Re: Any feedback for MNG-6261?

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
5 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Any feedback for MNG-6261?

Dawid Weiss
I added a comment on that issue. The problem is not deterministic for
some reason -- I can execute identical maven on identical code on the
same computer and get two different outcomes (consistenly). I have no
idea how to debug it better too -- passing -X doesn't yield any
reasonable logs. Looks very strange.

Dawid


On Thu, Jul 27, 2017 at 7:37 PM, Robert Scholte <[hidden email]> wrote:

> I had a look at it, but can't reproduce the issue with the attached project
> Even when using exactly the same directory
> (D:\repositories\carrotsearch.com\lingo4g) my build succeeds.
> A failing project is really required to fix this.
>
> Robert
>
>
> On Thu, 27 Jul 2017 13:41:16 +0200, Dawid Weiss <[hidden email]>
> wrote:
>
>> Hello,
>>
>> Just wanted to hear if anybody has any idea about MNG-6261 I filed
>> recently -- there is a relatively simple repro attached and it fails
>> with Maven 3.5.0+; I've been wondering if it's a bug or an illegal
>> abuse of the submodule/parent pom relationship.
>>
>> Dawid
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: [hidden email]
>> For additional commands, e-mail: [hidden email]
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]
>

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Any feedback for MNG-6261?

rfscholte
There must be a difference between the 2 projects.
But as long as there's no attachment which reproduces the issue, we can't  
help. Ensure that you've seen the attached project fail.

Just some suggestions to help:
- rename the directory (for backup), do a fresh checkout and try to do it  
again.
- try to follow the path as displayed in the exception.  It shows all  
required Maven coordinates and even the precise location in the pom.xml
- try to run it at different (valid) locations within the multimodule  
project.

Robert

On Thu, 27 Jul 2017 22:07:24 +0200, Dawid Weiss <[hidden email]>  
wrote:

> I added a comment on that issue. The problem is not deterministic for
> some reason -- I can execute identical maven on identical code on the
> same computer and get two different outcomes (consistenly). I have no
> idea how to debug it better too -- passing -X doesn't yield any
> reasonable logs. Looks very strange.
>
> Dawid
>
>
> On Thu, Jul 27, 2017 at 7:37 PM, Robert Scholte <[hidden email]>  
> wrote:
>> I had a look at it, but can't reproduce the issue with the attached  
>> project
>> Even when using exactly the same directory
>> (D:\repositories\carrotsearch.com\lingo4g) my build succeeds.
>> A failing project is really required to fix this.
>>
>> Robert
>>
>>
>> On Thu, 27 Jul 2017 13:41:16 +0200, Dawid Weiss <[hidden email]>
>> wrote:
>>
>>> Hello,
>>>
>>> Just wanted to hear if anybody has any idea about MNG-6261 I filed
>>> recently -- there is a relatively simple repro attached and it fails
>>> with Maven 3.5.0+; I've been wondering if it's a bug or an illegal
>>> abuse of the submodule/parent pom relationship.
>>>
>>> Dawid
>>>
>>> ---------------------------------------------------------------------
>>> To unsubscribe, e-mail: [hidden email]
>>> For additional commands, e-mail: [hidden email]
>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: [hidden email]
>> For additional commands, e-mail: [hidden email]
>>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Any feedback for MNG-6261?

Thomas Matthijs
In reply to this post by Dawid Weiss
Have you double checked file permissions? if it can't read the parent
pom (maybe not all dirs are +x etc) then it wil have the behavior you
see

On Thu, Jul 27, 2017 at 10:07 PM, Dawid Weiss <[hidden email]> wrote:

> I added a comment on that issue. The problem is not deterministic for
> some reason -- I can execute identical maven on identical code on the
> same computer and get two different outcomes (consistenly). I have no
> idea how to debug it better too -- passing -X doesn't yield any
> reasonable logs. Looks very strange.
>
> Dawid
>
>
> On Thu, Jul 27, 2017 at 7:37 PM, Robert Scholte <[hidden email]> wrote:
>> I had a look at it, but can't reproduce the issue with the attached project
>> Even when using exactly the same directory
>> (D:\repositories\carrotsearch.com\lingo4g) my build succeeds.
>> A failing project is really required to fix this.
>>
>> Robert
>>
>>
>> On Thu, 27 Jul 2017 13:41:16 +0200, Dawid Weiss <[hidden email]>
>> wrote:
>>
>>> Hello,
>>>
>>> Just wanted to hear if anybody has any idea about MNG-6261 I filed
>>> recently -- there is a relatively simple repro attached and it fails
>>> with Maven 3.5.0+; I've been wondering if it's a bug or an illegal
>>> abuse of the submodule/parent pom relationship.
>>>
>>> Dawid
>>>
>>> ---------------------------------------------------------------------
>>> To unsubscribe, e-mail: [hidden email]
>>> For additional commands, e-mail: [hidden email]
>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: [hidden email]
>> For additional commands, e-mail: [hidden email]
>>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]
>

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Any feedback for MNG-6261?

Dawid Weiss
In reply to this post by rfscholte
> There must be a difference between the 2 projects.

There is no difference. It's the same repo on the same commit, I
called "git clean -xfd ." prior to executing those maven commands (and
obviously there are no other changes).

> But as long as there's no attachment which reproduces the issue, we can't
> help. Ensure that you've seen the attached project fail.

I know, Robert. But you have to trust me that it does fail (non-predictably).

> - rename the directory (for backup), do a fresh checkout and try to do it again.

git clean -xfd .

> - try to follow the path as displayed in the exception.  It shows all
> required Maven coordinates and even the precise location in the pom.xml

It's an identical (relative path); the resolution of those relative
paths is broken somehow.

Where is the Maven code (on github) that's responsible for this pom
resolution? Adding -X doesn't really yield any useful logs as to where
it tries to locate those relative parent POMs, prior to just breaking.
I'll take a look at it, but it'd save me some time if you could point
me at the code that performs this.

Dawid

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Any feedback for MNG-6261?

rfscholte
In reply to this post by Thomas Matthijs
Nice catch. Looks a lot like MNG-6262

On Fri, 28 Jul 2017 14:12:19 +0200, Dawid Weiss <[hidden email]>  
wrote:

>> IIUC the failing project isn't always failing. This could be an  
>> explanation.
>
> I didn't say it isn't always failing -- I said it is failing
> deterministically (as in: always) from the same location, but
> sometimes isn't failing on the same project when executed from
> *another* location.
>
> I know what the problem is and I attached a longer comment on the
> issue, please take a look. Don't know if we should discuss it here or
> in Jira, up to you.
>
> Dawid
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Loading...