FWD » Core Development » Runtime Infrastructure

07/06/2022

05:45 PM Support #5567: move FWD to Java 11

Not yet. 5567a has not been merged to 3821c. Greg Shah

05:35 PM Support #5567: move FWD to Java 11

What is the status of this task? Am I able to do devops image building like Ansible and Docker _without_ including Ja... Roger Borrello

06/27/2022

03:10 AM Bug #6084: NIO SSL performance is slower than standard Java SSL

Constantin Asofiei wrote:
> Igor, in Eclipse I get a 'package not found' for this, in @SecureSocketsRegistrar@:
> [... Igor Skornyakov

02:34 AM Bug #6084: NIO SSL performance is slower than standard Java SSL

Igor, in Eclipse I get a 'package not found' for this, in @SecureSocketsRegistrar@:... Constantin Asofiei

06/26/2022

02:05 PM Bug #6084: NIO SSL performance is slower than standard Java SSL

Re-worked Conscrypt support based on the code review. Added JMC counters for @send@, @wrap@, and @unwrap@ operations ... Igor Skornyakov

06/24/2022

02:37 PM Bug #6084: NIO SSL performance is slower than standard Java SSL

Greg Shah wrote:
> Please also implement any JMX instrumentation needed to help us measure the performance of the SS... Igor Skornyakov

02:37 PM Bug #6084: NIO SSL performance is slower than standard Java SSL

Greg Shah wrote:
> Code Review Task Branch 3821c Revision 14001
>
> 1. @BCProbe@ and @BouncyCastle@ should not be... Igor Skornyakov

02:34 PM Bug #6084: NIO SSL performance is slower than standard Java SSL

Please also implement any JMX instrumentation needed to help us measure the performance of the SSL processing. Greg Shah

02:33 PM Bug #6084: NIO SSL performance is slower than standard Java SSL

Code Review Task Branch 3821c Revision 14001
1. @BCProbe@ and @BouncyCastle@ should not be referenced in @Security... Greg Shah

12:19 PM Bug #6084 (Test): NIO SSL performance is slower than standard Java SSL

Hynek Cihlar wrote:
> Code review 6084-1.diff. The changes are good.
Thank you.
Committed to 3821c/14001. Igor Skornyakov

12:11 PM Bug #6084: NIO SSL performance is slower than standard Java SSL

Code review 6084-1.diff. The changes are good. Hynek Cihlar

11:09 AM Bug #6084: NIO SSL performance is slower than standard Java SSL

Hynek Cihlar wrote:
> If you remove the empty branch and add @provider != null@ to the last branch, then @null@ will... Igor Skornyakov

11:04 AM Bug #6084: NIO SSL performance is slower than standard Java SSL

Igor Skornyakov wrote:
> Hynek Cihlar wrote:
> > I see. In that case please remove the empty branch as it adds nois... Hynek Cihlar

10:58 AM Bug #6084: NIO SSL performance is slower than standard Java SSL

Hynek Cihlar wrote:
> I see. In that case please remove the empty branch as it adds noise to the code and do @else i... Igor Skornyakov

10:52 AM Bug #6084: NIO SSL performance is slower than standard Java SSL

Igor Skornyakov wrote:
> Hynek Cihlar wrote:
> > > > @null@ should be handled similarly as the invalid value case. ... Hynek Cihlar

09:58 AM Bug #6084: NIO SSL performance is slower than standard Java SSL

Hynek Cihlar wrote:
> > > @null@ should be handled similarly as the invalid value case. So that the condition is not... Igor Skornyakov

09:44 AM Bug #6084: NIO SSL performance is slower than standard Java SSL

Igor Skornyakov wrote:
> Hynek Cihlar wrote:
> > Code review @6084.diff@.
> >
> > In @ServerDriver.java@:
> >
... Hynek Cihlar

09:30 AM Bug #6084: NIO SSL performance is slower than standard Java SSL

Hynek Cihlar wrote:
> Code review @6084.diff@.
>
> In @ServerDriver.java@:
>
> [...]
>
> @null@ should be h... Igor Skornyakov

08:40 AM Bug #6084: NIO SSL performance is slower than standard Java SSL

Code review @6084.diff@.
In @ServerDriver.java@:... Hynek Cihlar

07:05 AM Bug #6084 (Review): NIO SSL performance is slower than standard Java SSL

