Re: Probleme with installation

[Ryan J Ollos <rj...@apache.org>]

On Thu, Feb 20, 2014 at 6:19 PM, Ryan Ollos <ry...@wandisco.com> wrote:

> On Tue, Feb 11, 2014 at 12:23 AM, Ryan Ollos <ry...@wandisco.com>
> wrote:
>
>> On Mon, Feb 10, 2014 at 10:31 AM, Guillaume Meurice <
>> guillaume.meurice@gustaveroussy.fr> wrote:
>>
>>>
>>> On Mon, Feb 10, 2014 at 10:18 AM, Guillaume Meurice <
>>> guillaume.meurice@gustaveroussy.fr> wrote:
>>>
>>>> Thanks for quick reply.
>>>>
>>>> Unfortunately, the link you gave me seems corrupted (I haven't been
>>>> able to untar the archive.
>>>> So I've downloaded the tar.gz from here :
>>>> https://archive.apache.org/dist/bloodhound/
>>>>
>>>
>>> That is strange. I just pulled from one of the mirrors and had not
>>> problem extracting the archive. The checksum is correct too. Maybe you are
>>> just having trouble pulling from a particular mirror.
>>>
>>>
>>> yes probably.
>>>
>>>
>>>
>>> and re-reun the installation process.
>>>>
>>>> Still, I've been stuck at the same step :
>>>> ===============================================================
>>>> (bloodhound)[meurice@kototama : installer] $ python
>>>> bloodhound_setup.py
>>>>
>>>> This installer is able to install Apache Bloodhound with either SQLite
>>>> or
>>>> PostgreSQL databases. SQLite is an easier option for installing
>>>> Bloodhound as
>>>> SQLite is usually built into Python and also requires no special
>>>> permissions to
>>>> run. However, PostgreSQL is generally expected to be more robust for
>>>> production
>>>> use.
>>>> Do you want to install to a PostgreSQL database [Y/n]: n
>>>>
>>>> Please supply a username for the admin user [admin]: admin
>>>> Enter a new password for "admin":
>>>> Please reenter the password:
>>>> Initenv for
>>>> '/Users/meurice/Downloads/apache-bloodhound-0.7/installer/bloodhound/environments/main'
>>>> failed.
>>>> Wrong number of arguments: 3
>>>> Traceback (most recent call last):
>>>>   File "bloodhound_setup.py", line 458, in <module>
>>>>     bsetup.setup()
>>>>   File "bloodhound_setup.py", line 244, in setup
>>>>     self.writeconfig(baseini, [base_config, accounts_config])
>>>>   File "bloodhound_setup.py", line 313, in writeconfig
>>>>     config.save()
>>>>   File "build/bdist.macosx-10.6-universal/egg/trac/config.py", line
>>>> 237, in save
>>>>   File "build/bdist.macosx-10.6-universal/egg/trac/util/__init__.py",
>>>> line 171, in __init__
>>>>   File
>>>> "/System/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/tempfile.py",
>>>> line 293, in mkstemp
>>>>     return _mkstemp_inner(dir, prefix, suffix, flags)
>>>>   File
>>>> "/System/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/tempfile.py",
>>>> line 228, in _mkstemp_inner
>>>>     fd = _os.open(file, flags, 0600)
>>>> OSError: [Errno 2] No such file or directory:
>>>> '/Users/meurice/Downloads/apache-bloodhound-0.7/installer/bloodhound/environments/main/conf/base.ini-aryq5k'
>>>> (bloodhound)[meurice@kototama : installer] $
>>>> ===============================================================
>>>>
>>>
>>> I'm not sure what is going on with that garbage being appended to the
>>> base.ini filename. I guess you are on OSX platform?
>>>
>>>
>>> yes, I'm under OSX 10.6.8.
>>>
>>> I'll investigate and get back to you.
>>>
>>>
>>>
>>> many thanks !!
>>>
>>> Bests,
>>>
>>>
>>
>> I hadn't looked closely enough at the traceback earlier. The "garbage"
>> being appended to the file is due to a temporary file being created in an
>> atomic save operation. In 0.8dev we've added (r1539120) a SystemExit in the
>> case of errors while creating the environment, which should give a cleaner
>> traceback.
>>
>> It seems like the significant part of the traceback is "Wrong number of
>> arguments: 3". We've seen an issue like that in the past due to spaces in
>> the path,
>> https://issues.apache.org/bloodhound/ticket/459
>>
>> However, space in the path doesn't seem to be an issue here. I really
>> don't have any ideas of what the problem could be.
>>
>> Would you add the following print statement in bloodhound_setup.py? Maybe
>> that will give us a hint.
>>
>> --- bloodhound_setup.py.old 2014-02-11 00:18:38.860305578 -0800
>> +++ bloodhound_setup.py 2014-02-11 00:18:44.452397897 -0800
>> @@ -212,7 +212,6 @@
>>          trac = TracAdmin(os.path.abspath(new_env))
>>          if not trac.env_check():
>>              try:
>> +                print options
>>                  trac.do_initenv('%(project)s %(db)s '
>>                                  '%(repo_type)s %(repo_path)s '
>>                                  '--inherit=%(inherit)s '
>>
>>
> Hi Guillaume, did you have an success, or are you able to provide any more
> info? I don't have access to OSX to reproduce any potential issues, so I'm
> particularly interested to see if we can work through this one with you.
>
> I've cc'ed you on the email in case you aren't subscribed to the users
> mailing list, and in case you missed my previous follow-up.
>


Another instance of this issue was reported on StackOverflow (1). It could
be an issue only with an earlier version of Bloodhound. Bloodhound 0.7 is
used here and in the SO thread. I haven't been able to reproduce with the
trunk.

The issue was encountered some time ago (2), but in that case it was due to
specifying repo_type and an empty repo_dir. That condition is checked for
in the latest version of the script and it does not appear that is the
issue here or in the SO thread.

(1) http://stackoverflow.com/q/25403662/121694
(2) http://markmail.org/message/qxlpshxdpyw6knfw