Fixed memory issues with increasing heap size and forcing G1GC.
Do we need all these plugins loaded for inspections?
I've found a 'disable plugin' option in TC Inspections build configuration,
but it is unclear how to disable plugin correctly.
Can someone take over this?
> 46 plugins initialized in 1031 ms
> 2018-12-13 10:55:24,875 [ 1342] INFO - llij.ide.plugins.PluginManager -
> Loaded bundled plugins: Android Support (10.2.3), Ant Support (1.0), CSS
> Support (172.4574.11), Database Tools and SQL (172.4574.11), Eclipse
> Integration (3.0), FreeMarker support (1.0), GWT Support (1.0), Gradle
> (172.4574.11), Groovy (9.0), Guice (8.0), HTML Tools (2.0), Hibernate
> Support (1.0), I18n for Java (172.4574.11), IDEA CORE (172.4574.11),
> IntelliLang (8.0), JBoss Seam Support (1.0), JUnit (1.0), Java EE: Bean
> Validation Support (1.1), Java EE: Contexts and Dependency Injection (1.1),
> Java EE: EJB, JPA, Servlets (1.0), Java EE: Java Server Faces (2.2.X.),
> Java EE: Web Services (JAX-WS) (1.9), Java Server Pages (JSP) Integration
> (1.0), JavaScript Support (1.0), Kotlin (1.1.4-release-IJ2017.2-3), Maven
> Integration (172.4574.11), Persistence Frameworks Support (1.0), Plugin
> DevKit (1.0), Properties Support (172.4574.11), QuirksMode (172.4574.11),
> Spring AOP/@AspectJ (1.0), Spring Batch (1.0), Spring Data (1.0), Spring
> Integration Patterns (1.0), Spring Security (1.0), Spring Support (1.0),
> Spring Web Flow (1.0), Spring Web Services (1.0), Struts 1.x (2.0), Struts
> 2 (1.0), TestNG-J (8.0), UI Designer (172.4574.11), Velocity support (1.0),
> W3C Validators (2.0), WebLogic Integration (1.0), XPathView + XSLT Support
> (4)
Kotlin plugins fails to start, let's disable it.
>
> 2018-12-13 10:55:27,623 [ 4090] INFO - il.indexing.FileBasedIndexImpl -
> Rebuild requested for index
> org.jetbrains.kotlin.idea.versions.KotlinJvmMetadataVersionIndex
> java.lang.Throwable
> at
> com.intellij.util.indexing.FileBasedIndex.requestRebuild(FileBasedIndex.java:68)
> at
> org.jetbrains.kotlin.idea.versions.KotlinUpdatePluginComponent.initComponent(KotlinUpdatePluginComponent.kt:54)
> at
> com.intellij.openapi.components.impl.ComponentManagerImpl$ComponentConfigComponentAdapter.getComponentInstance(ComponentManagerImpl.java:492)
> at
> com.intellij.openapi.components.impl.ComponentManagerImpl.createComponents(ComponentManagerImpl.java:118)
> at
> com.intellij.openapi.application.impl.ApplicationImpl.a(ApplicationImpl.java:462)
> at
> com.intellij.openapi.application.impl.ApplicationImpl.createComponents(ApplicationImpl.java:466)
> at
> com.intellij.openapi.components.impl.ComponentManagerImpl.init(ComponentManagerImpl.java:102)
> at
> com.intellij.openapi.application.impl.ApplicationImpl.load(ApplicationImpl.java:421)
> at
> com.intellij.openapi.application.impl.ApplicationImpl.load(ApplicationImpl.java:407)
> at com.intellij.idea.IdeaApplication.run(IdeaApplication.java:203)
On Thu, Dec 13, 2018 at 1:45 PM Dmitriy Pavlov <dpav...@apache.org> wrote:
> Sure, let's apply. I hope all TC agents may handle 4G heap.
>
> чт, 13 дек. 2018 г. в 12:54, Andrey Mashenkov <andrey.mashen...@gmail.com
> >:
>
> > Guys,
> >
> > I've just creates a copy of Inspections TC build task with GC logs turned
> > on to check if there is any issues
> > and found Inspections task spent too much time in STW due to long Full GC
> > pauses.
> >
> > I've tried to increase Xmx up to 4Gb and use G1GC got 2+ times better
> > execution time down to ~15 min (~17 for 2G heap).
> > Increasing heap size only is not very helpful as it just postpone Full GC
> > issues, but changing GC to G1GC gives noticeable result.
> >
> > Let's apply this optimization.
> > Thoughts?
> >
> >
> > On Sun, Dec 9, 2018 at 12:43 PM Vyacheslav Daradur <daradu...@gmail.com>
> > wrote:
> >
> > > Hi, Maxim, Nikolay, I have the following questions regarding
> inspections:
> > >
> > > Should 'gnite_inspections_teamcity.xml' been imported into IDEA, since
> > > 'ignite_inspections.xml' has been removed in actual master?
