You are viewing a plain text version of this content. The canonical link for it is here.
Posted to derby-dev@db.apache.org by "John H. Embretsen (JIRA)" <de...@db.apache.org> on 2006/09/20 21:26:22 UTC
[jira] Created: (DERBY-1872) Backport lang/wisconsin.java at
10.2.0.4-alpha to the 10.1 branch
Backport lang/wisconsin.java at 10.2.0.4-alpha to the 10.1 branch
-----------------------------------------------------------------
Key: DERBY-1872
URL: http://issues.apache.org/jira/browse/DERBY-1872
Project: Derby
Issue Type: Test
Components: Test
Affects Versions: 10.1.3.1
Environment: N/A
Reporter: John H. Embretsen
Assigned To: John H. Embretsen
Priority: Minor
Fix For: 10.1.3.2
Investigations relating to DERBY-1564 ("wisconsin.java test failed in DerbyNet or DerbyNetClient frameworks, VM for network server got OutOfMemoryError") revealed that comparing memory usage for the lang/wisconsin.[java|sql] test in 10.1.x versus 10.2.x was like comparing apples and oranges. After the 10.1 branch was created, several updates were made to the wisconsin test in trunk that were not backported to 10.1, e.g:
* giving the optimizer unlimited time to choose query plans
* compressing the tables to avoid the instabilities reported in DERBY-937
Backporting the test will make it less unfair to compare 10.1 vs. 10.2 (or trunk) test results and memory usage, and it will make it easier to determine if failures (such as DERBY-1564) are regressions or not. On the other hand, other differences between the branches that may influence the test will remain, so this will not be a wonder cure.
The failures seen in DERBY-1564 were mainly caused by bugs (DERBY-1091, DERBY-1614) in the test harness, causing the wisconsin test for the 10.2.0.4 snapshot to run with low memory settings. These issues have been fixed.
The most practical backporting approach will probably be to take the wisconsin test as it was in the 10.2.0.4-snapshot (SVN revision 423199), and port it to the current 10.1 branch. Other changes have been made to the wisconsin test in trunk/10.2 after 10.2.0.4, but to avoid various dependencies (e.g. DERBY-1609), these should not be backported.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] Closed: (DERBY-1872) Backport lang/wisconsin.java at
10.2.0.4-alpha to the 10.1 branch
Posted by "John H. Embretsen (JIRA)" <de...@db.apache.org>.
[ http://issues.apache.org/jira/browse/DERBY-1872?page=all ]
John H. Embretsen closed DERBY-1872.
------------------------------------
Thank you for committing; I have verified that the changes made it into the branch, and that the Tinderbox test of revision 450908 (including these changes) did not fail more than previous revisions. This patch got rid of a couple of huge master files, which is always a good thing...
For details on how to set heap sizes when testing different Derby versions, see my comment to DERBY-1568, http://issues.apache.org/jira/browse/DERBY-1568#action_12436469.
> Backport lang/wisconsin.java at 10.2.0.4-alpha to the 10.1 branch
> -----------------------------------------------------------------
>
> Key: DERBY-1872
> URL: http://issues.apache.org/jira/browse/DERBY-1872
> Project: Derby
> Issue Type: Test
> Components: Test
> Affects Versions: 10.1.3.1
> Environment: N/A
> Reporter: John H. Embretsen
> Assigned To: John H. Embretsen
> Priority: Minor
> Fix For: 10.1.3.2
>
> Attachments: port-wisconsin-from-10.2.0.4-to-10.1_v2.stat, port-wisconsin-from-10.2.0.4-to-10.1_v2.zip
>
>
> Investigations relating to DERBY-1564 ("wisconsin.java test failed in DerbyNet or DerbyNetClient frameworks, VM for network server got OutOfMemoryError") revealed that comparing memory usage for the lang/wisconsin.[java|sql] test in 10.1.x versus 10.2.x was like comparing apples and oranges. After the 10.1 branch was created, several updates were made to the wisconsin test in trunk that were not backported to 10.1, e.g:
> * giving the optimizer unlimited time to choose query plans
> * compressing the tables to avoid the instabilities reported in DERBY-937
> Backporting the test will make it less unfair to compare 10.1 vs. 10.2 (or trunk) test results and memory usage, and it will make it easier to determine if failures (such as DERBY-1564) are regressions or not. On the other hand, other differences between the branches that may influence the test will remain, so this will not be a wonder cure.
