see
quarkus?
hey, wanna see
quarkus?
uhh … are you supposed
to shut down applications
after using them?
Slide 67
Slide 67 text
#Quarkus @holly_cummins
hey, wanna see
quarkus?
hey, wanna see
quarkus?
hey, wanna see
quarkus?
hey, wanna see
quarkus?
hey, wanna see
quarkus?
uhh … are you supposed
to shut down applications
after using them? 120 instances (!)
Slide 68
Slide 68 text
ok but does memory footprint matter?
Slide 69
Slide 69 text
ok but does memory footprint matter?
in the cloud, memory footprint is money
Slide 70
Slide 70 text
@holly_cummins #RedHat
remember the airline scheduling application?
Slide 71
Slide 71 text
@holly_cummins #RedHat
remember the airline scheduling application?
“[With Quarkus], we can run 3 times denser
deployments without sacrificing availability and
response times of services. ”
– Thorsten Pohl
Slide 72
Slide 72 text
let’s talk about throughput.
https://www.redhat.com/en/resources/mi-quarkus-lab-validation-idc-analyst-paper
48 concurrent
connections
Traditional cloud native stack 3555 req/s
Slide 73
Slide 73 text
let’s talk about throughput.
quarkus native 3212 req/s
https://www.redhat.com/en/resources/mi-quarkus-lab-validation-idc-analyst-paper
48 concurrent
connections
Traditional cloud native stack 3555 req/s
Slide 74
Slide 74 text
let’s talk about throughput.
quarkus native 3212 req/s
https://www.redhat.com/en/resources/mi-quarkus-lab-validation-idc-analyst-paper
a trade-off of throughput
against footprint
48 concurrent
connections
Traditional cloud native stack 3555 req/s
Slide 75
Slide 75 text
@holly_cummins
Native compilation trade-offs
throughput
startup time +
footprint
Slide 76
Slide 76 text
@holly_cummins
Native compilation trade-offs
throughput
startup time +
footprint
Slide 77
Slide 77 text
@holly_cummins
this is a classic tradeoff
Slide 78
Slide 78 text
but …
traditional cloud native stack 3555 req/s
quarkus native 3212 req/s
https://www.redhat.com/en/resources/mi-quarkus-lab-validation-idc-analyst-paper
a trade-off of throughput
against footprint
48 concurrent
connections
Slide 79
Slide 79 text
but …
traditional cloud native stack 3555 req/s
quarkus on jvm 6389 req/s
quarkus native 3212 req/s
https://www.redhat.com/en/resources/mi-quarkus-lab-validation-idc-analyst-paper
a trade-off of throughput
against footprint
48 concurrent
connections
Slide 80
Slide 80 text
but …
traditional cloud native stack 3555 req/s
quarkus on jvm 6389 req/s
quarkus native 3212 req/s
https://www.redhat.com/en/resources/mi-quarkus-lab-validation-idc-analyst-paper
no trade-off, just better :)
a trade-off of throughput
against footprint
48 concurrent
connections
Slide 81
Slide 81 text
@holly_cummins
we beat the trade-off.
throughput
startup time
+ footprint
Slide 82
Slide 82 text
@holly_cummins
we beat the trade-off.
throughput
startup time
+ footprint
Slide 83
Slide 83 text
@holly_cummins
we beat the trade-off.
throughput
startup time
+ footprint
it’s a double-win.
Slide 84
Slide 84 text
No content
Slide 85
Slide 85 text
@holly_cummins #RedHat
https://medium.com/arconsis/spring-boot-vs-quarkus-part-2-jvm-runtime-performance-af45d0db116e
Slide 86
Slide 86 text
@holly_cummins #RedHat
https://medium.com/arconsis/spring-boot-vs-quarkus-part-2-jvm-runtime-performance-af45d0db116e
spring boot quarkus
response time 1901 ms 294 ms
throughput 523 req/s 3374 req/s
Slide 87
Slide 87 text
#Quarkus @holly_cummins
users care (a lot)
about response time
Slide 88
Slide 88 text
better startup
better footprint
better throughput
wait, isn’t that
something for nothing?