> > >
> > > Also, I've faced mismatching: if I use
> > > '@SuppressWarnings("ErrorNotRethrown")' in code, then this will be
> > > marked on TC as "Redundant suppression". If I removed this suppression
> > > in "main" code base (not in tests) then it's fine and IDE does not
> > > mark the code by inspection. But, if I use
> > > 'GridTestUtils#assertThrows' in 'tests' code base, then IDE requires
> > > to suppress the inspection, if I have done it then TC marks this as
> > > "Redundant suppression".
> > >
> > > What should I do in this case?
> > >
> > > On Mon, Dec 3, 2018 at 10:26 PM Andrey Mashenkov
> > > <andrey.mashen...@gmail.com> wrote:
> > > >
> > > > Hi,
> > > >
> > > > Have someone tried to investigate the issue related to Inspection TC
> > task
> > > > execution time variation (from 0.5 up to 1,5 hours)?
> > > > Can we enable GC logs for this task or may be even get CPU, Disk,
> > Network
> > > > metrics?
> > > > Can someone check if there are unnecessary Idea plugins starts that
> can
> > > be
> > > > safely disabled?
> > > >
> > > >
> > > > On Tue, Nov 27, 2018 at 5:52 PM Dmitriy Pavlov <dpav...@apache.org>
> > > wrote:
> > > >
> > > > > I'm totally with you in this decision, let's move the file.
> > > > >
> > > > > вт, 27 нояб. 2018 г. в 16:24, Maxim Muzafarov <maxmu...@gmail.com
> >:
> > > > >
> > > > > > Igniters,
> > > > > >
> > > > > > I propose to make inspection configuration default on the project
> > > > > > level. I've created a new issue [1] for it. It can be easily done
> > and
> > > > > > recommend by IntelliJ documentation [2].
> > > > > > Thoughts?
> > > > > >
> > > > > >
> > > > > > Vyacheslav,
> > > > > >
> > > > > > Can you share an example of your warnings?
> > > > > > Currently, we have different inspection configurations:
> > > > > > - ignite_inspections.xml - to import inspections as default and
> use
> > > it
> > > > > > daily.
> > > > > > - ignite_inspections_teamcity.xml - config to run it on TC. Only
> > > fixed
> > > > > > rules in the project code are enabled. Each of these rules are
> > marked
> > > > > > with ERROR level.
> > > > > >
> > > > > > [1] https://issues.apache.org/jira/browse/IGNITE-10422
> > > > > > [2] https://www.jetbrains.com/help/idea/code-inspection.html
> > > > > > On Tue, 20 Nov 2018 at 13:58, Nikolay Izhikov <
> nizhi...@apache.org
> > >
> > > > > wrote:
> > > > > > >
> > > > > > > Hello, Vyacheslav.
> > > > > > >
> > > > > > > Yes, we have.
> > > > > > >
> > > > > > > Maxim Muzafarov, can you fix it, please?
> > > > > > >
> > > > > > > вт, 20 нояб. 2018 г., 13:10 Vyacheslav Daradur
> > daradu...@gmail.com
> > > :
> > > > > > >
> > > > > > > > Guys, why we have 2 different inspection files in the repo?
> > > > > > > > idea\ignite_inspections.xml
> > > > > > > > idea\ignite_inspections_teamcity.xml
> > > > > > > >
> > > > > > > > AFAIK TeamCity is able to use the same inspection file with
> > IDE.
> > > > > > > >
> > > > > > > > I've imported 'idea\ignite_inspections.xml' in the IDE, but
> now
> > > see
> > > > > > > > inspection warnings for my PR on TC because of different
> rules.
> > > > > > > >
> > > > > > > >
> > > > > > > > On Sun, Nov 11, 2018 at 6:06 PM Maxim Muzafarov <
> > > maxmu...@gmail.com>
> > > > > > > > wrote:
> > > > > > > > >
> > > > > > > > > Yakov, Dmitry,
> > > > > > > > >
> > > > > > > > > Which example of unsuccessful suite execution do we need?
> > > > > > > > > Does the current fail [1] in the master branch enough to
> > > configure
> > > > > > > > > notifications by TC.Bot?
> > > > > > > > >
> > > > > > > > > > Please consider adding more checks
> > > > > > > > > > - line endings. I think we should only have \n
> > > > > > > > > > - ensure blank line at the end of file
> > > > > > > > >
> > > > > > > > > It seems to me that `line endings` is easy to add, but for
> > the
> > > > > `blank
> > > > > > > > > line at the end` we need as special regexp. Can we focus on
> > > > > built-in
> > > > > > > > > IntelliJ inspections at first and fix others special
> further?
