Quarkus 1.12.0.CR1 released

[Guillaume Smet]

Hi,

We just released Quarkus 1.12.0.CR1.

One of the most important changes of this release is the switch to the fast-jar packaging by default, which is explained in the Migration guide: https://github.com/quarkusio/quarkus/wiki/Migration-Guide-1.12 .

If you have some spare time, we would be pleased to have feedback from the field on upgrading your applications to Quarkus 1.12.0.CR1.

If everything works well, just reply on this thread.

If you encounter an issue, please open a GitHub issue with a reproducer so that we can have a look.

As usual when testing a CR:

- you need to upgrade both the Quarkus version and the Quarkus Maven plugin version to 1.12.0.CR1

- you need to use the quarkus-bom, not the quarkus-universe-bom

We will prepare the Final core artifacts in a week.

Thanks!

--

Guillaume

[swidersk...@gmail.com]

Just a quick note based on my tests (RESTEasy classic with Jackson, Kafka, MQTT, Camel via reactive messaging, hibernate, Funqy http, Funqy Knative events). It pretty much all worked fine. Ran into one issue about reactive messaging emitters that throw following error

SRMSG00075: Invalid Emitter injection found for `io.quarkus.arc.impl.CurrentInjectionPointProvider$InjectionPointImpl@3ee62bf2`. Injecting an `Emitter<Message<T>>` is invalid. You can use an `Emitter<T>` to send instances of `T` and `Message<T>`.

So it seems like it started to be more restrictive when it comes to injection of emitter that it must have typed message on emitter. Haven’t found any note about this in the migration guide so mentioning just in case. 

Great work everyone!

Maciej

--
You received this message because you are subscribed to the Google Groups "Quarkus Development mailing list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to quarkus-dev...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/quarkus-dev/CALt0%2Bo_bv4MmYLoSPSaVwO%2B%2BH2nzAQwRVhB6wftCQxS6HLoHdQ%40mail.gmail.com.

[Loïc MATHIEU]

@Maciej, see the following PR on SR reactive messaging: https://github.com/smallrye/smallrye-reactive-messaging/pull/929

Using an Emitter<Message<T>> was always wrong as it means that the payload is a Message<T> not T, so you are sending a Message<Message<T>> inside your emitter and not a Message<T> ...

This PR generates an exception when it detects such programmatic error.

To view this discussion on the web visit https://groups.google.com/d/msgid/quarkus-dev/06F8FD96-5FD2-4021-BBB2-ECD55ADF0EC2%40gmail.com.

[swidersk...@gmail.com]

Thanks for the reference Loïc!

And agreed it does make sense to fail fast on this helps developers to improve the code base which I did :)

Maciej

[V. Sevel]

Hi all,

I finished migrating to CR1.

I created https://github.com/quarkusio/quarkus/issues/15103

here is the list of adaptations I had to do after I compared a newly created app and our current setup:

• move to fast jar
• change distroless from quay.io/quarkus/quarkus-distroless-image:20.3-java11 to quay.io/quarkus/quarkus-distroless-image:1.0
• change quarkus-maven-plugin config with:
  • add <extensions>true</extensions>
  • change goals prepare, prepare-tests, build to build, generate-code, generate-code-tests

I don't know if it exists, but it would be nice, when a CR is released, to have the migration guide available.

that way we can apply the required known modifications, and avoid loosing time, or make others loose time with invalid issues.

that was not the case this time, but there was an occurrence for 1.11.0.CR1

Vincent

[Guillaume Smet]

Hi Vincent,

On Tue, Feb 16, 2021 at 11:36 AM V. Sevel <vvs...@gmail.com> wrote:

I created https://github.com/quarkusio/quarkus/issues/15103

I created a PR to fix this one.

 

here is the list of adaptations I had to do after I compared a newly created app and our current setup:

• move to fast jar
• change distroless from quay.io/quarkus/quarkus-distroless-image:20.3-java11 to quay.io/quarkus/quarkus-distroless-image:1.0
• change quarkus-maven-plugin config with:
  • add <extensions>true</extensions>
  • change goals prepare, prepare-tests, build to build, generate-code, generate-code-tests

I don't know if it exists, but it would be nice, when a CR is released, to have the migration guide available.

The migration guide was already there: https://github.com/quarkusio/quarkus/wiki/Migration-Guide-1.12 ... but I might have forgotten to point to it.