Added support for the @Conscrypt@ SSL provider.
I've also changed the way for configuring non-default SSL providers.... Igor Skornyakov

06/23/2022

10:27 AM Support #5567: move FWD to Java 11

> Greg, there are 2 clamscan processes still running on devsrv01... this usually affects ChUI regression testing.
... Greg Shah

10:23 AM Support #5567: move FWD to Java 11

Greg, there are 2 clamscan processes still running on devsrv01... this usually affects ChUI regression testing.
Constantin Asofiei

10:11 AM Support #5567: move FWD to Java 11

The main regression testing should complete in less than 4 hours. Anything more than that suggests many unexpected f... Greg Shah

04:58 AM Support #5567: move FWD to Java 11

For now I have baseline for ctrl-c tests.
I started main regression tests yesterday but they are still running after... Tomasz Domin

06/17/2022

10:35 AM Bug #6529: Incompatible uppercasing.

I've noticed the following incompatibility.
Consider the following code:... Igor Skornyakov

10:33 AM Bug #6529 (New): Incompatible uppercasing.

Igor Skornyakov

06/14/2022

05:26 PM Support #5567: move FWD to Java 11

Feel free to fix the issue in the harness as well. Greg Shah

04:52 PM Support #5567: move FWD to Java 11

Greg Shah wrote:
> You mean this?
>
> [...]
>
> Go ahead and add it everywhere.
>
> > or maybe it's an easy... Roger Borrello

04:27 PM Support #5567: move FWD to Java 11

You mean this?... Greg Shah

03:32 PM Support #5567: move FWD to Java 11

Tomasz Domin wrote:
> I will try to check what is going on by shortening ctrlc0 and all1 scenarios, as there are no ... Roger Borrello

11:39 AM Support #5567: move FWD to Java 11

Tomasz Domin wrote:
> As I've changed linked @run@ directory to @deploy@ directory I have to manualy update @client.... Tomasz Domin

07:06 AM Support #5567: move FWD to Java 11

The old @run/@ directory is gone and will not come back. Any remaining references should be found and fixed. It is ... Greg Shah

06:16 AM Support #5567: move FWD to Java 11

As I've changed linked @run@ directory to @deploy@ directory I have to manualy update @client.sh@ script:
Note that ... Tomasz Domin

06:07 AM Support #5567: move FWD to Java 11

It seems I found an issue in regression tests - runDir is not properly set in tests despite its correctly set in harn... Tomasz Domin

03:38 AM Support #5567: move FWD to Java 11

Main Regression tests were running all night and they are still running, so something is wrong there.
I need to rest... Tomasz Domin

09:13 AM Bug #6084: NIO SSL performance is slower than standard Java SSL

Greg Shah wrote:
> I appreciate the concern. Thank you for noting this, it is important.
>
> On the other hand, ... Igor Skornyakov

08:25 AM Bug #6084: NIO SSL performance is slower than standard Java SSL

I appreciate the concern. Thank you for noting this, it is important.
On the other hand, I want us to have a choi... Greg Shah

03:08 AM Bug #6084: NIO SSL performance is slower than standard Java SSL

I've an impression that @Conscrypt@ is mostly used on the Android platform.
Moreover, it is based on @BoringSSL@. He... Igor Skornyakov

06/13/2022

03:47 PM Support #5567: move FWD to Java 11

Due to issues with regression testing on latest trunk I still need to hold, reported issue in #5034#note-2358 Tomasz Domin

12:04 PM Bug #6084: NIO SSL performance is slower than standard Java SSL

Yes, it is OK. Greg Shah

12:00 PM Bug #6084: NIO SSL performance is slower than standard Java SSL

As far as I understand @Conscrypt@ code is not pure Java - it contains native parts. Is it acceptable?
Thank you. Igor Skornyakov

11:32 AM Bug #6084 (WIP): NIO SSL performance is slower than standard Java SSL

Igor Skornyakov

06/10/2022

10:28 AM Feature #5776: reduce memory requirements for the FWD client

I fired two curls (interval of some seconds) in my terminal and I got:... Alexandru Lungu