> The failures seen in DERBY-1564 were mainly caused by bugs (DERBY-1091, DERBY-1614) in the test harness, causing the wisconsin test for the 10.2.0.4 snapshot to run with low memory settings. These issues have been fixed.
> The most practical backporting approach will probably be to take the wisconsin test as it was in the 10.2.0.4-snapshot (SVN revision 423199), and port it to the current 10.1 branch. Other changes have been made to the wisconsin test in trunk/10.2 after 10.2.0.4, but to avoid various dependencies (e.g. DERBY-1609), these should not be backported.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] Updated: (DERBY-1872) Backport lang/wisconsin.java at
10.2.0.4-alpha to the 10.1 branch
Posted by "Daniel John Debrunner (JIRA)" <de...@db.apache.org>.
[ http://issues.apache.org/jira/browse/DERBY-1872?page=all ]
Daniel John Debrunner updated DERBY-1872:
-----------------------------------------
Ignore last comment, attached to wrong issue.
> Backport lang/wisconsin.java at 10.2.0.4-alpha to the 10.1 branch
> -----------------------------------------------------------------
>
> Key: DERBY-1872
> URL: http://issues.apache.org/jira/browse/DERBY-1872
> Project: Derby
> Issue Type: Test
> Components: Test
> Affects Versions: 10.1.3.1
> Environment: N/A
> Reporter: John H. Embretsen
> Assigned To: John H. Embretsen
> Priority: Minor
> Fix For: 10.1.3.2
>
> Attachments: port-wisconsin-from-10.2.0.4-to-10.1_v2.stat, port-wisconsin-from-10.2.0.4-to-10.1_v2.zip
>
>
> Investigations relating to DERBY-1564 ("wisconsin.java test failed in DerbyNet or DerbyNetClient frameworks, VM for network server got OutOfMemoryError") revealed that comparing memory usage for the lang/wisconsin.[java|sql] test in 10.1.x versus 10.2.x was like comparing apples and oranges. After the 10.1 branch was created, several updates were made to the wisconsin test in trunk that were not backported to 10.1, e.g:
> * giving the optimizer unlimited time to choose query plans
> * compressing the tables to avoid the instabilities reported in DERBY-937
> Backporting the test will make it less unfair to compare 10.1 vs. 10.2 (or trunk) test results and memory usage, and it will make it easier to determine if failures (such as DERBY-1564) are regressions or not. On the other hand, other differences between the branches that may influence the test will remain, so this will not be a wonder cure.
> The failures seen in DERBY-1564 were mainly caused by bugs (DERBY-1091, DERBY-1614) in the test harness, causing the wisconsin test for the 10.2.0.4 snapshot to run with low memory settings. These issues have been fixed.
> The most practical backporting approach will probably be to take the wisconsin test as it was in the 10.2.0.4-snapshot (SVN revision 423199), and port it to the current 10.1 branch. Other changes have been made to the wisconsin test in trunk/10.2 after 10.2.0.4, but to avoid various dependencies (e.g. DERBY-1609), these should not be backported.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] Updated: (DERBY-1872) Backport lang/wisconsin.java at
10.2.0.4-alpha to the 10.1 branch
Posted by "John H. Embretsen (JIRA)" <de...@db.apache.org>.
[ http://issues.apache.org/jira/browse/DERBY-1872?page=all ]
John H. Embretsen updated DERBY-1872:
-------------------------------------
Attachment: port-wisconsin-from-10.2.0.4-to-10.1_v2.stat
port-wisconsin-from-10.2.0.4-to-10.1_v2.zip
Attaching patch for porting the lang/wisconsin.java test from its state in 10.2.0.4 to the 10.1 branch.
This patch, 'port-wisconsin-from-10.2.0.4-to-10.1_v2.zip' ('.diff' when unzipped), replaces the _v1, patch that was attached to DERBY-1564. There is only one change from the _v1 patch:
* Added lang/wisc_setup.sql to lang/copyfiles.ant, so that it gets included in
the build. copyfiles.ant is no longer needed in trunk or 10.2 (DERBY-892), so it
took me a while to figure out what caused my build problem with the _v1-patch,
and why. The build works fine with the _v2 patch.