I added some paragraphs regarding the issues you had that were not already covered.

Thanks a lot for the feedback!

--

Guillaume

[Alex Soto Bueno]

There is some problem on how the migration guide is rendered.

--

You received this message because you are subscribed to the Google Groups "Quarkus Development mailing list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to quarkus-dev...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/quarkus-dev/CALt0%2Bo-Lhr9b7-dEmOy00rwgGhupQtk12y_7HvZpfkipU_PgNg%40mail.gmail.com.

[Guillaume Smet]

Looks like a GitHub bug. I resaved the same page and it seems to be working now.

[Guillaume Smet]

And it's wrong again.Yeah well :).

[Guillaume Smet]

Yeah, I confirm it's a bug. It's mostly wrong but when refreshing, sometimes I get it right.

Hopefully they will fix it before we release 1.12 :).

[Antonio Goncalves]

Hi,

I just realized that 1.12 was published a few days ago[1] but couldn't see any announcement. Did I miss something or is the announcement coming soon ?

Thanks

Antonio

[1] https://search.maven.org/artifact/io.quarkus/quarkus-bom/1.12.0.Final/pom

To view this discussion on the web visit https://groups.google.com/d/msgid/quarkus-dev/CALt0%2Bo9pmRE7%2BBEkUBKbaWNpijzoCaugF7DEMZE91SK%3DF1qzvg%40mail.gmail.com.

--

Antonio Goncalves 
Contractor, Java Champion and Pluralsight author

Blog | Twitter | LinkedIn | Author | Pluralsight | Devoxx France | Voxxed Microservices

[Georgios Andrianakis]

For .Final releases we only announce when the entire platform has been released.

That will be done soon and thus the announcement will come soon :)

To view this discussion on the web visit https://groups.google.com/d/msgid/quarkus-dev/CAKawoObEWjzjUd4BXNfzV0Zt%2BHn7qSBNaR_gmCTBCBcK1hR9pw%40mail.gmail.com.

[Antonio Goncalves]

I knew it was coming ;o) Thanks

[Guillaume Smet]

Platform release is mostly done, I'll write the announcement and then announce it.

To view this discussion on the web visit https://groups.google.com/d/msgid/quarkus-dev/CAKawoOa18Z46fws4qWEDB068PKJbHPBCLo03L57OZADGunmXUQ%40mail.gmail.com.

[Loïc MATHIEU]

Hello,

I did my usual comparison between 1.11 and 1.12, with an app with the following extensions: agroal, cdi, flyway, google-cloud-bigquery, hibernate-orm, hibernate-orm-panache, hibernate-validator, jaeger, jdbc-postgresql, mutiny, narayana-jta, rest-client, resteasy, resteasy-jsonb, security, security-oauth2, security-properties-file, smallrye-context-propagation, smallrye-fault-tolerance, smallrye-health, smallrye-metrics, smallrye-openapi, smallrye-opentracing.

I must say a big BRAVO for the startup improvements made on dev mode (60%) and fast jar (25%) ! This is impressive, especially the dev mode that starts more than two times quicker !

I also noticed a slight memory footprint improvement of fast jar.

Performance is on par (I didn't test with a high load).

Here are the results:

1.11:

- Startup dev mode: 2.38

- Startup jar: 2.08

- Memory footprint (jar): 187m

1.12:

- Startup dev mode: 1s

- Startup legacy jar: 2.05

- Startup fast jar: 1.58

- Footprint fast jar: 171m

I attached various profiles for those interested.

I will also test dev mode on another apps with kafka reactive messaging and check if it had this kind of huge performance improvements, this apps is also deployed on constrained environment (on k8s with very low cpu/mem) so I'm very interested to see if it gives some improvements also in these conditions.

Regards,

Loïc

To view this discussion on the web visit https://groups.google.com/d/msgid/quarkus-dev/CALt0%2Bo9-sYCy%2BT7aCU-Nhc%3D7byJ1Oe2AL9ctZMgrD2KiUcqQ3A%40mail.gmail.com.

[Georgios Andrianakis]

Thanks for checking Loic!

To view this discussion on the web visit https://groups.google.com/d/msgid/quarkus-dev/CAJLxjVFjuUuo6dKX9jvAVY6_S9Us%2BXd0F7rJSEF_PipnAtytng%40mail.gmail.com.