> > > > > > > > >
> > > > > > > > > [1]
> > > > > > > >
> > > > > >
> > > > >
> > >
> >
> https://ci.ignite.apache.org/viewType.html?buildTypeId=IgniteTests24Java8_InspectionsCore&branch_IgniteTests24Java8=%3Cdefault%3E&tab=buildTypeStatusDiv
> > > > > > > > > On Sun, 11 Nov 2018 at 17:55, Maxim Muzafarov <
> > > maxmu...@gmail.com>
> > > > > > > > wrote:
> > > > > > > > > >
> > > > > > > > > > Igniters,
> > > > > > > > > >
> > > > > > > > > > Since the inspection rules are included in RunAll a few
> > > members
> > > > > of
> > > > > > the
> > > > > > > > > > community mentioned a wide distributed execution time on
> TC
> > > > > agents:
> > > > > > > > > > - 1h:27m:38s publicagent17_9094
> > > > > > > > > > - 38m:04s publicagent17_9094
> > > > > > > > > > - 33m:29s publicagent17_9094
> > > > > > > > > > - 17m:13s publicagent17_9094
> > > > > > > > > > It seems that we should configure the resources
> > distribution
> > > > > > across TC
> > > > > > > > > > containers. Can anyone take a look at it?
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > I've also prepared the short list of rules to work on:
> > > > > > > > > > + Inconsistent line separators (6 matches)
> > > > > > > > > > + Problematic whitespace (4 matches)
> > > > > > > > > > + expression.equals("literal")' rather than
> > > > > > > > > > '"literal".equals(expression) (53 matches)
> > > > > > > > > > + Unnecessary 'null' check before 'instanceof' expression
> > or
> > > call
> > > > > > (42
> > > > > > > > matches)
> > > > > > > > > > + Redundant 'if' statement (69 matches)
> > > > > > > > > > + Redundant interface declaration (28 matches)
> > > > > > > > > > + Double negation (0 matches)
> > > > > > > > > > + Unnecessary code block (472 matches)
> > > > > > > > > > + Line is longer than allowed by code style (2614
> matches)
> > > (Is it
> > > > > > > > > > possible to implement?)
> > > > > > > > > >
> > > > > > > > > > WDYT?
> > > > > > > > > >
> > > > > > > > > > On Fri, 26 Oct 2018 at 23:43, Dmitriy Pavlov <
> > > > > > dpavlov....@gmail.com>
> > > > > > > > wrote:
> > > > > > > > > > >
> > > > > > > > > > > Hi Maxim,
> > > > > > > > > > >
> > > > > > > > > > > thank you for your efforts to make this happen. Keep
> the
> > > pace!
> > > > > > > > > > >
> > > > > > > > > > > Could you please provide an example of how Inspections
> > can
> > > > > fail,
> > > > > > so
> > > > > > > > I or
> > > > > > > > > > > another contributor could implement support of these
> > > failures
> > > > > > > > validation in
> > > > > > > > > > > the Tc Bot.
> > > > > > > > > > >
> > > > > > > > > > > Sincerely,
> > > > > > > > > > > Dmitriy Pavlov
> > > > > > > > > > >
> > > > > > > > > > > пт, 26 окт. 2018 г. в 18:27, Yakov Zhdanov <
> > > > > yzhda...@apache.org
> > > > > > >:
> > > > > > > > > > >
> > > > > > > > > > > > Maxim,
> > > > > > > > > > > >
> > > > > > > > > > > > Thanks for response, let's do it the way you
> suggested.
> > > > > > > > > > > >
> > > > > > > > > > > > Please consider adding more checks
> > > > > > > > > > > > - line endings. I think we should only have \n
> > > > > > > > > > > > - ensure blank line in the end of file
> > > > > > > > > > > >
> > > > > > > > > > > > All these are code reviews issues I pointed out many
> > > times
> > > > > when
> > > > > > > > reviewing
> > > > > > > > > > > > conributions. It would be cool if we have TC build
> > > failing if
> > > > > > > > there is any.
> > > > > > > > > > > >
> > > > > > > > > > > > Thanks!
> > > > > > > > > > > >
> > > > > > > > > > > > --Yakov
> > > > > > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > --
> > > > > > > > Best Regards, Vyacheslav D.
> > > > > > > >
> > > > > >
> > > > >
> > > >
> > > >
> > > > --
> > > > Best regards,
> > > > Andrey V. Mashenkov
> > >
> > >
> > >
> > > --
> > > Best Regards, Vyacheslav D.
> > >
> >
> >
> > --
> > Best regards,
> > Andrey V. Mashenkov
> >
>
--
Best regards,
Andrey V. Mashenkov
