'make check' I believe. 

 

On Sep 29, 2010, at 7:05 PM, Dave Viner <[email protected]> wrote:

> Will do.
> 
> How do I run the perl unit-tests?  I tried the following:
> 
> % cd thrift-trunk/lib/perl
> % perl Makefile.PL
> % make
> % make test
> 
> But this doesn't seem to find the tests:
> 
> % make test
> PERL_DL_NONLAZY=1 /usr/bin/perl "-Iblib/lib" "-Iblib/arch" test.pl
> Files=0, Tests=0,  0 wallclock secs ( 0.00 usr +  0.00 sys =  0.00 CPU)
> Result: NOTESTS
> %
> 
> I want to make sure that my patch doesn't break existing tests, as well as
> expand the unit test to cover the scenario I'm adding.
> 
> Thanks
> Dave Viner
> 
> 
> 
> On Wed, Sep 29, 2010 at 2:45 AM, Jake Luciani <[email protected]> wrote:
> 
>> Hi Dave
>> 
>> No reason why we shouldn't check it. Could you please open a ticket? If you
>> have a patch that would be great.
>> 
>> Thanks.
>> 
>> Jake
>> 
>> 
>> 
>> On Sep 29, 2010, at 3:13 AM, Dave Viner <[email protected]> wrote:
>> 
>>> Hi all,
>>> 
>>> I'm using Thrift inside Cassandra, and I'm seeing conditions where I
>> receive
>>> lots of these error messages:
>>> 
>>> problem running get_slice: $VAR1 = bless( {
>>>                'code' => 0,
>>>                'message' => 'TSocket: timed out reading 4 bytes from
>>> server-ip:9160'
>>>              }, 'Thrift::TException' );
>>> 
>>> get_slice is the cassandra call.  The TSocket exception is thrown from
>>> 
>> http://svn.apache.org/repos/asf/incubator/thrift/trunk/lib/perl/lib/Thrift/Socket.pm
>>> 
>>> The issue seems to be in the handling of read(), readAll(), and write().
>>> The code checks to see if the $self->{'handle'} variable (which is an
>>> IO::Select object) is defined.  However, it seems possible for an
>> IO::Select
>>> object to be defined, but to have no handles that are valid or connected.
>>> It seems that the code could be a bit more robust in checking for a
>>> condition where the handle is not open, and throw a more explicit
>> exception.
>>> Then it can check for data in the Exception handle.
>>> 
>>> I'm happy to create a patch for it, but before I do so, is there a reason
>>> why the code doesn't currently check for this error condition?
>>> 
>>> Thanks
>>> Dave Viner
>> 

Reply via email to