Slide 89
Slide 89 text
@holly_cummins #RedHat
the two ways of improving performance
•trade off one thing against another
•eliminate waste
Slide 90
Slide 90 text
traditional cloud native stack 3555 req/s
quarkus on jvm 6389 req/s
quarkus native 3212 req/s
https://www.redhat.com/en/resources/mi-quarkus-lab-validation-idc-analyst-paper
Slide 91
Slide 91 text
traditional cloud native stack 3555 req/s
quarkus on jvm 6389 req/s
quarkus native 3212 req/s
https://www.redhat.com/en/resources/mi-quarkus-lab-validation-idc-analyst-paper
a trade-off of throughput
against footprint
Slide 92
Slide 92 text
traditional cloud native stack 3555 req/s
quarkus on jvm 6389 req/s
quarkus native 3212 req/s
https://www.redhat.com/en/resources/mi-quarkus-lab-validation-idc-analyst-paper
no trade-off, just better :)
a trade-off of throughput
against footprint
Slide 93
Slide 93 text
@holly_cummins
#RedHat
where the win
comes from
Slide 94
Slide 94 text
@holly_cummins
#RedHat
application frameworks were
optimised for dynamism
Slide 95
Slide 95 text
@holly_cummins
#RedHat
application frameworks
optimised for dynamism
dynamism has a cost
Slide 96
Slide 96 text
@holly_cummins
#RedHat
paying a dynamism tax
… even though the app is not dynamic
Slide 97
Slide 97 text
@holly_cummins #RedHat
cloud apps
are immutable
now
Slide 98
Slide 98 text
@holly_cummins #RedHat
cloud apps
are immutable
now
Slide 99
Slide 99 text
@holly_cummins #RedHat
cloud apps
are immutable
now
Slide 100
Slide 100 text
@holly_cummins #RedHat
cloud apps
are immutable
now
Slide 101
Slide 101 text
@holly_cummins #RedHat
cloud apps
are immutable
now
Slide 102
Slide 102 text
@holly_cummins #RedHat
cloud apps
are immutable
now
Slide 103
Slide 103 text
@holly_cummins
#RedHat
a highly dynamic
runtime in a container
is pointless
Slide 104
Slide 104 text
@holly_cummins
#RedHat
a highly dynamic
runtime in a container
is pointless
loading classes that aren’t needed
Slide 105
Slide 105 text
@holly_cummins
#RedHat
a highly dynamic
runtime in a container
is pointless
loading classes that aren’t needed
expensive, slow, reflection
Slide 106
Slide 106 text
@holly_cummins
#RedHat
a highly dynamic
runtime in a container
is pointless
loading classes that aren’t needed
expensive, slow, reflection
the same initialisation on every startup
Slide 107
Slide 107 text
@holly_cummins #RedHat
how does a java framework start?
Slide 108
Slide 108 text
@holly_cummins #RedHat
how does a java framework start?
build time
Slide 109
Slide 109 text
@holly_cummins #RedHat
how does a java framework start?
build time runtime
Slide 110
Slide 110 text
@holly_cummins #RedHat
how does a java framework start?
build time runtime
Slide 111
Slide 111 text
@holly_cummins #RedHat
how does a java framework start?
packaging
(maven, gradle…)
build time runtime
Slide 112
Slide 112 text
@holly_cummins #RedHat
how does a java framework start?
build time runtime
Slide 113
Slide 113 text
@holly_cummins #RedHat
how does a java framework start?
build time runtime
Slide 114
Slide 114 text
@holly_cummins #RedHat
how does a java framework start?
>
build time runtime
load and parse
• config files
• properties
• yaml
• xml
• etc.
Slide 115
Slide 115 text
@holly_cummins #RedHat
how does a java framework start?
>
build time runtime
Slide 116
Slide 116 text
@holly_cummins #RedHat
how does a java framework start?
@
@
>
build time runtime
• classpath scanning and
annotation discovery
• attempt to load class to
enable/disable features
Slide 117
Slide 117 text
@holly_cummins #RedHat
how does a java framework start?
@
@
>
build time runtime
Slide 118
Slide 118 text
@holly_cummins #RedHat
how does a java framework start?
@
@
>
build time runtime
build a metamodel
of the world
Slide 119
Slide 119 text
@holly_cummins #RedHat
how does a java framework start?
@
@
>
build time runtime
Slide 120
Slide 120 text
@holly_cummins #RedHat
how does a java framework start?
