[erlang-questions] epgsql coordination?

Anton Lebedevich <>
Thu Feb 13 18:25:55 CET 2014


Did you add lines mentioned in the beginning of test_schema.sql to
pg_hba.conf ?
You'll need to set up ssl authentification for you server using crt and
keys from test_data/.

On 02/13/2014 09:07 PM, David Welton wrote:
> The output I'm getting, after having run test_schema.sql is:
> 
> pgsql_tests: all_test_ (connect_test(pgsql))...*failed*
> in function pgsql_tests:connect_only/2 (test/pgsql_tests.erl, line 711)
> **error:{badmatch,{error,invalid_password}}
> 
> 
> pgsql_tests: all_test_ (connect_to_db_test(pgsql))...*failed*
> in function pgsql_tests:connect_only/2 (test/pgsql_tests.erl, line 711)
> **error:{badmatch,{error,invalid_password}}
> 
> 
> pgsql_tests: all_test_ (connect_as_test(pgsql))...*failed*
> in function pgsql_tests:connect_only/2 (test/pgsql_tests.erl, line 711)
> **error:{badmatch,{error,invalid_password}}
> 
> 
> pgsql_tests: all_test_ (connect_with_ssl_test(pgsql))...*skipped*
> undefined
> *unexpected termination of test process*
> ::{ssl_negotiation_failed,ssl_not_started}
> 
> =======================================================
>   Failed: 3.  Skipped: 0.  Passed: 4.
> One or more tests were cancelled.
> 
> =ERROR REPORT==== 13-Feb-2014::18:07:13 ===
> ** Generic server <0.57.0> terminating
> ** Last message in was {connect,"localhost","epgsql_test",[],
>                                 [{port,5432},
>                                  {database,"epgsql_test_db1"},
>                                  {ssl,true}]}
> ** When Server state == {state,undefined,undefined,<<>>,undefined,undefined,
>                                {[],[]},
>                                undefined,[],[],[],[],[],[],undefined,
>                                undefined}
> ** Reason for termination ==
> ** {ssl_negotiation_failed,ssl_not_started}
> 
> On Thu, Feb 13, 2014 at 5:23 PM, Anton Lebedevich <> wrote:
>> On 02/13/2014 07:51 PM, David Welton wrote:
>>>> Does it make sense to have it shown as a fork? I would suggest pushing a new repo there instead of forking from davidw :)
>>>
>>> I'm not sure how to accomplish that, and to be honest, it is a fork of
>>> my repository which I had forked from mabrek's.  It'd be cleaner to
>>> post his as where it was forked from.
>>>
>>> I think the first order of business is to see if I can get the tests
>>> he had in there working 100%.  With tests, it gets easier to move
>>> forward without too many worries...
>>
>> Timeout tests are broken in my fork (mabrek/epgsql) because I haven't
>> implemented timeout support yet. Client side timeouts turned out to be
>> quite complicated in asynchronous model. Maybe we need to steal ideas
>> from semiocast/pgsql and use server-side timeouts?
>>
>> These tests pass in original (wg/epgsql) but timeout support is not
>> complete there too (there are situations when connection could stuck
>> after timeout).
>>
>> Regards,
>> Anton Lebedevich.
>> _______________________________________________
>> erlang-questions mailing list
>> 
>> http://erlang.org/mailman/listinfo/erlang-questions
> 
> 
> 



More information about the erlang-questions mailing list