You are viewing a plain text version of this content. The canonical link for it is here.
Posted to notifications@accumulo.apache.org by "Keith Turner (JIRA)" <ji...@apache.org> on 2012/11/27 19:39:58 UTC
[jira] [Created] (ACCUMULO-878) Decide fate of Mock Accumulo
Keith Turner created ACCUMULO-878:
-------------------------------------
Summary: Decide fate of Mock Accumulo
Key: ACCUMULO-878
URL: https://issues.apache.org/jira/browse/ACCUMULO-878
Project: Accumulo
Issue Type: Sub-task
Reporter: Keith Turner
Fix For: 1.5.0
Now that we have MiniAccumuloCluster, we need to decide what we want to do with Mock Accumulo. Seems like there are two options for 1.5, either deperecate it or continue to maintain it. The main advantage of keeping it around is that its fast, however its behavior is not always correct (ACCUMULO-664, ACCUMULO-843, ACCUMULO-877). If we choose to deperecate it, we could probably close out the tickets against mock for 1.5.
What are peoples thoughts?
--
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] (ACCUMULO-878) Decide fate of Mock Accumulo
Posted by "Keith Turner (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/ACCUMULO-878?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13504969#comment-13504969 ]
Keith Turner commented on ACCUMULO-878:
---------------------------------------
I am going to try to figure out if MiniAccumuloCluster works on Windows and what can be done to make it faster.
> Decide fate of Mock Accumulo
> ----------------------------
>
> Key: ACCUMULO-878
> URL: https://issues.apache.org/jira/browse/ACCUMULO-878
> Project: Accumulo
> Issue Type: Sub-task
> Components: client
> Reporter: Keith Turner
> Fix For: 1.5.0
>
>
> Now that we have MiniAccumuloCluster, we need to decide what we want to do with Mock Accumulo. Seems like there are two options for 1.5, either deperecate it or continue to maintain it. The main advantage of keeping it around is that its fast, however its behavior is not always correct (ACCUMULO-664, ACCUMULO-843, ACCUMULO-877). If we choose to deperecate it, we could probably close out the tickets against mock for 1.5.
> What are peoples thoughts?
--
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] (ACCUMULO-878) Decide fate of Mock Accumulo
Posted by "Billie Rinaldi (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/ACCUMULO-878?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13504835#comment-13504835 ]
Billie Rinaldi commented on ACCUMULO-878:
-----------------------------------------
I'd be in favor of deprecating mock accumulo.
> Decide fate of Mock Accumulo
> ----------------------------
>
> Key: ACCUMULO-878
> URL: https://issues.apache.org/jira/browse/ACCUMULO-878
> Project: Accumulo
> Issue Type: Sub-task
> Components: client
> Reporter: Keith Turner
> Fix For: 1.5.0
>
>
> Now that we have MiniAccumuloCluster, we need to decide what we want to do with Mock Accumulo. Seems like there are two options for 1.5, either deperecate it or continue to maintain it. The main advantage of keeping it around is that its fast, however its behavior is not always correct (ACCUMULO-664, ACCUMULO-843, ACCUMULO-877). If we choose to deperecate it, we could probably close out the tickets against mock for 1.5.
> What are peoples thoughts?
--
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] (ACCUMULO-878) Decide fate of Mock Accumulo
Posted by "Christopher Tubbs (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/ACCUMULO-878?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13504949#comment-13504949 ]
Christopher Tubbs commented on ACCUMULO-878:
--------------------------------------------
I like the idea of both. However, I do think that there are probably some "unit tests" that are really integration tests and should be tested against a real instance. Many such tests can be done during the integration-test phase of the build lifecycle (perhaps within the test module or a dedicated integration test module, if it will depend on other modules being built). MiniAccumuloCluster can be set up during the pre-integration-test phase and shut down during the post-integration-test phase, with maven-failsafe-plugin being used to run these tests previously treated as "unit tests". This can probably be done using maven-exec-plugin, but it might be easier to write a custom maven-mini-accumulo-cluster-plugin, like the maven-jetty-plugin, for launching/shutting down the MiniAccumuloCluster. If this option is pursued, I'm curious if there will still be any remaining need for mock... it might be easier to determine if we need this after we implement proper integration testing. If there's a problem running integration tests on Windows, that can probably be fixed.
> Decide fate of Mock Accumulo
> ----------------------------
>
> Key: ACCUMULO-878
> URL: https://issues.apache.org/jira/browse/ACCUMULO-878
> Project: Accumulo
> Issue Type: Sub-task
> Components: client
> Reporter: Keith Turner
> Fix For: 1.5.0
>
>
> Now that we have MiniAccumuloCluster, we need to decide what we want to do with Mock Accumulo. Seems like there are two options for 1.5, either deperecate it or continue to maintain it. The main advantage of keeping it around is that its fast, however its behavior is not always correct (ACCUMULO-664, ACCUMULO-843, ACCUMULO-877). If we choose to deperecate it, we could probably close out the tickets against mock for 1.5.