@
@
>
build time runtime
start
• thread pools
• I/O
• etc.
Slide 121
Slide 121 text
@holly_cummins #RedHat
how does a java framework start?
@
@
>
build time runtime
ready to
do work!
Slide 122
Slide 122 text
@holly_cummins #RedHat
what if we start the application more than once?
@
@
>
Slide 123
Slide 123 text
@holly_cummins #RedHat
what if we start the application more than once?
@
@
>
@
@
>
Slide 124
Slide 124 text
@holly_cummins #RedHat
what if we start the application more than once?
@
@
>
@
@
>
@
@
>
Slide 125
Slide 125 text
@holly_cummins #RedHat
what if we start the application more than once?
@
@
>
@
@
>
@
@
>
@
@
>
Slide 126
Slide 126 text
@holly_cummins #RedHat
what if we start the application more than once?
@
@
>
@
@
>
@
@
>
@
@
>
Slide 127
Slide 127 text
@holly_cummins #RedHat
what if we start the application more than once?
@
@
>
@
@
>
@
@
>
@
@
>
so much work gets
redone every time
Slide 128
Slide 128 text
@holly_cummins
Hibernate
speed example: JTA auto-wiring
Slide 129
Slide 129 text
@holly_cummins
Hibernate
speed example: JTA auto-wiring
Class.forName(“LikelyJTAImplementation”);
Slide 130
Slide 130 text
@holly_cummins
Hibernate
speed example: JTA auto-wiring
Class.forName(“LikelyJTAImplementation”);
Slide 131
Slide 131 text
@holly_cummins
Hibernate
speed example: JTA auto-wiring
Class.forName(“LikelyJTAImplementation”);
Class.forName(“APossibleJTAImplementation”);
Slide 132
Slide 132 text
@holly_cummins
Hibernate
speed example: JTA auto-wiring
Class.forName(“LikelyJTAImplementation”);
Class.forName(“APossibleJTAImplementation”);
Class.forName(“AnotherJTAImplementation”);
Slide 133
Slide 133 text
@holly_cummins
Hibernate
speed example: JTA auto-wiring
Class.forName(“LikelyJTAImplementation”);
Class.forName(“APossibleJTAImplementation”);
Class.forName(“AnotherJTAImplementation”);
…
Slide 134
Slide 134 text
@holly_cummins
Hibernate
speed example: JTA auto-wiring
Class.forName(“LikelyJTAImplementation”);
Class.forName(“APossibleJTAImplementation”);
Class.forName(“AnotherJTAImplementation”);
Class.forName(“NicheJTAImplementation”);
…
Slide 135
Slide 135 text
@holly_cummins
Hibernate
speed example: JTA auto-wiring
Class.forName(“LikelyJTAImplementation”);
Class.forName(“APossibleJTAImplementation”);
Class.forName(“AnotherJTAImplementation”);
Class.forName(“NicheJTAImplementation”);
Class.forName(“VeryNicheJTAImplementation”);
…
Slide 136
Slide 136 text
@holly_cummins
Hibernate
speed example: JTA auto-wiring
Class.forName(“LikelyJTAImplementation”);
Class.forName(“APossibleJTAImplementation”);
Class.forName(“AnotherJTAImplementation”);
Class.forName(“NicheJTAImplementation”);
Class.forName(“VeryNicheJTAImplementation”);
…
Slide 137
Slide 137 text
@holly_cummins
Hibernate
speed example: JTA auto-wiring
Class.forName(“LikelyJTAImplementation”);
Class.forName(“APossibleJTAImplementation”);
Class.forName(“AnotherJTAImplementation”);
Class.forName(“NicheJTAImplementation”);
Class.forName(“VeryNicheJTAImplementation”);
…
~129
auto-wiring attempts
Slide 138
Slide 138 text
@holly_cummins
Hibernate
speed example: JTA auto-wiring
Class.forName(“LikelyJTAImplementation”);
Class.forName(“APossibleJTAImplementation”);
Class.forName(“AnotherJTAImplementation”);
Class.forName(“NicheJTAImplementation”);
Class.forName(“VeryNicheJTAImplementation”);
…
~129
auto-wiring attempts
every single start.