Here is a summary of changes provided by the patch:
- renames the test from lang/wisconsin.sql to lang/wisconsin.java (DERBY-398)
- adds wisc_setup.sql support file (DERBY-398)
- removes framework-specific master files (for DerbyNet and DerbyNetClient) (DERBY-1014)
- runs compress on tables before retreiving runtimestatistics (DERBY-937)
- adds property derby.optimizer.noTimeout=true to wisconsin_derby.properties (DERBY-413/DERBY-937)
(I do not expect any problems related to DERBY-822 (as mentioned in the _v1 patch description), as the changes relating to the wisconsin test stems from DERBY-1014 _in preparation for_ DERBY-822, not DERBY-822 itself.)
Please note that the patch requires several 'svn add' and 'svn delete' commands, as shown in 'port-wisconsin-from-10.2.0.4-to-10.1_v2.stat'.
The patch is zipped because it is 3.3 MB uncompressed.
I have tested this patch by running:
* lang/wisconsin.java (10.1), single-runs in all three frameworks on
* Solaris 10 x86, Sun JVM 1.5.0_07 - passed
* Solaris 10 x86, Sun JVM 1.4.2_12 - passed
* Solaris 10 x86, Sun JVM 1.3.1_15 - passed
* derbyall on Solaris 10 x86, Sun JVM 1.5.0_07 - 1 unrelated failure, derbynetmats/testSecMec.diff
Please review/commit if you think this adds value to the 10.1 branch.
> Backport lang/wisconsin.java at 10.2.0.4-alpha to the 10.1 branch
> -----------------------------------------------------------------
>
> Key: DERBY-1872
> URL: http://issues.apache.org/jira/browse/DERBY-1872
> Project: Derby
> Issue Type: Test
> Components: Test
> Affects Versions: 10.1.3.1
> Environment: N/A
> Reporter: John H. Embretsen
> Assigned To: John H. Embretsen
> Priority: Minor
> Fix For: 10.1.3.2
>
> Attachments: port-wisconsin-from-10.2.0.4-to-10.1_v2.stat, port-wisconsin-from-10.2.0.4-to-10.1_v2.zip
>
>
> Investigations relating to DERBY-1564 ("wisconsin.java test failed in DerbyNet or DerbyNetClient frameworks, VM for network server got OutOfMemoryError") revealed that comparing memory usage for the lang/wisconsin.[java|sql] test in 10.1.x versus 10.2.x was like comparing apples and oranges. After the 10.1 branch was created, several updates were made to the wisconsin test in trunk that were not backported to 10.1, e.g:
> * giving the optimizer unlimited time to choose query plans
> * compressing the tables to avoid the instabilities reported in DERBY-937
> Backporting the test will make it less unfair to compare 10.1 vs. 10.2 (or trunk) test results and memory usage, and it will make it easier to determine if failures (such as DERBY-1564) are regressions or not. On the other hand, other differences between the branches that may influence the test will remain, so this will not be a wonder cure.
> The failures seen in DERBY-1564 were mainly caused by bugs (DERBY-1091, DERBY-1614) in the test harness, causing the wisconsin test for the 10.2.0.4 snapshot to run with low memory settings. These issues have been fixed.
> The most practical backporting approach will probably be to take the wisconsin test as it was in the 10.2.0.4-snapshot (SVN revision 423199), and port it to the current 10.1 branch. Other changes have been made to the wisconsin test in trunk/10.2 after 10.2.0.4, but to avoid various dependencies (e.g. DERBY-1609), these should not be backported.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] Resolved: (DERBY-1872) Backport lang/wisconsin.java at
10.2.0.4-alpha to the 10.1 branch
Posted by "Andrew McIntyre (JIRA)" <de...@db.apache.org>.
[ http://issues.apache.org/jira/browse/DERBY-1872?page=all ]
Andrew McIntyre resolved DERBY-1872.
------------------------------------
Resolution: Fixed
Derby Info: (was: [Patch Available])
Committed to 10.1 branch with 450904. Thanks, John, as long as there is some way to adjust the heap size when running the test, then it's not as important that we port the other harness changes over. Thanks for looking into this.
> Backport lang/wisconsin.java at 10.2.0.4-alpha to the 10.1 branch
> -----------------------------------------------------------------
>
> Key: DERBY-1872
> URL: http://issues.apache.org/jira/browse/DERBY-1872
> Project: Derby
> Issue Type: Test
> Components: Test
> Affects Versions: 10.1.3.1
> Environment: N/A
> Reporter: John H. Embretsen
> Assigned To: John H. Embretsen
> Priority: Minor
> Fix For: 10.1.3.2
>
> Attachments: port-wisconsin-from-10.2.0.4-to-10.1_v2.stat, port-wisconsin-from-10.2.0.4-to-10.1_v2.zip
>
>
> Investigations relating to DERBY-1564 ("wisconsin.java test failed in DerbyNet or DerbyNetClient frameworks, VM for network server got OutOfMemoryError") revealed that comparing memory usage for the lang/wisconsin.[java|sql] test in 10.1.x versus 10.2.x was like comparing apples and oranges. After the 10.1 branch was created, several updates were made to the wisconsin test in trunk that were not backported to 10.1, e.g:
> * giving the optimizer unlimited time to choose query plans
> * compressing the tables to avoid the instabilities reported in DERBY-937
> Backporting the test will make it less unfair to compare 10.1 vs. 10.2 (or trunk) test results and memory usage, and it will make it easier to determine if failures (such as DERBY-1564) are regressions or not. On the other hand, other differences between the branches that may influence the test will remain, so this will not be a wonder cure.
> The failures seen in DERBY-1564 were mainly caused by bugs (DERBY-1091, DERBY-1614) in the test harness, causing the wisconsin test for the 10.2.0.4 snapshot to run with low memory settings. These issues have been fixed.
> The most practical backporting approach will probably be to take the wisconsin test as it was in the 10.2.0.4-snapshot (SVN revision 423199), and port it to the current 10.1 branch. Other changes have been made to the wisconsin test in trunk/10.2 after 10.2.0.4, but to avoid various dependencies (e.g. DERBY-1609), these should not be backported.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] Commented: (DERBY-1872) Backport lang/wisconsin.java at
10.2.0.4-alpha to the 10.1 branch
Posted by "Daniel John Debrunner (JIRA)" <de...@db.apache.org>.
[ http://issues.apache.org/jira/browse/DERBY-1872?page=comments#action_12438253 ]
Daniel John Debrunner commented on DERBY-1872:
----------------------------------------------
Link to wiki page capturing the performance progress due to changes made related to this issue.
http://wiki.apache.org/db-derby/DerbyBug1872
> Backport lang/wisconsin.java at 10.2.0.4-alpha to the 10.1 branch
> -----------------------------------------------------------------
>
> Key: DERBY-1872
> URL: http://issues.apache.org/jira/browse/DERBY-1872
> Project: Derby
> Issue Type: Test
> Components: Test
> Affects Versions: 10.1.3.1
> Environment: N/A
> Reporter: John H. Embretsen
> Assigned To: John H. Embretsen
> Priority: Minor
> Fix For: 10.1.3.2
>
> Attachments: port-wisconsin-from-10.2.0.4-to-10.1_v2.stat, port-wisconsin-from-10.2.0.4-to-10.1_v2.zip
>
>
> Investigations relating to DERBY-1564 ("wisconsin.java test failed in DerbyNet or DerbyNetClient frameworks, VM for network server got OutOfMemoryError") revealed that comparing memory usage for the lang/wisconsin.[java|sql] test in 10.1.x versus 10.2.x was like comparing apples and oranges. After the 10.1 branch was created, several updates were made to the wisconsin test in trunk that were not backported to 10.1, e.g:
> * giving the optimizer unlimited time to choose query plans
> * compressing the tables to avoid the instabilities reported in DERBY-937
> Backporting the test will make it less unfair to compare 10.1 vs. 10.2 (or trunk) test results and memory usage, and it will make it easier to determine if failures (such as DERBY-1564) are regressions or not. On the other hand, other differences between the branches that may influence the test will remain, so this will not be a wonder cure.
> The failures seen in DERBY-1564 were mainly caused by bugs (DERBY-1091, DERBY-1614) in the test harness, causing the wisconsin test for the 10.2.0.4 snapshot to run with low memory settings. These issues have been fixed.
> The most practical backporting approach will probably be to take the wisconsin test as it was in the 10.2.0.4-snapshot (SVN revision 423199), and port it to the current 10.1 branch. Other changes have been made to the wisconsin test in trunk/10.2 after 10.2.0.4, but to avoid various dependencies (e.g. DERBY-1609), these should not be backported.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] Updated: (DERBY-1872) Backport lang/wisconsin.java at
10.2.0.4-alpha to the 10.1 branch
Posted by "John H. Embretsen (JIRA)" <de...@db.apache.org>.
