[
https://issues.apache.org/jira/browse/HADOOP-1325?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12495302
]
udanax commented on HADOOP-1325:
--------------------------------
I've tested it in my test mapreduce class for storing 10000 by 10000 dense
matrix data of row i of Table and column j of ColumnFamily.
It works good. :) I will do make linear algebra library for Hadoop & Hbase.
{code}
public void reduce(WritableComparable key, Iterator values,
OutputCollector output, Reporter reporter) throws
IOException {
Configuration conf = new Configuration();
HClient client = new HClient(conf);
client.openTable(new Text("matrices"));
long lockId = client.startUpdate(new Text(key.toString()));
while(values.hasNext()){
HbaseTestDatum m = (HbaseTestDatum)values.next();
client.put(lockId, new Text("dense:"+m.getColumnkey()),
m.getEntry().getBytes());
}
client.commit(lockId);
}
{code}
> HBase code update 0.0.0-2
> -------------------------
>
> Key: HADOOP-1325
> URL: https://issues.apache.org/jira/browse/HADOOP-1325
> Project: Hadoop
> Issue Type: Improvement
> Components: contrib/hbase
> Reporter: Jim Kellerman
> Assigned To: Jim Kellerman
> Fix For: 0.13.0
>
> Attachments: patch.txt
>
>
> The latest HBase code which contains fixes and updates for:
> - Eliminate multiple conversions to/from byte[] <-> BytesWritable
> - Scanners have a problem find all the elements in a column family if they
> were saved at different times
> - Master continuously allocates META region
> - Scanners do not properly handle deleted values
> - HMemcache and HStore scanners return oldest results first. They should
> return the most recent values first
> - New test cases for:
> o get
> o HBaseMiniCluster
> o HMemcache
> o Scanners
> o object toString methods
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
