You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@hbase.apache.org by "Gregory Chanan (JIRA)" <ji...@apache.org> on 2012/10/19 23:36:11 UTC
[jira] [Created] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Gregory Chanan created HBASE-7018:
-------------------------------------
Summary: Fix and Improve TableDescriptor caching for bulk assignment
Key: HBASE-7018
URL: https://issues.apache.org/jira/browse/HBASE-7018
Project: HBase
Issue Type: Bug
Components: regionserver
Reporter: Gregory Chanan
Assignee: Gregory Chanan
Fix For: 0.94.3, 0.96.0
HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
{code}
HTableDescriptor htd = null;
if (htds == null) {
htd = this.tableDescriptors.get(region.getTableName());
} else {
htd = htds.get(region.getTableNameAsString());
if (htd == null) {
htd = this.tableDescriptors.get(region.getTableName());
htds.put(region.getRegionNameAsString(), htd);
}
}
{code}
i.e. we get the tableName from the map but write the regionName.
Even fixing this, it looks like there are areas for improvement:
1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
#regions + #tables
(with caching):
min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
We can make this only one RPC, yielding:
#tables
Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gregory Chanan updated HBASE-7018:
----------------------------------
Status: Patch Available (was: Open)
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: HBASE-7018-94.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Hadoop QA (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13482003#comment-13482003 ]
Hadoop QA commented on HBASE-7018:
----------------------------------
{color:red}-1 overall{color}. Here are the results of testing the latest attachment
http://issues.apache.org/jira/secure/attachment/12550373/HBASE-7018-v4-trunk.patch
against trunk revision .
{color:green}+1 @author{color}. The patch does not contain any @author tags.
{color:green}+1 tests included{color}. The patch appears to include 3 new or modified tests.
{color:green}+1 hadoop2.0{color}. The patch compiles against the hadoop 2.0 profile.
{color:red}-1 javadoc{color}. The javadoc tool appears to have generated 82 warning messages.
{color:green}+1 javac{color}. The applied patch does not increase the total number of javac compiler warnings.
{color:red}-1 findbugs{color}. The patch appears to introduce 4 new Findbugs (version 1.3.9) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase the total number of release audit warnings.
{color:green}+1 core tests{color}. The patch passed unit tests in .
Test results: https://builds.apache.org/job/PreCommit-HBASE-Build/3116//testReport/
Findbugs warnings: https://builds.apache.org/job/PreCommit-HBASE-Build/3116//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-hadoop2-compat.html
Findbugs warnings: https://builds.apache.org/job/PreCommit-HBASE-Build/3116//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-hadoop1-compat.html
Findbugs warnings: https://builds.apache.org/job/PreCommit-HBASE-Build/3116//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-common.html
Findbugs warnings: https://builds.apache.org/job/PreCommit-HBASE-Build/3116//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-server.html
Findbugs warnings: https://builds.apache.org/job/PreCommit-HBASE-Build/3116//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-hadoop-compat.html
Console output: https://builds.apache.org/job/PreCommit-HBASE-Build/3116//console
This message is automatically generated.
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: 7018-trunk.v2, HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-94-v3.patch, HBASE-7018-trunk.patch, HBASE-7018-v3-trunk.patch, HBASE-7018-v4-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gregory Chanan updated HBASE-7018:
----------------------------------
Attachment: HBASE-7018-v3-trunk.patch
v3 of trunk patch. Working on v3 of 94 patch.
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: 7018-trunk.v2, HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-trunk.patch, HBASE-7018-v3-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Lars Hofhansl (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13482151#comment-13482151 ]
Lars Hofhansl commented on HBASE-7018:
--------------------------------------
Should we bother with checking the mod time if refresh is false. We could just load the file again, since getting the mod stamp is a NN roundtrip anyway (just avoid the getting the data from a DN).
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: 7018-trunk.v2, HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-94-v3.patch, HBASE-7018-trunk.patch, HBASE-7018-v3-trunk.patch, HBASE-7018-v4-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Ted Yu (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13480600#comment-13480600 ]
Ted Yu commented on HBASE-7018:
-------------------------------
The test failure in TestHRegionInfo#testGetSetOfHTD is reproducible.