> What are peoples thoughts?
--
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] (ACCUMULO-878) Decide fate of Mock Accumulo
Posted by "Keith Turner (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/ACCUMULO-878?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13504860#comment-13504860 ]
Keith Turner commented on ACCUMULO-878:
---------------------------------------
bq. used extensively for testing complicated iterator stacks
One benefit of doing this type of testing with MiniAccumuloCluster is that you may test continuance case where your iterator is reseeked.
> Decide fate of Mock Accumulo
> ----------------------------
>
> Key: ACCUMULO-878
> URL: https://issues.apache.org/jira/browse/ACCUMULO-878
> Project: Accumulo
> Issue Type: Sub-task
> Components: client
> Reporter: Keith Turner
> Fix For: 1.5.0
>
>
> Now that we have MiniAccumuloCluster, we need to decide what we want to do with Mock Accumulo. Seems like there are two options for 1.5, either deperecate it or continue to maintain it. The main advantage of keeping it around is that its fast, however its behavior is not always correct (ACCUMULO-664, ACCUMULO-843, ACCUMULO-877). If we choose to deperecate it, we could probably close out the tickets against mock for 1.5.
> What are peoples thoughts?
--
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] (ACCUMULO-878) Decide fate of Mock Accumulo
Posted by "Josh Elser (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/ACCUMULO-878?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13504951#comment-13504951 ]
Josh Elser commented on ACCUMULO-878:
-------------------------------------
There were some issues previously inside of MockAccumulo that caused things to not run in mock but run correctly on a real instance. Until we have something that can provide an equivalent in-memory instance that can be used for testing, I'm hesitant to mark MockAccumulo as deprecated. Figuring out the actual scope/functionality of MiniAccumuloCluster is dependent on this, I believe.
> Decide fate of Mock Accumulo
> ----------------------------
>
> Key: ACCUMULO-878
> URL: https://issues.apache.org/jira/browse/ACCUMULO-878
> Project: Accumulo
> Issue Type: Sub-task
> Components: client
> Reporter: Keith Turner
> Fix For: 1.5.0
>
>
> Now that we have MiniAccumuloCluster, we need to decide what we want to do with Mock Accumulo. Seems like there are two options for 1.5, either deperecate it or continue to maintain it. The main advantage of keeping it around is that its fast, however its behavior is not always correct (ACCUMULO-664, ACCUMULO-843, ACCUMULO-877). If we choose to deperecate it, we could probably close out the tickets against mock for 1.5.
> What are peoples thoughts?
--
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] (ACCUMULO-878) Decide fate of Mock Accumulo
Posted by "William Slacum (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/ACCUMULO-878?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13504901#comment-13504901 ]
William Slacum commented on ACCUMULO-878:
-----------------------------------------
I'm leaning towards keeping it. There have been several behavioral differences between MockAccumulo and a real instance that can go undetected (and possibly remain undocumented), but I prefer to have some framework that is completely in memory and doesn't dirty the user's file system.
> Decide fate of Mock Accumulo
> ----------------------------
>
> Key: ACCUMULO-878
> URL: https://issues.apache.org/jira/browse/ACCUMULO-878
> Project: Accumulo
> Issue Type: Sub-task
> Components: client
> Reporter: Keith Turner
> Fix For: 1.5.0
>
>
> Now that we have MiniAccumuloCluster, we need to decide what we want to do with Mock Accumulo. Seems like there are two options for 1.5, either deperecate it or continue to maintain it. The main advantage of keeping it around is that its fast, however its behavior is not always correct (ACCUMULO-664, ACCUMULO-843, ACCUMULO-877). If we choose to deperecate it, we could probably close out the tickets against mock for 1.5.
> What are peoples thoughts?
--
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] (ACCUMULO-878) Decide fate of Mock Accumulo
Posted by "Keith Turner (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/ACCUMULO-878?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13504855#comment-13504855 ]
Keith Turner commented on ACCUMULO-878:
---------------------------------------
bq. even for those poor souls who develop on Windows
I am not sure if MiniAccumuloCluster will work on Windows. I have never tried it. Need to test that.
> Decide fate of Mock Accumulo
> ----------------------------
>
> Key: ACCUMULO-878
> URL: https://issues.apache.org/jira/browse/ACCUMULO-878
> Project: Accumulo
> Issue Type: Sub-task
> Components: client
> Reporter: Keith Turner
> Fix For: 1.5.0
>
>
> Now that we have MiniAccumuloCluster, we need to decide what we want to do with Mock Accumulo. Seems like there are two options for 1.5, either deperecate it or continue to maintain it. The main advantage of keeping it around is that its fast, however its behavior is not always correct (ACCUMULO-664, ACCUMULO-843, ACCUMULO-877). If we choose to deperecate it, we could probably close out the tickets against mock for 1.5.