09:56 AM Feature #5776: reduce memory requirements for the FWD client

Alexandru Lungu wrote:
> I am running the same version of Gradle. It is odd the fact that when I access https://repo... Tomasz Domin

09:51 AM Feature #5776: reduce memory requirements for the FWD client

I am running the same version of Gradle. It is odd the fact that when I access https://repository.jboss.org/nexus/con... Alexandru Lungu

08:31 AM Feature #5776: reduce memory requirements for the FWD client

Alexandru Lungu wrote:
> I have an issue when using @./gradlew all@ to build FWD:
>
> [...]
>
> Th... Tomasz Domin

08:11 AM Feature #5776: reduce memory requirements for the FWD client

I have an issue when using @./gradlew all@ to build FWD:... Alexandru Lungu

06/09/2022

04:04 PM Support #5567: move FWD to Java 11

Greg Shah wrote:
> I know we have regressions in the ChUI testing, but I suspect those regressions were already ther... Tomasz Domin

07:10 AM Support #5567: move FWD to Java 11

I know we have regressions in the ChUI testing, but I suspect those regressions were already there before 5567a. Do ... Greg Shah

01:45 PM Feature #5776: reduce memory requirements for the FWD client

Tomasz Domin wrote:
> Attached please find corrected @5776-3.diff@
The changes look good.
Hynek Cihlar

01:39 PM Feature #5776: reduce memory requirements for the FWD client

Hynek Cihlar wrote:
> Tomasz Domin wrote:
> > @out.write()@ always sends out full array no matter of length and off... Tomasz Domin

11:38 AM Feature #5776: reduce memory requirements for the FWD client

Tomasz Domin wrote:
> @out.write()@ always sends out full array no matter of length and offset values,
> so this i... Hynek Cihlar

10:47 AM Feature #5776: reduce memory requirements for the FWD client

Thank you Hynek.
> Code review 3821c revision 13944. The changes are good. But my worry is about maintaining the T... Tomasz Domin

09:35 AM Feature #5776: reduce memory requirements for the FWD client

Code review 3821c revision 13944.
@Protocol.java@ and @OutputStreamWrapper.java@ require copyright year to be bump... Hynek Cihlar

09:12 AM Feature #5776: reduce memory requirements for the FWD client

Code review 3821c revision 13944. The changes are good. But my worry is about maintaining the TwelveMonkeys changes. ... Hynek Cihlar

01:21 AM Feature #5776: reduce memory requirements for the FWD client

Greg Shah wrote:
> The directory changes are in relation to #4065. They aren't necessary for this task, per se.
... Tomasz Domin

06/08/2022

01:52 PM Feature #5776: reduce memory requirements for the FWD client

The directory changes are in relation to #4065. They aren't necessary for this task, per se. Greg Shah

01:46 PM Feature #5776: reduce memory requirements for the FWD client

Are there changes to our projects' directories that we should plan on making, in order to benefit from this great enh... Roger Borrello

01:45 PM Feature #5776: reduce memory requirements for the FWD client

Hynek: Please review.
I'm OK with closing the task at this level. We can always look further in the future if nee... Greg Shah

10:55 AM Feature #5776: reduce memory requirements for the FWD client

I checked non-heap memory usage and it is correctly allocated and de-allocated.
Above mentioned 500 GB heap seems to... Tomasz Domin

06:32 AM Feature #5776 (Review): reduce memory requirements for the FWD client

I've run more tests and it seems hotel_chui swing/console client also executes the test correctly when having set "-X... Tomasz Domin

06/07/2022

04:17 PM Feature #5776: reduce memory requirements for the FWD client

Implemented #5776#note-62
Tested with chui_hotel swing client with heap limit set to @-Xmx64m@, configuration as i... Tomasz Domin

06:41 AM Feature #5776: reduce memory requirements for the FWD client

Good plan. Greg Shah

04:49 AM Feature #5776: reduce memory requirements for the FWD client

Regarding writing LOB files input/output.
Reading LOB files is implemented using @InputStreamWrapper@ which reads da... Tomasz Domin

03:28 AM Feature #5776: reduce memory requirements for the FWD client

I've added security configuration and @server-side-resources=memptr@ feature started to work correctly with following... Tomasz Domin