It is on this line in testGetSetOfHTD():
{code}
assertTrue(htd.equals(htd2));
{code}
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Ted Yu (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13481944#comment-13481944 ]
Ted Yu commented on HBASE-7018:
-------------------------------
Thanks for the quick turnaround.
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: 7018-trunk.v2, HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-trunk.patch, HBASE-7018-v3-trunk.patch, HBASE-7018-v4-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gregory Chanan updated HBASE-7018:
----------------------------------
Attachment: HBASE-7018-94-v2.patch
v2:
- Removed mention of TableExistsException; this seems to be an outdated comment
- Properly pass "false" as refresh parameter for existing calls to FSTableDescriptor.get
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: HBASE-7018-94.patch, HBASE-7018-94-v2.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13480552#comment-13480552 ]
Gregory Chanan commented on HBASE-7018:
---------------------------------------
Testing a patch for trunk now.
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: HBASE-7018-94.patch, HBASE-7018-94-v2.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Lars Hofhansl (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13482537#comment-13482537 ]
Lars Hofhansl commented on HBASE-7018:
--------------------------------------
+1
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: 7018-trunk.v2, HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-94-v3.patch, HBASE-7018-trunk.patch, HBASE-7018-v3-trunk.patch, HBASE-7018-v4-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Ted Yu (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13481921#comment-13481921 ]
Ted Yu commented on HBASE-7018:
-------------------------------
Patch v3 looks good.
{code}
+ // If refresh, check mod time has not changed (this is trip to NN).
+ if (getTableInfoModtime(this.fs, this.rootdir, tablename) <= cachedtdm.getModtime()) {
{code}
Was the above comment leftover from previous patch ?
{code}
+ return tdmt == null? null: tdmt.getTableDescriptor();
{code}
nit: insert space following the words 'null'.
I ran TestFromClientSide3#testAdvancedConfigOverride, TestConstraint and TestMasterObserver
They passed with patch v3.
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: 7018-trunk.v2, HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-trunk.patch, HBASE-7018-v3-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13482516#comment-13482516 ]
Gregory Chanan commented on HBASE-7018:
---------------------------------------
no problem rajeshbabu.
Going to commit to trunk and 94 later today if I hear no objections.
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: 7018-trunk.v2, HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-94-v3.patch, HBASE-7018-trunk.patch, HBASE-7018-v3-trunk.patch, HBASE-7018-v4-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Hadoop QA (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13481001#comment-13481001 ]
Hadoop QA commented on HBASE-7018:
----------------------------------
{color:red}-1 overall{color}. Here are the results of testing the latest attachment
http://issues.apache.org/jira/secure/attachment/12550189/7018-trunk.v2
against trunk revision .
{color:green}+1 @author{color}. The patch does not contain any @author tags.
{color:green}+1 tests included{color}. The patch appears to include 3 new or modified tests.
{color:green}+1 hadoop2.0{color}. The patch compiles against the hadoop 2.0 profile.
{color:red}-1 javadoc{color}. The javadoc tool appears to have generated 82 warning messages.
{color:green}+1 javac{color}. The applied patch does not increase the total number of javac compiler warnings.
{color:red}-1 findbugs{color}. The patch appears to introduce 5 new Findbugs (version 1.3.9) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase the total number of release audit warnings.
{color:red}-1 core tests{color}. The patch failed these unit tests:
org.apache.hadoop.hbase.client.TestFromClientSide3
org.apache.hadoop.hbase.coprocessor.TestClassLoading
org.apache.hadoop.hbase.client.TestAdmin
org.apache.hadoop.hbase.coprocessor.TestMasterObserver
org.apache.hadoop.hbase.constraint.TestConstraint
org.apache.hadoop.hbase.coprocessor.TestRegionServerCoprocessorExceptionWithAbort
org.apache.hadoop.hbase.master.TestZKBasedOpenCloseRegion
Test results: https://builds.apache.org/job/PreCommit-HBASE-Build/3108//testReport/
Findbugs warnings: https://builds.apache.org/job/PreCommit-HBASE-Build/3108//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-hadoop2-compat.html
Findbugs warnings: https://builds.apache.org/job/PreCommit-HBASE-Build/3108//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-server.html
Findbugs warnings: https://builds.apache.org/job/PreCommit-HBASE-Build/3108//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-hadoop1-compat.html
Findbugs warnings: https://builds.apache.org/job/PreCommit-HBASE-Build/3108//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-common.html
Findbugs warnings: https://builds.apache.org/job/PreCommit-HBASE-Build/3108//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-hadoop-compat.html
Console output: https://builds.apache.org/job/PreCommit-HBASE-Build/3108//console
This message is automatically generated.