Slide 139
Slide 139 text
@holly_cummins
it’s not just JTA
this happens for lots of internal service bindings
Slide 140
Slide 140 text
@holly_cummins
JVM
footprint example: Hibernate
Slide 141
Slide 141 text
@holly_cummins
JVM spends time
loading classes for
specific databases
JVM
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
footprint example: Hibernate
Slide 142
Slide 142 text
@holly_cummins
JVM spends time
loading classes for
specific databases
JVM
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
class for
unused
database
turns out they’re
never used
footprint example: Hibernate
Slide 143
Slide 143 text
@holly_cummins
JVM spends time
loading classes for
specific databases
JVM
turns out they’re
never used
JIT spends time
unloading classes
footprint example: Hibernate
Slide 144
Slide 144 text
@holly_cummins
Hibernate
example:
~500 classes
which are only useful if you're running an Oracle database
loaded and then unloaded
Slide 145
Slide 145 text
@holly_cummins
Hibernate
example:
~500 classes
which are only useful if you're running an Oracle database
loaded and then unloaded
every single start.
Slide 146
Slide 146 text
@holly_cummins
the true cost of loaded classes isn’t just memory + start time
Slide 147
Slide 147 text
@holly_cummins
the true cost of loaded classes isn’t just memory + start time
method dispatching:
Slide 148
Slide 148 text
@holly_cummins
interface
the true cost of loaded classes isn’t just memory + start time
method dispatching:
Slide 149
Slide 149 text
@holly_cummins
unused
implementation
the one we
want
interface
unused
implementation
unused
implementation
the true cost of loaded classes isn’t just memory + start time
method dispatching:
Slide 150
Slide 150 text
@holly_cummins
unused
implementation
the one we
want
interface
unused
implementation
unused
implementation
the true cost of loaded classes isn’t just memory + start time
method dispatching:
Slide 151
Slide 151 text
@holly_cummins
unused
implementation
the one we
want
interface
megamorphic call
slow dispatching
unused
implementation
unused
implementation
the true cost of loaded classes isn’t just memory + start time
method dispatching:
Slide 152
Slide 152 text
@holly_cummins
the true cost of loaded classes isn’t just memory + start time
the one we
want
interface
Slide 153
Slide 153 text
@holly_cummins
the true cost of loaded classes isn’t just memory + start time
the one we
want
monomorphic call
fast dispatching
interface
Slide 154
Slide 154 text
@holly_cummins
how do we fix all this?
Slide 155
Slide 155 text
@holly_cummins #RedHat
@
@
>
build time runtime
what if we initialize at build time?
Slide 156
Slide 156 text
@holly_cummins #RedHat
@
@
>
build time runtime
what if we initialize at build time?
Slide 157
Slide 157 text
@holly_cummins #RedHat
@
@
>
build time runtime
start
• thread pools
• I/O
• etc.
what if we initialize at build time?
Slide 158
Slide 158 text
@holly_cummins #RedHat
@
@
>
build time runtime
ready to
do work!
start
• thread pools
• I/O
• etc.
what if we initialize at build time?
Slide 159
Slide 159 text
@holly_cummins #RedHat
@
@
>
repeated starts are now efficient
Slide 160
Slide 160 text
@holly_cummins #RedHat
@
@
>
repeated starts are now efficient
Slide 161
Slide 161 text
@holly_cummins #RedHat
@
@
>
repeated starts are now efficient
Slide 162
Slide 162 text
@holly_cummins #RedHat
@
@
>
repeated starts are now efficient
Slide 163
Slide 163 text
@holly_cummins #RedHat
@
@
>
repeated starts are now efficient
Slide 164
Slide 164 text
@holly_cummins #RedHat
@
@
>
repeated starts are now efficient
Slide 165
Slide 165 text
@holly_cummins #RedHat
@
@
>
repeated starts are now efficient
less wasted work
Slide 166
Slide 166 text
@holly_cummins #RedHat
the Quarkus way enables native compilation
native
(graalvm)
@
@
>
jvm
build time
Slide 167
Slide 167 text
@holly_cummins #RedHat
the Quarkus way enables native compilation
native
(graalvm)
@
@
>
jvm
build time
Slide 168
Slide 168 text
ok but we don’t need native
Slide 169
Slide 169 text
ok but we don’t need native
quarkus is faster and smaller than legacy
frameworks, even running on the jvm
Slide 170
Slide 170 text
@holly_cummins
Slide 171
Slide 171 text
@holly_cummins
doing more up-front
Slide 172
Slide 172 text
@holly_cummins
doing more up-front
- speeds up start
Slide 173
Slide 173 text
@holly_cummins
doing more up-front
- speeds up start
- shrinks memory footprint
Slide 174
Slide 174 text
@holly_cummins
doing more up-front
- speeds up start
- shrinks memory footprint
- improves throughput (!)
