maven-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Martin Todorov (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (MNG-6604) Intermittent failures while downloading GAVs from Nexus
Date Thu, 04 Jul 2019 18:53:00 GMT

    [ https://issues.apache.org/jira/browse/MNG-6604?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16878856#comment-16878856
] 

Martin Todorov commented on MNG-6604:
-------------------------------------

In this day and age saying the such a widely used tool for building code like Maven will not
support multiple VM access will be madness. I'm not a core Maven developer, but I've been
using it and developing plugins and tooling around it since it first popped up many years
ago. CI build farms need this to be properly implemented.

I think that we need to focus on implementing a proper locking mechanism that works for multiple
JVM-s.

I get your point about the complexity of fine-tuning this. I don't think one should worry
about large files, as much as tracking inactivity on files (a JVM may have terminated and
left a file behind and then lock things for other JVM-s; if this has happened an easy way
to check it would be to simply see when the last file write activity happened and see, if
the {{maxInactivity}} time has been exceeded). 

What are the classes of importance, if somebody would like to dig into this?


> Intermittent failures while downloading GAVs from Nexus
> -------------------------------------------------------
>
>                 Key: MNG-6604
>                 URL: https://issues.apache.org/jira/browse/MNG-6604
>             Project: Maven
>          Issue Type: Bug
>          Components: Command Line, Toolchains
>    Affects Versions: 3.6.0
>         Environment: Nexus OSS 3.15.2-01
> Docker 18.09.2 on Ubuntu 18.04.2 LTS
> Gitlab runner 11.8.0
>            Reporter: Ivan Rizzante
>            Priority: Major
>         Attachments: docker-env.txt, log.txt
>
>
> Hello
> we're running maven 3.6.0 builds in a docker container and we use Nexus OSS configured
as proxy for Maven Central.
> While running our builds using Gitlab CI, we're experiencing intermittent build failures
because Maven cannot find artifacts in Nexus which we verified they are actually available.
> Error example below:
> {noformat}
> 20744 [main] [ERROR] Failed to execute goal on project EcotransitWSClient: Could not
resolve dependencies for project it.sdb.ecotransit:EcotransitWSClient:jar:7.0.1-SNAPSHOT:
Could not transfer artifact commons-beanutils:commons-beanutils:jar:1.9.3 from/to nexus (http://maven-repo.sdb.it:8081/repository/maven-public/):
/builds/sdb/webportal/.m2/repository/commons-beanutils/commons-beanutils/1.9.3/commons-beanutils-1.9.3.jar.part
(No such file or directory) -> [Help 1]
> {noformat}
> I attached the full maven build log and our Docker env settings.
> We tried disabling the keep alive and also disabling the connection pooling but nothing
seems to fix the issue.
>  
>  



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Mime
View raw message