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: 7018-trunk.v2, HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gregory Chanan updated HBASE-7018:
----------------------------------
Resolution: Fixed
Status: Resolved (was: Patch Available)
Thanks for the reviews everyone.
Committed to 0.94 and trunk.
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: 7018-trunk.v2, HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-94-v3.patch, HBASE-7018-trunk.patch, HBASE-7018-v3-trunk.patch, HBASE-7018-v4-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Hadoop QA (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13480518#comment-13480518 ]
Hadoop QA commented on HBASE-7018:
----------------------------------
{color:red}-1 overall{color}. Here are the results of testing the latest attachment
http://issues.apache.org/jira/secure/attachment/12550107/HBASE-7018-94-v2.patch
against trunk revision .
{color:green}+1 @author{color}. The patch does not contain any @author tags.
{color:green}+1 tests included{color}. The patch appears to include 3 new or modified tests.
{color:red}-1 patch{color}. The patch command could not apply the patch.
Console output: https://builds.apache.org/job/PreCommit-HBASE-Build/3094//console
This message is automatically generated.
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: HBASE-7018-94.patch, HBASE-7018-94-v2.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13482159#comment-13482159 ]
Gregory Chanan commented on HBASE-7018:
---------------------------------------
Latest version doesn't have refresh.
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: 7018-trunk.v2, HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-94-v3.patch, HBASE-7018-trunk.patch, HBASE-7018-v3-trunk.patch, HBASE-7018-v4-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Hudson (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13482919#comment-13482919 ]
Hudson commented on HBASE-7018:
-------------------------------
Integrated in HBase-TRUNK #3478 (See [https://builds.apache.org/job/HBase-TRUNK/3478/])
HBASE-7018 Fix and Improve TableDescriptor caching for bulk assignment (Revision 1401525)
Result = FAILURE
gchanan :
Files :
* /hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/TableDescriptors.java
* /hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/util/FSTableDescriptors.java
* /hbase/trunk/hbase-server/src/test/java/org/apache/hadoop/hbase/master/TestCatalogJanitor.java
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: 7018-trunk.v2, HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-94-v3.patch, HBASE-7018-trunk.patch, HBASE-7018-v3-trunk.patch, HBASE-7018-v4-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13480581#comment-13480581 ]
Gregory Chanan commented on HBASE-7018:
---------------------------------------
Benchmarks for trunk:
First test: Bulk assign 500 empty regions in one table on my local box. Run each trial 3 times.
Without patch: 1.299 s
With patch: 1.194 s
Second test: Bulk assign 400 regions, all belonging to separate tables (so 400 tables). Same setup otherwise. This is really to test the RPC improvements.
Without patch: 3.338 s
With patch: 2.648 s
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Ted Yu (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13480528#comment-13480528 ]
Ted Yu commented on HBASE-7018:
-------------------------------
v2 looks good to me.
Is patch for trunk necessary ?
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: HBASE-7018-94.patch, HBASE-7018-94-v2.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13481889#comment-13481889 ]
Gregory Chanan commented on HBASE-7018:
---------------------------------------
Test failures seem real. I think that previously, we would load the latest version the HTableDescriptor once on a call openRegion, but now, if the HTableDescriptor is already in the cache, we won't load the latest version. This seems like a real problem.
We could just store which tables we've re-loaded, and use the current "refresh" code, but I think it's probably clearer to just blow that away and only keep the optimization about not making multiple roundtrips to the NN.
