I've recently been trying to get the pig contrib project working, and
as far as I can tell I have it all set up correctly, but when pig
executes the job, it shows 0 rows loaded from the table, even though
there's 50K rows in the table.  I used the helloworld example to
populate a hellotable.  The load line I'm using in Pig is

DATA = LOAD 'accumulo://hellotable?instance=myinst&password=mypass&zookeepers=namenode&columns=f:cnt'
using org.apache.accumulo.pig.AccumuloStorage() AS (row, cf, cq, cv,
s, val);

Thanks,
--
Chris

On Wed, Jan 23, 2013 at 5:59 AM, Chris Sigman <[EMAIL PROTECTED]> wrote:

> I've recently been trying to get the pig contrib project working, and
> as far as I can tell I have it all set up correctly, but when pig
> executes the job, it shows 0 rows loaded from the table, even though
> there's 50K rows in the table.  I used the helloworld example to
> populate a hellotable.  The load line I'm using in Pig is
>
> DATA = LOAD
> 'accumulo://hellotable?instance=myinst&password=mypass&zookeepers=namenode&columns=f:cnt'
> using org.apache.accumulo.pig.AccumuloStorage() AS (row, cf, cq, cv,
> s, val);
>

I think the "&columns=f:cnt" part of the URL is specifying a column
family:column qualifier to fetch.  If you loaded the data with the
helloworld example, its columns are of the form "colfam:colqual_#".  I
think you should just be able to remove "&columns=f:cnt" from the statement
and it will scan all columns.

Billie

>
> Thanks,
> --
> Chris
>

Thanks Billie, that did it! I didn't know what that was from the
readme that came with the accumulo-pig contrib, so I just kept it
there.  Thanks everyone for the help!

--
Chris
On Thu, Jan 24, 2013 at 1:38 PM, Billie Rinaldi <[EMAIL PROTECTED]> wrote:
> On Wed, Jan 23, 2013 at 5:59 AM, Chris Sigman <[EMAIL PROTECTED]> wrote:
>>
>> I've recently been trying to get the pig contrib project working, and
>> as far as I can tell I have it all set up correctly, but when pig
>> executes the job, it shows 0 rows loaded from the table, even though
>> there's 50K rows in the table.  I used the helloworld example to
>> populate a hellotable.  The load line I'm using in Pig is
>>
>> DATA = LOAD
>> 'accumulo://hellotable?instance=myinst&password=mypass&zookeepers=namenode&columns=f:cnt'
>> using org.apache.accumulo.pig.AccumuloStorage() AS (row, cf, cq, cv,
>> s, val);
>
>
> I think the "&columns=f:cnt" part of the URL is specifying a column
> family:column qualifier to fetch.  If you loaded the data with the
> helloworld example, its columns are of the form "colfam:colqual_#".  I think
> you should just be able to remove "&columns=f:cnt" from the statement and it
> will scan all columns.
>
> Billie
>
>
>>
>>
>> Thanks,
>> --
>> Chris
>
>

could it be a visibility issue?

On Wed, Jan 23, 2013 at 8:59 AM, Chris Sigman <[EMAIL PROTECTED]> wrote:
> when pig executes the job, it shows 0 rows loaded from the table

No, I don't think so. The first couple dozen rows at least don't have
any auths set.

--
Chris
On Wed, Jan 23, 2013 at 10:51 AM, David Medinets
<[EMAIL PROTECTED]> wrote:
> could it be a visibility issue?
>
> On Wed, Jan 23, 2013 at 8:59 AM, Chris Sigman <[EMAIL PROTECTED]> wrote:
>> when pig executes the job, it shows 0 rows loaded from the table

A little more info in case it helps, I'm connecting using the same
user (root) as the data was inserted with.  I can't imagine this, but
there aren't potentially any issues because there aren't any auths on
anything? I can scan the data just fine, so I'm rather befuddled by
it.

--
Chris
On Wed, Jan 23, 2013 at 11:55 AM, Chris Sigman <[EMAIL PROTECTED]> wrote:
> No, I don't think so. The first couple dozen rows at least don't have
> any auths set.
>
> --
> Chris
>
>
> On Wed, Jan 23, 2013 at 10:51 AM, David Medinets
> <[EMAIL PROTECTED]> wrote:
>> could it be a visibility issue?
>>
>> On Wed, Jan 23, 2013 at 8:59 AM, Chris Sigman <[EMAIL PROTECTED]> wrote:
>>> when pig executes the job, it shows 0 rows loaded from the table

The shell will use all of a user's auths if none are provided, as a
convenience.

-Eric
On Thu, Jan 24, 2013 at 12:40 PM, Chris Sigman <[EMAIL PROTECTED]> wrote:

> A little more info in case it helps, I'm connecting using the same
> user (root) as the data was inserted with.  I can't imagine this, but
> there aren't potentially any issues because there aren't any auths on
> anything? I can scan the data just fine, so I'm rather befuddled by
> it.
>
> --
> Chris
>
>
> On Wed, Jan 23, 2013 at 11:55 AM, Chris Sigman <[EMAIL PROTECTED]> wrote:
> > No, I don't think so. The first couple dozen rows at least don't have
> > any auths set.
> >
> > --
> > Chris
> >
> >
> > On Wed, Jan 23, 2013 at 10:51 AM, David Medinets
> > <[EMAIL PROTECTED]> wrote:
> >> could it be a visibility issue?
> >>
> >> On Wed, Jan 23, 2013 at 8:59 AM, Chris Sigman <[EMAIL PROTECTED]>
> wrote:
> >>> when pig executes the job, it shows 0 rows loaded from the table
>

Right, but since the rows don't have any auth's applied, and I'm not
passing any to the call to AccumuloStorage in Pig, Pig shouldn't have
an issue finding anything, should it? Or should I have auths on the
data and pass an auth into the LOAD statement?

--
Chris
On Thu, Jan 24, 2013 at 12:50 PM, Eric Newton <[EMAIL PROTECTED]> wrote:
> The shell will use all of a user's auths if none are provided, as a
> convenience.
>
> -Eric
>
>
> On Thu, Jan 24, 2013 at 12:40 PM, Chris Sigman <[EMAIL PROTECTED]> wrote:
>>
>> A little more info in case it helps, I'm connecting using the same
>> user (root) as the data was inserted with.  I can't imagine this, but
>> there aren't potentially any issues because there aren't any auths on
>> anything? I can scan the data just fine, so I'm rather befuddled by
>> it.
>>
>> --
>> Chris
>>
>>
>> On Wed, Jan 23, 2013 at 11:55 AM, Chris Sigman <[EMAIL PROTECTED]> wrote:
>> > No, I don't think so. The first couple dozen rows at least don't have
>> > any auths set.
>> >
>> > --
>> > Chris
>> >
>> >
>> > On Wed, Jan 23, 2013 at 10:51 AM, David Medinets
>> > <[EMAIL PROTECTED]> wrote:
>> >> could it be a visibility issue?
>> >>
>> >> On Wed, Jan 23, 2013 at 8:59 AM, Chris Sigman <[EMAIL PROTECTED]>
>> >> wrote:
>> >>> when pig executes the job, it shows 0 rows loaded from the table
>
>

I tried adding a row with a specific auth set and added an auths
parameter to the LOAD statement, but still didn't get anything.  Any
suggestions as to how I could figure out where the disconnect is?

--
Chris
On Thu, Jan 24, 2013 at 1:16 PM, Chris Sigman <[EMAIL PROTECTED]> wrote:
> Right, but since the rows don't have any auth's applied, and I'm not
> passing any to the call to AccumuloStorage in Pig, Pig shouldn't have
> an issue finding anything, should it? Or should I have auths on the
> data and pass an auth into the LOAD statement?
>
> --
> Chris
>
>
> On Thu, Jan 24, 2013 at 12:50 PM, Eric Newton <[EMAIL PROTECTED]> wrote:
>> The shell will use all of a user's auths if none are provided, as a
>> convenience.
>>
>> -Eric
>>
>>
>> On Thu, Jan 24, 2013 at 12:40 PM, Chris Sigman <[EMAIL PROTECTED]> wrote:
>>>
>>> A little more info in case it helps, I'm connecting using the same
>>> user (root) as the data was inserted with.  I can't imagine this, but
>>> there aren't potentially any issues because there aren't any auths on
>>> anything? I can scan the data just fine, so I'm rather befuddled by
>>> it.
>>>
>>> --
>>> Chris
>>>
>>>
>>> On Wed, Jan 23, 2013 at 11:55 AM, Chris Sigman <[EMAIL PROTECTED]> wrote:
>>> > No, I don't think so. The first couple dozen rows at least don't have
>>> > any auths set.
>>> >
>>> > --
>>> > Chris
>>> >
>>> >
>>> > On Wed, Jan 23, 2013 at 10:51 AM, David Medinets
>>> > <[EMAIL PROTECTED]> wrote:
>>> >> could it be a visibility issue?
>>> >>
>>> >> On Wed, Jan 23, 2013 at 8:59 AM, Chris Sigman <[EMAIL PROTECTED]>
>>> >> wrote:
>>> >>> when pig executes the job, it shows 0 rows loaded from the table
>>
>>