> What are peoples thoughts?
--
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] (ACCUMULO-878) Decide fate of Mock Accumulo
Posted by "Billie Rinaldi (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/ACCUMULO-878?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13504948#comment-13504948 ]
Billie Rinaldi commented on ACCUMULO-878:
-----------------------------------------
I think using the file system is fine as long as it's all in the tmp directory.
> Decide fate of Mock Accumulo
> ----------------------------
>
> Key: ACCUMULO-878
> URL: https://issues.apache.org/jira/browse/ACCUMULO-878
> Project: Accumulo
> Issue Type: Sub-task
> Components: client
> Reporter: Keith Turner
> Fix For: 1.5.0
>
>
> Now that we have MiniAccumuloCluster, we need to decide what we want to do with Mock Accumulo. Seems like there are two options for 1.5, either deperecate it or continue to maintain it. The main advantage of keeping it around is that its fast, however its behavior is not always correct (ACCUMULO-664, ACCUMULO-843, ACCUMULO-877). If we choose to deperecate it, we could probably close out the tickets against mock for 1.5.
> What are peoples thoughts?
--
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] (ACCUMULO-878) Decide fate of Mock Accumulo
Posted by "Eric Newton (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/ACCUMULO-878?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13504850#comment-13504850 ]
Eric Newton commented on ACCUMULO-878:
--------------------------------------
It is faster, and it works for in-memory unit tests, even for those poor souls who develop on Windows (eew). I know that it is used extensively for testing complicated iterator stacks. Perhaps Josh and/or Bill would like to chime in.
> Decide fate of Mock Accumulo
> ----------------------------
>
> Key: ACCUMULO-878
> URL: https://issues.apache.org/jira/browse/ACCUMULO-878
> Project: Accumulo
> Issue Type: Sub-task
> Components: client
> Reporter: Keith Turner
> Fix For: 1.5.0
>
>
> Now that we have MiniAccumuloCluster, we need to decide what we want to do with Mock Accumulo. Seems like there are two options for 1.5, either deperecate it or continue to maintain it. The main advantage of keeping it around is that its fast, however its behavior is not always correct (ACCUMULO-664, ACCUMULO-843, ACCUMULO-877). If we choose to deperecate it, we could probably close out the tickets against mock for 1.5.
> What are peoples thoughts?
--
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] (ACCUMULO-878) Decide fate of Mock Accumulo
Posted by "Keith Turner (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/ACCUMULO-878?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13504857#comment-13504857 ]
Keith Turner commented on ACCUMULO-878:
---------------------------------------
One more thing to consider if we continue to maintain mock is user experience. A new user comes to this project and sees two possibilities for they could use for testing. Its confusing. This can be solved with adequate documentation that outlines the pros and cons of each to help the user make a decision.
> Decide fate of Mock Accumulo
> ----------------------------
>
> Key: ACCUMULO-878
> URL: https://issues.apache.org/jira/browse/ACCUMULO-878
> Project: Accumulo
> Issue Type: Sub-task
> Components: client
> Reporter: Keith Turner
> Fix For: 1.5.0
>
>
> Now that we have MiniAccumuloCluster, we need to decide what we want to do with Mock Accumulo. Seems like there are two options for 1.5, either deperecate it or continue to maintain it. The main advantage of keeping it around is that its fast, however its behavior is not always correct (ACCUMULO-664, ACCUMULO-843, ACCUMULO-877). If we choose to deperecate it, we could probably close out the tickets against mock for 1.5.
> What are peoples thoughts?
--
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] (ACCUMULO-878) Decide fate of Mock Accumulo
Posted by "Keith Turner (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/ACCUMULO-878?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13504830#comment-13504830 ]
Keith Turner commented on ACCUMULO-878:
---------------------------------------
I am leaning twoards deprecating mock accumulo.
There are some unit test in accumulo-core and accumulo-server that use mock accumulo. MiniAccumuloCluster is in the accumulo-test module which depends on accumlo-core and accumulo-server. So I think the unit test in accumulo that use mock would need to move to accumulo-test.
> Decide fate of Mock Accumulo
> ----------------------------
>
> Key: ACCUMULO-878
> URL: https://issues.apache.org/jira/browse/ACCUMULO-878
> Project: Accumulo
> Issue Type: Sub-task
> Components: client
> Reporter: Keith Turner
> Fix For: 1.5.0
>
>
> Now that we have MiniAccumuloCluster, we need to decide what we want to do with Mock Accumulo. Seems like there are two options for 1.5, either deperecate it or continue to maintain it. The main advantage of keeping it around is that its fast, however its behavior is not always correct (ACCUMULO-664, ACCUMULO-843, ACCUMULO-877). If we choose to deperecate it, we could probably close out the tickets against mock for 1.5.
> What are peoples thoughts?
--
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