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: 7018-trunk.v2, HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Ted Yu (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13480495#comment-13480495 ]
Ted Yu commented on HBASE-7018:
-------------------------------
{code}
+ * @throws TableExistsException
{code}
I don't see the above exception declared to be thrown by the new get() methods.
Where is true passed for the refresh parameter in the new get() methods ?
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: HBASE-7018-94.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gregory Chanan updated HBASE-7018:
----------------------------------
Attachment: HBASE-7018-94.patch
Attached a patch against 0.94
This is version d as described in the benchmark.
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: HBASE-7018-94.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Ted Yu (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13480432#comment-13480432 ]
Ted Yu commented on HBASE-7018:
-------------------------------
Nice finding and nice improvement.
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Lars Hofhansl (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13481033#comment-13481033 ]
Lars Hofhansl commented on HBASE-7018:
--------------------------------------
@Ted: Good find.
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: 7018-trunk.v2, HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "rajeshbabu (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13482167#comment-13482167 ]
rajeshbabu commented on HBASE-7018:
-----------------------------------
bq. we get the tableName from the map but write the regionName.
Sorry for this. I have prepared patch internally and tested thoroughly but made mistake when it was rebased for community version.
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: 7018-trunk.v2, HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-94-v3.patch, HBASE-7018-trunk.patch, HBASE-7018-v3-trunk.patch, HBASE-7018-v4-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "stack (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13480861#comment-13480861 ]
stack commented on HBASE-7018:
------------------------------
Nice numbers G.
Minor, FileNotFoundException is an IOException so the FNFE redundant?
+1 on patch (after figuring the test fail above). Good stuff.
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Hudson (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13490365#comment-13490365 ]
Hudson commented on HBASE-7018:
-------------------------------
Integrated in HBase-0.94-security-on-Hadoop-23 #9 (See [https://builds.apache.org/job/HBase-0.94-security-on-Hadoop-23/9/])
HBASE-7018 Fix and Improve TableDescriptor caching for bulk assignment (Revision 1401526)
Result = FAILURE
gchanan :
Files :
* /hbase/branches/0.94/src/main/java/org/apache/hadoop/hbase/TableDescriptors.java
* /hbase/branches/0.94/src/main/java/org/apache/hadoop/hbase/regionserver/HRegionServer.java
* /hbase/branches/0.94/src/main/java/org/apache/hadoop/hbase/util/FSTableDescriptors.java
* /hbase/branches/0.94/src/test/java/org/apache/hadoop/hbase/master/TestCatalogJanitor.java
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: 7018-trunk.v2, HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-94-v3.patch, HBASE-7018-trunk.patch, HBASE-7018-v3-trunk.patch, HBASE-7018-v4-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Hadoop QA (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13480474#comment-13480474 ]
Hadoop QA commented on HBASE-7018:
----------------------------------
{color:red}-1 overall{color}. Here are the results of testing the latest attachment
http://issues.apache.org/jira/secure/attachment/12550099/HBASE-7018-94.patch
against trunk revision .
{color:green}+1 @author{color}. The patch does not contain any @author tags.
{color:green}+1 tests included{color}. The patch appears to include 3 new or modified tests.
{color:red}-1 patch{color}. The patch command could not apply the patch.
Console output: https://builds.apache.org/job/PreCommit-HBASE-Build/3093//console
This message is automatically generated.
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: HBASE-7018-94.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Hadoop QA (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13482042#comment-13482042 ]
Hadoop QA commented on HBASE-7018:
----------------------------------
{color:red}-1 overall{color}. Here are the results of testing the latest attachment
http://issues.apache.org/jira/secure/attachment/12550386/HBASE-7018-94-v3.patch
against trunk revision .
{color:green}+1 @author{color}. The patch does not contain any @author tags.
{color:green}+1 tests included{color}. The patch appears to include 3 new or modified tests.
{color:red}-1 patch{color}. The patch command could not apply the patch.
Console output: https://builds.apache.org/job/PreCommit-HBASE-Build/3119//console
This message is automatically generated.
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: 7018-trunk.v2, HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-94-v3.patch, HBASE-7018-trunk.patch, HBASE-7018-v3-trunk.patch, HBASE-7018-v4-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Lars Hofhansl (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13481035#comment-13481035 ]
Lars Hofhansl commented on HBASE-7018:
--------------------------------------
Patch looks good. I wonder whether we should even bother with checking the mod time if refresh is false. We could just load the file again, since getting the mod stamp is a NN roundtrip anyway.