[Loïc MATHIEU]

So the startup improvements on dev mode seems to be a time reporting issue => https://github.com/quarkusio/quarkus/issues/15371

[Max Rydahl Andersen]

On 1 Mar 2021, at 11:52, Loïc MATHIEU wrote:

> So the startup improvements on dev mode seems to be a time reporting
> issue
> => https://github.com/quarkusio/quarkus/issues/15371

so with that corrected what are the new times?

/max

>>>>>> For *.Final* releases we only announce when the entire platform

>>>>>>>>>>>>> - move to fast jar
>>>>>>>>>>>>> - change distroless from

>>>>>>>>>>>>> quay.io/quarkus/quarkus-distroless-image:20.3-java11 to
>>>>>>>>>>>>> quay.io/quarkus/quarkus-distroless-image:1.0

>>>>>>>>>>>>> - change quarkus-maven-plugin config with:
>>>>>>>>>>>>> - add <extensions>true</extensions>
>>>>>>>>>>>>> - change goals prepare, prepare-tests, build to

>>>>>>>>>>>>> build,
>>>>>>>>>>>>> generate-code, generate-code-tests
>>>>>>>>>>>>>
>>>>>>>>>>>>> I don't know if it exists, but it would be nice, when a CR
>>>>>>>>>>>>> is
>>>>>>>>>>>>> released, to have the migration guide available.
>>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>> The migration guide was already there:
>>>>>>>>>>>> https://github.com/quarkusio/quarkus/wiki/Migration-Guide-1.12
>>>>>>>>>>>> ... but I might have forgotten to point to it.
>>>>>>>>>>>>
>>>>>>>>>>>> I added some paragraphs regarding the issues you had that
>>>>>>>>>>>> were
>>>>>>>>>>>> not already covered.
>>>>>>>>>>>>
>>>>>>>>>>>> Thanks a lot for the feedback!
>>>>>>>>>>>>
>>>>>>>>>>>> --
>>>>>>>>>>>> Guillaume
>>>>>>>>>>>>
>>>>>>>>>>>> --
>>>>>>>>>>>> You received this message because you are subscribed to the
>>>>>>>>>>>> Google Groups "Quarkus Development mailing list" group.
>>>>>>>>>>>> To unsubscribe from this group and stop receiving emails
>>>>>>>>>>>> from
>>>>>>>>>>>> it, send an email to
>>>>>>>>>>>> quarkus-dev...@googlegroups.com.
>>>>>>>>>>>> To view this discussion on the web visit
>>>>>>>>>>>> https://groups.google.com/d/msgid/quarkus-dev/CALt0%2Bo-Lhr9b7-dEmOy00rwgGhupQtk12y_7HvZpfkipU_PgNg%40mail.gmail.com

>>>>>>>>>>>> <https://groups.google.com/d/msgid/quarkus-dev/CALt0%2Bo-Lhr9b7-dEmOy00rwgGhupQtk12y_7HvZpfkipU_PgNg%40mail.gmail.com?utm_medium=email&utm_source=footer>
>>>>>>>>>>>> .

>>>>>>>>>>>>
>>>>>>>>>>> --
>>>>>>>> You received this message because you are subscribed to the
>>>>>>>> Google
>>>>>>>> Groups "Quarkus Development mailing list" group.
>>>>>>>> To unsubscribe from this group and stop receiving emails from
>>>>>>>> it,
>>>>>>>> send an email to quarkus-dev...@googlegroups.com.
>>>>>>>> To view this discussion on the web visit

>>>>>>>> https://groups.google.com/d/msgid/quarkus-dev/CALt0%2Bo9pmRE7%2BBEkUBKbaWNpijzoCaugF7DEMZE91SK%3DF1qzvg%40mail.gmail.com
>>>>>>>> <https://groups.google.com/d/msgid/quarkus-dev/CALt0%2Bo9pmRE7%2BBEkUBKbaWNpijzoCaugF7DEMZE91SK%3DF1qzvg%40mail.gmail.com?utm_medium=email&utm_source=footer>
>>>>>>>> .

