hbase-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Rohit Kelkar <rohitkel...@gmail.com>
Subject Re: help with filter combination
Date Sat, 06 Jul 2013 00:31:15 GMT
Thanks Ted, The deptId and empid are analogies to the case at hand. The
SingleColumnValueFilter expects (colfam, columnQualifier, CompareOp,
value). In my case colfam = pt, columnQualifier = empid. But the empid is
not a fixed string. Instead it varies based on my input data. So 100
employees would mean 100 qual:value pairs.
Now I could change the column structure such that qualifier = deptId and
value = list of empids. But then the process that fills up this table will
have to first fetch the list of existing empids, add the new empid and push
it back. To avoid this I turned it around so that the process just pushes
the empid:deptid in to the table. I hope the explanation makes sense. In
short the process that fills up this table is mission critical and I am
trying to avoid adding overheads to it.

- R


On Fri, Jul 5, 2013 at 6:49 PM, Ted Yu <yuzhihong@gmail.com> wrote:

> Consider SingleColumnValueFilter
>
> You can refer
> to
> src/test/java/org/apache/hadoop/hbase/filter/TestSingleColumnValueFilter.java
>
> BTW why is deptId stored in the pt:employeeId column ?
>
> Cheers
>
> On Fri, Jul 5, 2013 at 4:43 PM, Rohit Kelkar <rohitkelkar@gmail.com>
> wrote:
>
> > Each row in my hbase table contains the following data:
> >
> > rowkey   column=pt:np, value="abcd"
> >              column=pt:vb, value="efgh"
> >              column=pt:employeeId, value=deptId
> >
> > Using a combination of filters is it possible to get all rows and all
> > qualifiers within the pt column family where deptId = 100?
> >
> >
> > - R
> >
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message