cloudstack-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Hugo Trippaers <trip...@gmail.com>
Subject Re: Findbugs report on 4.3-forward
Date Mon, 27 Jan 2014 19:58:48 GMT
Hey David,

As always you are completely right. I feel a bit bad about getting the findbugs report through
at the last moment. However the coverity reports (with 6000+ issues) didn't make it easy to
find what really needed fixing.

However the fixes we made based on findbugs are pretty nasty. Especially the == vs equals
can have some serious impact in how things behave.

From my end I'm tackling this as follows. We (Schuberg) have a limited functional test environment
that runs the basic create zone with zen test. I have manual tests that will cover Nicira
support for advanced zones with VR, VPC and NVP router. I won't vote +1 before these tests
complete.

Maybe some other folks can respond with what tests they are doing pre release so we have an
idea on what is covered when somebody gives a +1. Actually I would rather have the report
now, any test findings now will save animesh quite some time in doing all the work to cut
the next RC.

Cheers,

Hugo 

Sent from my iPhone

> On 27 jan. 2014, at 20:46, David Nalley <david@gnsa.us> wrote:
> 
> So just curious if I am the only one concerned about a ton of fixes
> going in at the last minute. If the fixes are for serious bugs and we
> have consensus around their severity being high enough, indeed lets
> fix things. My concern is that much of the QA we do is manual; and
> while we are getting better; fixing tons of things at the last minute
> may have unintended consequences that we don't know about and won't
> easily find.
> 
> I yearn for the day when our automated testing is broad enough that we
> can do fixes right up to the wire and know that things still work, I
> am just not sure that I have confidence that we are there yet.
> Thoughts? I am being paranoid?
> 
> --David
> 
>> On Mon, Jan 27, 2014 at 3:11 AM, Daan Hoogland <daan.hoogland@gmail.com> wrote:
>> Animesh, I commented the once i made yesterday with findbugs:
>> 
>> I allready send a few and will get you a list of the rest later today.
>> 
>> regards,
>> 
>> On Mon, Jan 27, 2014 at 3:48 AM, Animesh Chaturvedi
>> <animesh.chaturvedi@citrix.com> wrote:
>>> Good job fellas. I see a number of commits 20+ into 4.3-forward branch. Are their
specific commits you want me to pick up out of these?
>>> 
>>> Animesh
>>> 
>>> -----Original Message-----
>>> From: Daan Hoogland [mailto:daan.hoogland@gmail.com]
>>> Sent: Sunday, January 26, 2014 2:41 AM
>>> To: dev
>>> Subject: Re: Findbugs report on 4.3-forward
>>> 
>>> I didn't get very far last night and will be looking at the server package again
this afternoon.
>>> 
>>> bon app├ętit,
>>> 
>>>> On Sun, Jan 26, 2014 at 1:36 AM, Ian Duffy <ian@ianduffy.ie> wrote:
>>>> Hi,
>>>> 
>>>> Fixed the issues highlighted in the ldap user authentication package.
>>>> 
>>>> Have pushed to 4.3-forward.
>>>> 
>>>> Thanks,
>>>> Ian
>>>> 
>>>> 
>>>> On 25 January 2014 22:26, Daan Hoogland <daan.hoogland@gmail.com> wrote:
>>>> 
>>>>>> or reply to this mail with the filename you are working on
>>>>> I'll be looking at the server package as it seems to contain the most
>>>>> issues.
>>>>> 
>>>>>> On Sat, Jan 25, 2014 at 4:00 PM, Hugo Trippaers <hugo@trippaers.nl>
wrote:
>>>>>> I've also added a job to master with the Findbugs report and the
>>>>> cobertura code coverage report.
>>>>>> 
>>>>>> Good stuff, we have a 12% coverage of our classes with unit tests.
>>>>>> Huge
>>>>> improvement over the last release where we had 4% iirc. We have 306
>>>>> reports from Findbugs, of which the majority are internationalization
issues.
>>>>> (String.getBytes without charset mostly). On the coverity site we
>>>>> have
>>>>> 6000+ issues still open, but at least that number is relatively
>>>>> 6000+ stable, we
>>>>> fix as much issues as we introduce and it's untuned so we can assume
>>>>> a large number of false positives there.
>>>>>> 
>>>>>> I think that on average the automated tools tell us that code
>>>>>> quality is
>>>>> improving, which a good thing. Combined with the functional testing
>>>>> and the simulator build we can prove that we are doing quite well on
>>>>> the code quality angle.
>>>>>> 
>>>>>> http://jenkins.buildacloud.org/job/build-master-slowbuild/
>>>>>> 
>>>>>> 
>>>>>> Cheers,
>>>>>> 
>>>>>> Hugo
>>>>>> 
>>>>>> 
>>>>>>> On 25 jan. 2014, at 14:13, Daan Hoogland <daan.hoogland@gmail.com>
>>>>>> wrote:
>>>>>> 
>>>>>>> H Hugo,
>>>>>>> 
>>>>>>> I'll spend some time on it tonight. Do you have a work load
>>>>>>> distribution scheme or is it random access?
>>>>>>> ;)
>>>>>>> 
>>>>>>> regards
>>>>>>> 
>>>>>>> On Sat, Jan 25, 2014 at 12:39 PM, Hugo Trippaers
>>>>>>> <hugo@trippaers.nl>
>>>>> wrote:
>>>>>>>> Hey all,
>>>>>>>> 
>>>>>>>> I've made Jenkins run the findbugs analysis on 4.3-forward.
Is
>>>>>>>> there
>>>>> somebody who is willing to help triage the findings? Maybe there is
>>>>> some stuff that we need to fix?
>>>>>>>> 
>>>>>>>> the url is
>>>>> http://jenkins.buildacloud.org/job/cloudstack-4.3-forward-maven-build
>>>>> /3/findbugsResult/
>>>>>>>> 
>>>>>>>> Cheers,
>>>>>>>> 
>>>>>>>> Hugo
>>>>> 

Mime
View raw message