+1 otherwise
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: 7018-trunk.v2, HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gregory Chanan updated HBASE-7018:
----------------------------------
Attachment: HBASE-7018-trunk.patch
Patch for trunk
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gregory Chanan updated HBASE-7018:
----------------------------------
Attachment: HBASE-7018-v4-trunk.patch
Good catches, put up v4 addressing your comments.
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: 7018-trunk.v2, HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-trunk.patch, HBASE-7018-v3-trunk.patch, HBASE-7018-v4-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Ted Yu (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Ted Yu updated HBASE-7018:
--------------------------
Attachment: 7018-trunk.v2
Turns out that there was a bug in FSTableDescriptors.get(final String tablename, boolean refresh), line 198:
{code}
this.cache.put(tablename, tdmt);
tdm = tdmt;
{code}
When we have new tdmt which is not null, tdm should have been updated.
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: 7018-trunk.v2, HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Hudson (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13483172#comment-13483172 ]
Hudson commented on HBASE-7018:
-------------------------------
Integrated in HBase-TRUNK-on-Hadoop-2.0.0 #233 (See [https://builds.apache.org/job/HBase-TRUNK-on-Hadoop-2.0.0/233/])
HBASE-7018 Fix and Improve TableDescriptor caching for bulk assignment (Revision 1401525)
Result = FAILURE
gchanan :
Files :
* /hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/TableDescriptors.java
* /hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/util/FSTableDescriptors.java
* /hbase/trunk/hbase-server/src/test/java/org/apache/hadoop/hbase/master/TestCatalogJanitor.java
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: 7018-trunk.v2, HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-94-v3.patch, HBASE-7018-trunk.patch, HBASE-7018-v3-trunk.patch, HBASE-7018-v4-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gregory Chanan updated HBASE-7018:
----------------------------------
Attachment: HBASE-7018-94-v3.patch
Latest version against 94. I ran this against my local jenkins and all tests passed.
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: 7018-trunk.v2, HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-94-v3.patch, HBASE-7018-trunk.patch, HBASE-7018-v3-trunk.patch, HBASE-7018-v4-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13480414#comment-13480414 ]
Gregory Chanan commented on HBASE-7018:
---------------------------------------
First test: Bulk assign 500 empty regions in one table on my local box. Run each trial 3 times.
Versions: a) 0.94 today (broken cache)
b) 0.94 w/ fixed "extra" cache (the cache introduced in HBASE-6214)
c) 0.94 w/ fixed "extra" cache + faster existing cache (don't go to NN each time) + reduce RPC overhead
d) Same as c) but remove "extra cache"
Results:
a) 14 seconds (only ran test once)
b) 1.458 seconds
c) 1.379 seconds
d) 1.227 seconds
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Hadoop QA (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13480587#comment-13480587 ]
Hadoop QA commented on HBASE-7018:
----------------------------------
{color:red}-1 overall{color}. Here are the results of testing the latest attachment
http://issues.apache.org/jira/secure/attachment/12550120/HBASE-7018-trunk.patch
against trunk revision .
{color:green}+1 @author{color}. The patch does not contain any @author tags.
{color:green}+1 tests included{color}. The patch appears to include 3 new or modified tests.
{color:green}+1 hadoop2.0{color}. The patch compiles against the hadoop 2.0 profile.
{color:red}-1 javadoc{color}. The javadoc tool appears to have generated 82 warning messages.
{color:green}+1 javac{color}. The applied patch does not increase the total number of javac compiler warnings.
{color:red}-1 findbugs{color}. The patch appears to introduce 4 new Findbugs (version 1.3.9) warnings.
{color:green}+1 release audit{color}. The applied patch does not increase the total number of release audit warnings.