Slide 175
Slide 175 text
ok but is performance all there is?
Slide 176
Slide 176 text
ok but is performance all there is?
developer joy
Slide 177
Slide 177 text
@holly_cummins #RedHat
Slide 178
Slide 178 text
@holly_cummins #RedHat
with quarkus, developers
have to type less boilerplate
this is good, because we’re about to
do a live demo with a broken finger
Slide 179
Slide 179 text
to the code!
@holly_cummins #RedHat
Slide 180
Slide 180 text
@holly_cummins
#RedHat
mvn quarkus:dev
zero-config live coding
Slide 181
Slide 181 text
@holly_cummins
#RedHat
tests are run on every code change
“reverse code coverage” means only relevant tests run
mvn quarkus:dev
continuous testing
Slide 182
Slide 182 text
@holly_cummins
#RedHat
developer UI
Slide 183
Slide 183 text
@holly_cummins
#RedHat
zero-config testcontainers
Slide 184
Slide 184 text
@holly_cummins #RedHat
testcontainers integration … without quarkus
@TestConfiguration(proxyBeanMethods = false)
public class ContainersConfig {
@Bean
@ServiceConnection
public PostgreSQLContainer> postgres() {
return new PostgreSQLContainer<>(DockerImageName.parse("postgres:14"));
}
}
public class TestApplication {
public static void main(String[] args) {
SpringApplication
.from(MySpringDataApplication::main)
.with(ContainersConfig.class)
.run(args);
}
}
@Import(ContainersConfig.class)
Slide 185
Slide 185 text
@holly_cummins #RedHat
testcontainers integration … without quarkus
Slide 186
Slide 186 text
@holly_cummins #RedHat
zero-config testcontainers integration
the only thing you need to do to make
testcontainers work is not configure anything
# configure your datasource
quarkus.datasource.db-kind = postgresql
quarkus.datasource.username = sarah
quarkus.datasource.password = connor
quarkus.datasource.jdbc.url = jdbc:postgresql://localhost:5432/mydatabase
# drop and create the database at startup
quarkus.hibernate-orm.database.generation = drop-and-create
Slide 187
Slide 187 text
@holly_cummins #RedHat
zero-config testcontainers integration
the only thing you need to do to make
testcontainers work is not configure anything
# drop and create the database at startup
quarkus.hibernate-orm.database.generation = drop-and-create
Slide 188
Slide 188 text
@holly_cummins #RedHat
zero-config testcontainers integration
the only thing you need to do to make
testcontainers work is not configure anything
Slide 189
Slide 189 text
@holly_cummins #RedHat
zero-config testcontainers integration
the only thing you need to do to make
testcontainers work is not configure anything
quarkus also auto-invokes flyway and liquibase
Slide 190
Slide 190 text
@holly_cummins #RedHat
zero-config testcontainers integration
realistically, use profiles so things work in
production :)
# configure your datasource
%prod.quarkus.datasource.db-kind = postgresql
%prod.quarkus.datasource.username = sarah
%prod.quarkus.datasource.password = connor
%prod.quarkus.datasource.jdbc.url = jdbc:postgresql://localhost:5432/mydatabase
# on real databases, defaults to ‘none’, but let’s validate
%prod.quarkus.hibernate-orm.database.generation = validate
Slide 191
Slide 191 text
@holly_cummins
#RedHat
• databases
• redis
• keycloak
• kafka
• elasticsearch
• kubernetes
• … or add your own
auto-provision services
“dev services”
using testcontainers under the hood
Slide 192
Slide 192 text
@holly_cummins
#RedHat
more opinions, less boilerplate
Slide 193
Slide 193 text
@holly_cummins
#RedHat
package com.example;
import org.jboss.logging.Logger;
public class MyService {
private static final Logger log = Logger.getLogger(MyService.class);
public void doSomething() {
log.info("It works!");
}
}
example: logging
Slide 194
Slide 194 text
@holly_cummins
#RedHat
package com.example;
import org.jboss.logging.Logger;
public class MyService {
private static final Logger log = Logger.getLogger(MyService.class);
public void doSomething() {
log.info("It works!");
}
}
example: logging
import io.quarkus.logging.Log;
Log
Slide 195
Slide 195 text
@holly_cummins
“but isn’t that dynamism expensive?”
Slide 196
Slide 196 text
@holly_cummins
no.
“but isn’t that dynamism expensive?”
Slide 197
Slide 197 text
@holly_cummins
no.
it’s done at build time
“but isn’t that dynamism expensive?”
Slide 198
Slide 198 text
@holly_cummins
logging: compiled version
public class MyService {
// injected
private static final Logger $logger = Logger.getLogger(Thing.class)
public void doSomething() {
$logger.info(“It works!”);
}
}
Slide 199
Slide 199 text
@holly_cummins
#RedHat
example: query parameters
public String hello(@QueryParam("name") String name) {
Slide 200
Slide 200 text
@holly_cummins
#RedHat
example: query parameters
public String hello(@RestQuery String name) {
Slide 201
Slide 201 text
@holly_cummins
#RedHat
package org.acme;
import org.springframework.boot.SpringApplication;
import org.springframework.boot.autoconfigure.SpringBootApplication;
@SpringBootApplication
public class SpringDemo {
public static void main(String[] args) {
SpringApplication.run(SpringDemo.class, args);
}
}
example: declaring an application
Slide 202
Slide 202 text
@holly_cummins
#RedHat
example: declaring an application
Slide 203
Slide 203 text
@holly_cummins
what if…
you could inherit boilerplate
Hibernate queries from a superclass,
instead of having to write them all?
example: hibernate
Slide 204
Slide 204 text
@holly_cummins
#RedHat
@ApplicationScoped
public class GreetingRepository {
public Entity findByName(int name) {
return find("name", name).firstResult();
}
void persist(Entity entity) {}
void delete(Entity entity) {}
Entity findById(Id id) {}
List list(String query, Sort sort, Object... params) {
return null;
}
Stream stream(String query, Object... params) {
return null;
}
long count() {
return 0;
}
long count(String query, Object... params) {
return 0;
}
}
example: panache + hibernate
Slide 205
Slide 205 text
@holly_cummins
#RedHat
example: panache + hibernate
@ApplicationScoped
public class GreetingRepository implements PanacheRepository {
public Entity findByName(int name) {
return find("name", name).firstResult();
}
}
Slide 206
Slide 206 text
@holly_cummins
#RedHat
DAO
example: panache + hibernate
@ApplicationScoped
public class GreetingRepository implements PanacheRepository {
public Entity findByName(int name) {
return find("name", name).firstResult();
}
}
repository pattern
Slide 207
Slide 207 text
@holly_cummins
#RedHat
example: panache + hibernate
Slide 208
Slide 208 text
@holly_cummins
#RedHat
example: panache + hibernate
active record pattern
@Entity
public class Greeting extends PanacheEntity {
public String name;
public LocalDate issued;
@Version
public int version;
public static List getTodaysGreetings() {
return list("date", LocalDate.now());
}
}
Slide 209
Slide 209 text
@holly_cummins
#RedHat
int port = 8081;
ClientInterceptor[] interceptors = new ClientInterceptor[3];
interceptors[0] = new EventLoopBlockingCheckInterceptor();
interceptors[1] = new IOThreadClientInterceptor();
interceptors[2] = new StorkMeasuringGrpcInterceptor();
// etc
Channel channel = ManagedChannelBuilder
.forAddress(“localhost", port)
.usePlaintext()
.build();
GreeterGrpc.GreeterStub greeter = GreeterGrpc
.newStub(channel)
.withInterceptors(interceptors);
example: gRPC
Slide 210
Slide 210 text
@holly_cummins
#RedHat
example: gRPC
@GrpcClient
Greeter greeter;
Slide 211
Slide 211 text
@holly_cummins
#RedHat
ok but i don’t like magic
Slide 212
Slide 212 text
@holly_cummins
#RedHat
ok but i don’t like magic
enrichment happens at build time
Slide 213
Slide 213 text
@holly_cummins
#RedHat
ok but i don’t like magic
enrichment happens at build time
no performance drag
Slide 214
Slide 214 text
doing more up-front enables better devex
runtime
build time
@
@
>
Slide 215
Slide 215 text
doing more up-front enables better devex
runtime
build time
we can do cool code introspections
here that would be too expensive and
annoying to do at runtime
@
@
>
Slide 216
Slide 216 text
@holly_cummins
#RedHat
ok but i still don’t like magic
Slide 217
Slide 217 text
@holly_cummins
#RedHat
the old ways all still work
ok but i still don’t like magic
Slide 218
Slide 218 text
@holly_cummins
#RedHat
the old ways all still work
but you don’t have to type all
the stuff unless you want to
ok but i still don’t like magic
Slide 219
Slide 219 text
@holly_cummins
#RedHat
ok but we use spring
for existing applications, there is a trade-off there.
Slide 220
Slide 220 text
@holly_cummins
#RedHat
io.quarkus
quarkus-spring-web
io.quarkus
quarkus-spring-data-jpa
ok but we use spring
option 1: compatibility libraries
option 2: migration tooling
• migration toolkit for applications (mta)
• windup
• open rewrite
Slide 221
Slide 221 text
@holly_cummins #RedHat
“After a week of development with
quarkus, I was able to regain the
same level of productivity as when
I was developing with Spring Boot.”
– Fawaz Paraïso, Decathlon
Slide 222
Slide 222 text
@holly_cummins #RedHat
one last thing …
Slide 223
Slide 223 text
@holly_cummins #RedHat
does being faster
and lighter mean
quarkus is greener?
Slide 224
Slide 224 text
digression:
how do you measure carbon?
Slide 225
Slide 225 text
@holly_cummins #RedHat
step 1: measure power usage
Slide 226
Slide 226 text
@holly_cummins #RedHat
step 1: measure power usage
wall power measurement
Slide 227
Slide 227 text
@holly_cummins #RedHat
step 1: measure power usage
wall power measurement
RAPL
Slide 228
Slide 228 text
@holly_cummins #RedHat
Slide 229
Slide 229 text
@holly_cummins #RedHat
Slide 230
Slide 230 text
@holly_cummins #RedHat
Slide 231
Slide 231 text
@holly_cummins #RedHat
load
Slide 232
Slide 232 text
@holly_cummins #RedHat
Source: Teads EC2 instances carbon dataset
Slide 233
Slide 233 text
@holly_cummins #RedHat
density
Source: Clement Escoffier
experiment 1: cloud
Setup:
• 800 requests/second, over 20 days
• SLA > 99%
• AWS instances
Assumptions:
• Costs are for us-east-1 data centre
Slide 234
Slide 234 text
@holly_cummins #RedHat
Setup:
• 800 requests/second, over 20
days
• SLA > 99%
Assumptions:
Source: Clement Escoffier x Teads
cloud carbon impact of framework choice
interpolated carbon metrics
Slide 235
Slide 235 text
@holly_cummins #RedHat
Setup:
• 800 requests/second, over 20
days
• SLA > 99%
Assumptions:
Source: Clement Escoffier x Teads
cloud carbon impact of framework choice
the carbon is lower because
the cost is lower
interpolated carbon metrics
Slide 236
Slide 236 text
@holly_cummins #RedHat
Setup:
• REST + CRUD
• large heap
• RAPL energy measurement
• multiple instances to
support high load
Assumptions:
• US energy mix
Source: John O’Hara
experiment 2: RAPL measurements
Slide 237
Slide 237 text
@holly_cummins #RedHat
Setup:
• REST + CRUD
• large heap
• RAPL energy measurement
• multiple instances to
support high load
Assumptions:
• US energy mix
Source: John O’Hara
experiment 2: RAPL measurements
quarkus on JVM has the lowest carbon …
because it has the highest throughput
Slide 238
Slide 238 text
• quarkus cuts carbon by ~2-3x*
• native consumes more carbon
than JVM
carbon measurements:
conclusions
Slide 239
Slide 239 text
@holly_cummins #RedHat
another last thing …
Slide 240
Slide 240 text
@holly_cummins #RedHat
open source community of contributors
Slide 241