>>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> --
>>>>>>> Antonio Goncalves
>>>>>>> Contractor, Java Champion and Pluralsight author
>>>>>>>

>>>>>>> Blog <https://antoniogoncalves.org/> | Twitter
>>>>>>> <https://twitter.com/agoncal> | LinkedIn
>>>>>>> <https://www.linkedin.com/in/agoncal> | A
>>>>>>> <http://amazon.com/author/agoncal> uthor
>>>>>>> <http://amazon.com/author/agoncal> | Pluralsight
>>>>>>> <https://app.pluralsight.com/profile/author/antonio-goncalves> |
>>>>>>> Devoxx
>>>>>>> France <http://www.devoxx.fr/> | Voxxed Microservices
>>>>>>> <https://voxxeddays.com/microservices>

>>>>>>>
>>>>>>> --
>>>>>>> You received this message because you are subscribed to the
>>>>>>> Google
>>>>>>> Groups "Quarkus Development mailing list" group.
>>>>>>> To unsubscribe from this group and stop receiving emails from
>>>>>>> it,
>>>>>>> send an email to quarkus-dev...@googlegroups.com.
>>>>>>> To view this discussion on the web visit

>>>>>>> https://groups.google.com/d/msgid/quarkus-dev/CAKawoObEWjzjUd4BXNfzV0Zt%2BHn7qSBNaR_gmCTBCBcK1hR9pw%40mail.gmail.com
>>>>>>> <https://groups.google.com/d/msgid/quarkus-dev/CAKawoObEWjzjUd4BXNfzV0Zt%2BHn7qSBNaR_gmCTBCBcK1hR9pw%40mail.gmail.com?utm_medium=email&utm_source=footer>
>>>>>>> .

>>>>>>>
>>>>>>
>>>>>
>>>>> --
>>>>> Antonio Goncalves
>>>>> Contractor, Java Champion and Pluralsight author
>>>>>

>>>>> Blog <https://antoniogoncalves.org/> | Twitter
>>>>> <https://twitter.com/agoncal> | LinkedIn
>>>>> <https://www.linkedin.com/in/agoncal> | A
>>>>> <http://amazon.com/author/agoncal> uthor
>>>>> <http://amazon.com/author/agoncal> | Pluralsight
>>>>> <https://app.pluralsight.com/profile/author/antonio-goncalves> |
>>>>> Devoxx
>>>>> France <http://www.devoxx.fr/> | Voxxed Microservices
>>>>> <https://voxxeddays.com/microservices>

>>>>>
>>>>> --
>>>>> You received this message because you are subscribed to the Google
>>>>> Groups "Quarkus Development mailing list" group.
>>>>> To unsubscribe from this group and stop receiving emails from it,
>>>>> send
>>>>> an email to quarkus-dev...@googlegroups.com.
>>>>> To view this discussion on the web visit

>>>>> https://groups.google.com/d/msgid/quarkus-dev/CAKawoOa18Z46fws4qWEDB068PKJbHPBCLo03L57OZADGunmXUQ%40mail.gmail.com
>>>>> <https://groups.google.com/d/msgid/quarkus-dev/CAKawoOa18Z46fws4qWEDB068PKJbHPBCLo03L57OZADGunmXUQ%40mail.gmail.com?utm_medium=email&utm_source=footer>
>>>>> .

>>>>>
>>>> --
>>>> You received this message because you are subscribed to the Google
>>>> Groups "Quarkus Development mailing list" group.
>>>> To unsubscribe from this group and stop receiving emails from it,
>>>> send
>>>> an email to quarkus-dev...@googlegroups.com.
>>>> To view this discussion on the web visit

>>>> https://groups.google.com/d/msgid/quarkus-dev/CALt0%2Bo9-sYCy%2BT7aCU-Nhc%3D7byJ1Oe2AL9ctZMgrD2KiUcqQ3A%40mail.gmail.com
>>>> <https://groups.google.com/d/msgid/quarkus-dev/CALt0%2Bo9-sYCy%2BT7aCU-Nhc%3D7byJ1Oe2AL9ctZMgrD2KiUcqQ3A%40mail.gmail.com?utm_medium=email&utm_source=footer>
>>>> .

>>>>
>>> --
>>> You received this message because you are subscribed to the Google
>>> Groups
>>> "Quarkus Development mailing list" group.
>>> To unsubscribe from this group and stop receiving emails from it,
>>> send an
>>> email to quarkus-dev...@googlegroups.com.
>>> To view this discussion on the web visit

>>> https://groups.google.com/d/msgid/quarkus-dev/CAJLxjVFjuUuo6dKX9jvAVY6_S9Us%2BXd0F7rJSEF_PipnAtytng%40mail.gmail.com
>>> <https://groups.google.com/d/msgid/quarkus-dev/CAJLxjVFjuUuo6dKX9jvAVY6_S9Us%2BXd0F7rJSEF_PipnAtytng%40mail.gmail.com?utm_medium=email&utm_source=footer>
>>> .