{color:red}-1 core tests{color}. The patch failed these unit tests:
org.apache.hadoop.hbase.regionserver.TestHRegionInfo
Test results: https://builds.apache.org/job/PreCommit-HBASE-Build/3098//testReport/
Findbugs warnings: https://builds.apache.org/job/PreCommit-HBASE-Build/3098//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-hadoop2-compat.html
Findbugs warnings: https://builds.apache.org/job/PreCommit-HBASE-Build/3098//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-hadoop1-compat.html
Findbugs warnings: https://builds.apache.org/job/PreCommit-HBASE-Build/3098//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-common.html
Findbugs warnings: https://builds.apache.org/job/PreCommit-HBASE-Build/3098//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-server.html
Findbugs warnings: https://builds.apache.org/job/PreCommit-HBASE-Build/3098//artifact/trunk/patchprocess/newPatchFindbugsWarningshbase-hadoop-compat.html
Console output: https://builds.apache.org/job/PreCommit-HBASE-Build/3098//console
This message is automatically generated.
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Gregory Chanan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13480418#comment-13480418 ]
Gregory Chanan commented on HBASE-7018:
---------------------------------------
Second test: Bulk assign 400 regions, all belonging to separate tables (so 400 tables). Same setup otherwise. This is really to test the RPC improvements.
Results
a) N/A (Didn't test)
b) 3.725 seconds
c) 2.848 seconds
d) 2.860 seconds
Talking off almost an entire second here isn't too shabby.
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-7018) Fix and Improve TableDescriptor
caching for bulk assignment
Posted by "Hudson (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-7018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13482910#comment-13482910 ]
Hudson commented on HBASE-7018:
-------------------------------
Integrated in HBase-0.94 #550 (See [https://builds.apache.org/job/HBase-0.94/550/])
HBASE-7018 Fix and Improve TableDescriptor caching for bulk assignment (Revision 1401526)
Result = FAILURE
gchanan :
Files :
* /hbase/branches/0.94/src/main/java/org/apache/hadoop/hbase/TableDescriptors.java
* /hbase/branches/0.94/src/main/java/org/apache/hadoop/hbase/regionserver/HRegionServer.java
* /hbase/branches/0.94/src/main/java/org/apache/hadoop/hbase/util/FSTableDescriptors.java
* /hbase/branches/0.94/src/test/java/org/apache/hadoop/hbase/master/TestCatalogJanitor.java
> Fix and Improve TableDescriptor caching for bulk assignment
> -----------------------------------------------------------
>
> Key: HBASE-7018
> URL: https://issues.apache.org/jira/browse/HBASE-7018
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Reporter: Gregory Chanan
> Assignee: Gregory Chanan
> Fix For: 0.94.3, 0.96.0
>
> Attachments: 7018-trunk.v2, HBASE-7018-94.patch, HBASE-7018-94-v2.patch, HBASE-7018-94-v3.patch, HBASE-7018-trunk.patch, HBASE-7018-v3-trunk.patch, HBASE-7018-v4-trunk.patch
>
>
> HBASE-6214 backported HBASE-5998 (Bulk assignment: regionserver optimization by using a temporary cache for table descriptors when receiving an open regions request), but it's buggy on 0.94 (0.96 appears correct):
> {code}
> HTableDescriptor htd = null;
> if (htds == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> } else {
> htd = htds.get(region.getTableNameAsString());
> if (htd == null) {
> htd = this.tableDescriptors.get(region.getTableName());
> htds.put(region.getRegionNameAsString(), htd);
> }
> }
> {code}
> i.e. we get the tableName from the map but write the regionName.
> Even fixing this, it looks like there are areas for improvement:
> 1) FSTableDescriptors already has a cache (though it goes to the NameNode each time through to check we have the latest copy. May as well combine these two caches, might be a performance win as well since we don't need to write to multiple caches.
> 2) FSTableDescriptors makes two RPCs to the NameNode when it encounters a new table. So the total number of RPCs necessary for a bulk assign (without caching is):
> #regions + #tables
> (with caching):
> min(#regions,#tables) + #tables = #tables + #tables = 2 * #tables
> We can make this only one RPC, yielding:
> #tables
> Probably not a big deal for most users, but in a multi-tenant situation where the number of regions being bulk assigned approaches the number of tables being bulk assigned, this could be a nice performance win.
> Benchmarks coming.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira