Hi,
So you will have multiple table elements instead in
gora-hbase-mapping.xml. This will enable you to push data to a different
tables.
Changing this at runtime (in an adhoc fashion) would suppose that you could
check for some identifying characteristics and assign the dataflow to the
particular datastore instance. Some justification is below.
// We keep track of all tables, because even though we
// only build a mapping for one table. We do this because of the way
// the mapping file is set up.
// (First family properties are defined, whereafter columns are defined).
//
// HBaseMapping in fact does not need to support multiple tables,
// because a Store itself only supports a single table. (Every store
// instance simply creates one mapping instance for itself).
//
// TODO A nice solution would be to redefine the mapping file structure.
// For example nest columns in families. Of course this would
break compatibility.
hth
Lewis
On Wed, Aug 21, 2013 at 1:30 PM, Sznajder ForMailingList <
[email protected]> wrote:
> Hi Lewis
>
> Thanks a lot for your answer.
> I plan to use HBASE.
>
> Is it different?
> How can I define keyspace in Gora for HBASE? I need during my running
> time, push into different keyspace. Is it possible?
>
> thanks again
>
> Benjamin
>
>
> On Wed, Aug 21, 2013 at 6:25 PM, Lewis John Mcgibbney <
> [email protected]> wrote:
>
>> Hi Benjamin,
>> You didn't mention which Gora module you are using or planning on using
>> so I'll just state my most recent Gora project and hopefull
>> the analogy will be clear.
>> I was building large dictionary maps for different document frequency
>> thresholds of Solr indexes. I wanted to store the dictionary maps and some
>> other metadata into Cassandra. For simplicity lets say thresholds were 10,
>> 20 and 50. This meant that each grouping would be stored as a different
>> keyspace in Cassandra. I could then push data there by editing the keyspace
>> name in gora-cassandra-mapping.xml accordingly, ensuring that the correct
>> file was on the classpath and running my code to load this into Cassandra.
>> I must admit not much of the Gora code is used at all.
>> This of course uses Cassandra... and it may not be the best solution but
>> it did the job for me no bother at all and was very very quick and writing
>> the data down into Cassandra.
>> hth
>> Lewis
>>
>> On Wednesday, August 21, 2013, Sznajder ForMailingList <
>> [email protected]> wrote:
>> > Hi
>> >
>> > I need to keep to a DB values (String) with status 0 or status 1.
>> > I need to fetch from the DB some values with status 0 and process them
>> for then passing their status to 1 and re-pushing them...
>> >
>> >
>> > I thought about a solution:
>> > Keeping two tables "Table_status_0" and "Table_status_1"...
>> >
>> > But, how can we keep data in two different tables using GORA (the two
>> tables have the same key/values objects)
>> >
>> >
>> > Or, maybe, do you have a better idea for fixing this issue?
>> >
>> > Benjamin
>> >
>>
>> --
>> *Lewis*
>>
>>
>
--
*Lewis*