>>>
>>
>
> --
> You received this message because you are subscribed to the Google
> Groups "Quarkus Development mailing list" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to quarkus-dev...@googlegroups.com.
> To view this discussion on the web visit

> https://groups.google.com/d/msgid/quarkus-dev/CAJLxjVGOsRMf8WhxsOrtbwa-Bg4UdAj2ayNM3R3%3DLy83mCZV9A%40mail.gmail.com.


/max
https://xam.dk/about

[Georgios Andrianakis]

It has been corrected in master.

[Loïc MATHIEU]

Hi,

Comparing master with 1.11, dev mode seems to be a little slower 2.3s => 2.5.
I launch the test multiple times and always come to these numbers.

I'll try to produce profiles tomorrow so we can have a look, and I'll also provide extension timing.

[Guillaume Smet]

On Tue, Mar 2, 2021 at 2:09 PM Loïc MATHIEU <loik...@gmail.com> wrote:

Comparing master with 1.11, dev mode seems to be a little slower 2.3s => 2.5.

I launch the test multiple times and always come to these numbers.

I'll try to produce profiles tomorrow so we can have a look, and I'll also provide extension timing.

I'm not totally surprised given we have the Dev UI now. But profiles would say for sure.

[Loïc MATHIEU]

Some more timings on the improvement of fast jar on a different application (reactive messaging kafka) with the following features:

 Installed features: [cdi, jaeger, micrometer, mutiny, smallrye-context-propagation, smallrye-health, smallrye-opentracing, smallrye-reactive-messaging, smallrye-reactive-messaging-kafka, vertx]

Local env
- 1.11 - legacy jar: 1.43s
- 1.12 - fast jar: 1.19s
=> 17% startup improvement
 
K8s container with 256m(req)/512m(limit) memory and 250mi(req)/500mi(limit) CPU
- 1.11 - legacy jar: 14,1s
- 1.12 - fast jar: 11,5s
=> 18% startup improvement

K8s container with 512m(req)/1024m(limit) memory and 500mi(req)/1000mi(limit) CPU
- 1.11 - legacy jar:  7.43s
- 1.12 - fast jar: 5.73s
=> 23% improvement

By the way, reactive messaging is awesome, with very low memory and CPU (256m memory and 250mi CPU) I can handle more than 50k msg/min without issue (with spikes at 250k msg/min).

[Georgios Andrianakis]

Very interesting numbers Loic, thanks!

On Wed, Mar 3, 2021 at 4:48 PM Loïc MATHIEU <loik...@gmail.com> wrote:

Some more timings on the improvement of fast jar on a different application (reactive messaging kafka) with the following features:

 Installed features: [cdi, jaeger, micrometer, mutiny, smallrye-context-propagation, smallrye-health, smallrye-opentracing, smallrye-reactive-messaging, smallrye-reactive-messaging-kafka, vertx]

Local env
- 1.11 - legacy jar: 1.43s
- 1.12 - fast jar: 1.19s
=> 17% startup improvement
 
K8s container with 256m(req)/512m(limit) memory and 250mi(req)/500mi(limit) CPU
- 1.11 - legacy jar: 14,1s
- 1.12 - fast jar: 11,5s
=> 18% startup improvement

K8s container with 512m(req)/1024m(limit) memory and 500mi(req)/1000mi(limit) CPU
- 1.11 - legacy jar:  7.43s
- 1.12 - fast jar: 5.73s
=> 23% improvement

Can you redo this experiment with  500mi(req)/3000mi(limit) ? It would be interesting to see how multiple CPUs at startup affects the boot time vs a single CPU.

[Loïc MATHIEU]

So with 1.12 - fastjar with 500mi(req)/3000(limit) cpu it starts in 2.37s.

So yes, multiple CPU allow very quick startup time, I usually only set limit to 2 times the request, and my request are realy low because I didn't need too much CPU (smallrye reactive message based component only need very low CPU to handle tens of thouthands of message per min). I may reconsider it now :).

Regards,

Loïc

[Georgios Andrianakis]

👍