Apache
/
hbase
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

HBASE-1812 Document fact that Scanners do not respect row locks 

git-svn-id: https://svn.apache.org/repos/asf/hadoop/hbase/trunk@811568 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
Michael Stack 2009-09-04 23:33:45 +00:00
parent 2a45323857
commit d5e3f6c894
2 changed files with 14 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
src/java/org/apache/hadoop/hbase/client/HTable.java
View File

@ -56,7 +56,10 @@ import org.apache.hadoop.hbase.util.Writables;
/** /**
* Used to communicate with a single HBase table. * Used to communicate with a single HBase table.
* <p> * <p>
* This class is not MT safe for writes. * This class is not thread safe for writes.
* Gets, puts, and deletes take out a row lock for the duration
* of their operation. Scans (currently) do not respect
* row locking.
*/ */
public class HTable implements HTableInterface { public class HTable implements HTableInterface {
private final HConnection connection; private final HConnection connection;

16
src/java/org/apache/hadoop/hbase/client/package-info.java
View File

@ -30,14 +30,13 @@ Provides HBase Client
<p>To administer HBase, create and drop tables, list and alter tables, <p>To administer HBase, create and drop tables, list and alter tables,
use {@link org.apache.hadoop.hbase.client.HBaseAdmin}. Once created, table access is via an instance use {@link org.apache.hadoop.hbase.client.HBaseAdmin}. Once created, table access is via an instance
of {@link org.apache.hadoop.hbase.client.HTable}. You add content to a table a row at a time. To insert, of {@link org.apache.hadoop.hbase.client.HTable}. You add content to a table a row at a time. To insert,
create an instance of a {@link org.apache.hadoop.hbase.client.Put} object and after setting it appropriately, create an instance of a {@link org.apache.hadoop.hbase.client.Put} object. Specify value, target column
commit your update using {@link org.apache.hadoop.hbase.client.HTable#put(Put)}. and optionally a timestamp. Commit your update using {@link org.apache.hadoop.hbase.client.HTable#put(Put)}.
To fetch your inserted To fetch your inserted value, use {@link org.apache.hadoop.hbase.client.Get}. The Get can be specified to be broad -- get all
value, use {@link org.apache.hadoop.hbase.client.Get}. The Get can be specified to be broad -- get all on a particular row -- or narrow; i.e. return only a single cell value. After creating an instance of
on a particular row -- or narrow; return only a single cell value. Get, invoke {@link org.apache.hadoop.hbase.client.HTable#get(Get)}. Use
When finished with your Get settings, invoke {@link org.apache.hadoop.hbase.client.HTable#get(Get)}. Use
{@link org.apache.hadoop.hbase.client.Scan} to set up a scanner -- a Cursor- like access. After {@link org.apache.hadoop.hbase.client.Scan} to set up a scanner -- a Cursor- like access. After
configuring your Scan instance, call {@link org.apache.hadoop.hbase.client.HTable#getScanner(Scan)} and then creating and configuring your Scan instance, call {@link org.apache.hadoop.hbase.client.HTable#getScanner(Scan)} and then
invoke next on the returned object. Both {@link org.apache.hadoop.hbase.client.HTable#get(Get)} and invoke next on the returned object. Both {@link org.apache.hadoop.hbase.client.HTable#get(Get)} and
{@link org.apache.hadoop.hbase.client.HTable#getScanner(Scan)} return a {@link org.apache.hadoop.hbase.client.HTable#getScanner(Scan)} return a
{@link org.apache.hadoop.hbase.client.Result}. {@link org.apache.hadoop.hbase.client.Result}.
@ -47,6 +46,9 @@ in different formats.
You can remove individual cells or entire families, etc. Pass it to You can remove individual cells or entire families, etc. Pass it to
{@link org.apache.hadoop.hbase.client.HTable#delete(Delete)} to execute. {@link org.apache.hadoop.hbase.client.HTable#delete(Delete)} to execute.
</p> </p>
<p>Puts, Gets and Deletes take out a lock on the target row for the duration of their operation.
Scans (currently) operate without respect for row locks.
</p>
<p>Client code accessing a cluster finds the cluster by querying ZooKeeper. <p>Client code accessing a cluster finds the cluster by querying ZooKeeper.
This means that the ZooKeeper quorum to use must be on the client CLASSPATH. This means that the ZooKeeper quorum to use must be on the client CLASSPATH.
Usually this means make sure the client can find your <code>hbase-site.xml</code>. Usually this means make sure the client can find your <code>hbase-site.xml</code>.