Slide 241 text
@holly_cummins #RedHat
Emiliia Nesterovych
Emmanuel Bernard
Emre Kaplan
Enrique gonzález Martínez
Enrique Mingorance Cano
Eoin Gallinagh
Eric Deandrea
Eric Wittmann
Erik Åsén
Erik Mattheis
Erin Schnabel
Eugene Berman
Evan Shortiss
Fabricio Gregorio
faculbsz
Falko Modler
Fedor Dudinskiy
Felipe Carvalho dos Anjos Formentin
Felipe Henrique Gross Windmoller
Fernando Comunello
Fernando Henrique
fhavel
Fikru Mengesha
Filippe Spolti
Florian Beutel
Florian Bütler
Florian Heubeck
Florin Botis
Foivos Zakkak
Foobartender
Fouad Almalki
Francesco Nigro
Francisco Javier Tirado Sarti
Francois Steyn
Frank Eichfelder
franz1981
freakse-sa
Fred Bricon
Frédérc Blanc
Freeman Fang
Fu Cheng
Gabriele Cardosi
Galder Zamarreño
galiacheng
Gavin King
Gavin Ray
Geert Schuring
Geoffrey De Smet
Geoffrey GREBERT
Georg Leber
George Gastaldi
manofthepeace
Manyanda Chitimbo
Marat Gubaidullin
Marc Nuri
Marc Schlegel
Marc Wrobel
Marcel Hanser
Marcel Lohmann
Marcell Cruz
Marcelo Pereira
Marcin Czeczko
Marcin Kłopotek
Marco Bungart
Marco Schaub
Marco Yeung
Marco Zanghì
Marcus Paulo
Marek goldmann
Marek Skacelik
Marián Macik
Mario Fusco
MarioHNogueira
Mark Lambert
Mark Little
Mark McLaughlin
Mark Sailes
marko-bekhta
Markus Heberling
Markus Himmel
Markus Schwer
Martin C. Richards
Martin Grammelspacher
Martin Kouba
Martin Muzikar
Martin Panzer
Martin Weiler
martin-kofoed-jyskebank-dk
MartinWitt
Marvin B. Lillehaug
masini
Matej Novotny
Matej Vasek
Matheus Cruz
Mathias Holzer
Matteo Mortari
Matthias Andreas Benkard
Matthias Cullmann
mauroal
Max Andersen
Max Gabrielsson
Max Rydahl Andersen
Victor Hugo de Oliveira Molinar
Vincent Sevel
Vincent van Dam
Vinícius Ferraz Campos Florentino
Viswa Teja Nariboina
Vladimir Konkov
Vojtech Juranek
Vratislav Hais
w.glanzer
Walter Medvedeo
Wayne Ellis
Werner Glanzer
Willem Jan Glerum
William Antônio Siqueira
Wim goeman
Wippermueller, Frank
wojciech.stryjewski
Xavier
Xieshen
xstefank
Y. Luis
Yann-Thomas LE MOIGNE
Yannick Reifschneider
YassinHajaj
Yelzhas Suleimenov
yesunch9
Yoann Rodière
Yoshikazu Nojima
Youngmin Koo
Yubao Liu
yugoccp
Yukihiro Okada
Zaheed Beita
zanmagerl
zedbeit
Zheng Feng
Žiga Deisinger
Zineb Bendhiba
zohar
Zoran Regvart
Шумов Игорь Юрьевич
出
门
三不惹
open source community of contributors
Slide 242
Slide 242 text
@holly_cummins #RedHat
Slide 243
Slide 243 text
@holly_cummins #RedHat
tl;dpa
(too long didn’t pay attention)
deployment density
lower cloud bill
frictionless development experience
Medium Nano
auto-provision services
zero-config live coding
continuous testing
developer UI
greener
Slide 244
Slide 244 text
@holly_cummins #RedHat
tl;sdpa
(too long still didn’t pay attention)
what problems is quarkus solving?
tea
typing
tedium
tolls
taxes
Slide 245
Slide 245 text
slides
Holly Cummins
Red Hat
https://hollycummins.com/faster-greener-code-europe/