[ http://issues.apache.org/jira/browse/DERBY-1872?page=all ]
John H. Embretsen updated DERBY-1872:
-------------------------------------
Derby Info: [Patch Available]
> Backport lang/wisconsin.java at 10.2.0.4-alpha to the 10.1 branch
> -----------------------------------------------------------------
>
> Key: DERBY-1872
> URL: http://issues.apache.org/jira/browse/DERBY-1872
> Project: Derby
> Issue Type: Test
> Components: Test
> Affects Versions: 10.1.3.1
> Environment: N/A
> Reporter: John H. Embretsen
> Assigned To: John H. Embretsen
> Priority: Minor
> Fix For: 10.1.3.2
>
> Attachments: port-wisconsin-from-10.2.0.4-to-10.1_v2.stat, port-wisconsin-from-10.2.0.4-to-10.1_v2.zip
>
>
> Investigations relating to DERBY-1564 ("wisconsin.java test failed in DerbyNet or DerbyNetClient frameworks, VM for network server got OutOfMemoryError") revealed that comparing memory usage for the lang/wisconsin.[java|sql] test in 10.1.x versus 10.2.x was like comparing apples and oranges. After the 10.1 branch was created, several updates were made to the wisconsin test in trunk that were not backported to 10.1, e.g:
> * giving the optimizer unlimited time to choose query plans
> * compressing the tables to avoid the instabilities reported in DERBY-937
> Backporting the test will make it less unfair to compare 10.1 vs. 10.2 (or trunk) test results and memory usage, and it will make it easier to determine if failures (such as DERBY-1564) are regressions or not. On the other hand, other differences between the branches that may influence the test will remain, so this will not be a wonder cure.
> The failures seen in DERBY-1564 were mainly caused by bugs (DERBY-1091, DERBY-1614) in the test harness, causing the wisconsin test for the 10.2.0.4 snapshot to run with low memory settings. These issues have been fixed.
> The most practical backporting approach will probably be to take the wisconsin test as it was in the 10.2.0.4-snapshot (SVN revision 423199), and port it to the current 10.1 branch. Other changes have been made to the wisconsin test in trunk/10.2 after 10.2.0.4, but to avoid various dependencies (e.g. DERBY-1609), these should not be backported.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] Commented: (DERBY-1872) Backport lang/wisconsin.java at
10.2.0.4-alpha to the 10.1 branch
Posted by "John H. Embretsen (JIRA)" <de...@db.apache.org>.
[ http://issues.apache.org/jira/browse/DERBY-1872?page=comments#action_12438363 ]
John H. Embretsen commented on DERBY-1872:
------------------------------------------
Hi Andrew, thanks for looking at this...
As far as I know there are no failures on 10.1 that this patch will fix. Hopefully it won't introduce any either... but wisconsin is still a bit unstable in trunk (DERBY-1801), so you never know. But yes, the main point of this patch is just so that memory profiling using this test is more consistent across branches. Here's some history that may help you understand the background for this patch:
I ported the 10.2.0.4 wisconsin test to the 10.1 branch so that I could have a better foundation for determining whether or not DERBY-1564 was a 10.2 regression or not. Since I already had a patch for porting the test, Kathey (Marsden) asked me to submit it so that it could be committed to the 10.1 branch. That is what I did, after polishing it just a bit.
So I leave it to you and Kathey (and other interested parties, if any) to decide whether or not to commit the patch.
Yes, I agree that porting DERBY-1091 and DERBY-1614 to 10.1 would be good, but I have been reluctant to look into that too closely, since I fear that would open up another can of worms. Right now the jvmflags are not passed properly from the test _app.properties files in 10.1, but it seems to be possible to set a specific heap size at least, using the jvmflags property from the command line.
> Backport lang/wisconsin.java at 10.2.0.4-alpha to the 10.1 branch
> -----------------------------------------------------------------
>
> Key: DERBY-1872
> URL: http://issues.apache.org/jira/browse/DERBY-1872
> Project: Derby
> Issue Type: Test
> Components: Test
> Affects Versions: 10.1.3.1
> Environment: N/A
> Reporter: John H. Embretsen
> Assigned To: John H. Embretsen
> Priority: Minor
> Fix For: 10.1.3.2
>
> Attachments: port-wisconsin-from-10.2.0.4-to-10.1_v2.stat, port-wisconsin-from-10.2.0.4-to-10.1_v2.zip
>
>
> Investigations relating to DERBY-1564 ("wisconsin.java test failed in DerbyNet or DerbyNetClient frameworks, VM for network server got OutOfMemoryError") revealed that comparing memory usage for the lang/wisconsin.[java|sql] test in 10.1.x versus 10.2.x was like comparing apples and oranges. After the 10.1 branch was created, several updates were made to the wisconsin test in trunk that were not backported to 10.1, e.g:
> * giving the optimizer unlimited time to choose query plans
> * compressing the tables to avoid the instabilities reported in DERBY-937
> Backporting the test will make it less unfair to compare 10.1 vs. 10.2 (or trunk) test results and memory usage, and it will make it easier to determine if failures (such as DERBY-1564) are regressions or not. On the other hand, other differences between the branches that may influence the test will remain, so this will not be a wonder cure.
> The failures seen in DERBY-1564 were mainly caused by bugs (DERBY-1091, DERBY-1614) in the test harness, causing the wisconsin test for the 10.2.0.4 snapshot to run with low memory settings. These issues have been fixed.
> The most practical backporting approach will probably be to take the wisconsin test as it was in the 10.2.0.4-snapshot (SVN revision 423199), and port it to the current 10.1 branch. Other changes have been made to the wisconsin test in trunk/10.2 after 10.2.0.4, but to avoid various dependencies (e.g. DERBY-1609), these should not be backported.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] Commented: (DERBY-1872) Backport lang/wisconsin.java at
10.2.0.4-alpha to the 10.1 branch
Posted by "Andrew McIntyre (JIRA)" <de...@db.apache.org>.
[ http://issues.apache.org/jira/browse/DERBY-1872?page=comments#action_12438272 ]
Andrew McIntyre commented on DERBY-1872:
----------------------------------------
Hi John,
These changes look fine, and I will commit them if you like, since it is just a rearrangement of the test.
I'm a little unclear what the benefit to testing on the 10.1 codeline will be. Are there failures on 10.1 on a particular platform/VM combination that fail without these changes? Or is it just so that memory profiling using this test is more consistent across branches? I've tried to follow all the discussion in DERBY-1564 and elsewhere, and I'm wondering if it wouldn't also make sense to also port the DERBY-1641 (and maybe also DERBY-1091) harness changes to 10.1 so that we can be sure of the jvmflags that are being passed to the test?
Anyway, I'll wait to hear if you have any comments, otherwise I'll commit these changes tomorrow.
> Backport lang/wisconsin.java at 10.2.0.4-alpha to the 10.1 branch
> -----------------------------------------------------------------
>
> Key: DERBY-1872
> URL: http://issues.apache.org/jira/browse/DERBY-1872
> Project: Derby
> Issue Type: Test
> Components: Test
> Affects Versions: 10.1.3.1
> Environment: N/A
> Reporter: John H. Embretsen
> Assigned To: John H. Embretsen
> Priority: Minor
> Fix For: 10.1.3.2
>
> Attachments: port-wisconsin-from-10.2.0.4-to-10.1_v2.stat, port-wisconsin-from-10.2.0.4-to-10.1_v2.zip
>
>
> Investigations relating to DERBY-1564 ("wisconsin.java test failed in DerbyNet or DerbyNetClient frameworks, VM for network server got OutOfMemoryError") revealed that comparing memory usage for the lang/wisconsin.[java|sql] test in 10.1.x versus 10.2.x was like comparing apples and oranges. After the 10.1 branch was created, several updates were made to the wisconsin test in trunk that were not backported to 10.1, e.g:
> * giving the optimizer unlimited time to choose query plans
> * compressing the tables to avoid the instabilities reported in DERBY-937
> Backporting the test will make it less unfair to compare 10.1 vs. 10.2 (or trunk) test results and memory usage, and it will make it easier to determine if failures (such as DERBY-1564) are regressions or not. On the other hand, other differences between the branches that may influence the test will remain, so this will not be a wonder cure.
> The failures seen in DERBY-1564 were mainly caused by bugs (DERBY-1091, DERBY-1614) in the test harness, causing the wisconsin test for the 10.2.0.4 snapshot to run with low memory settings. These issues have been fixed.
> The most practical backporting approach will probably be to take the wisconsin test as it was in the 10.2.0.4-snapshot (SVN revision 423199), and port it to the current 10.1 branch. Other changes have been made to the wisconsin test in trunk/10.2 after 10.2.0.4, but to avoid various dependencies (e.g. DERBY-1609), these should not be backported.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira