Derby 10.3.1.1 RC1 - summary testreport

Release candidate 10.3.1.1 was created 2007-07-09 (svn revision 554502) and testing was started that same day.


Testing was done under jvm1.6 with individual tests (derbyall subsuites) and derbyall and suitesAll. Under jvm1.5 and jvm1.4 derbyall and suitesAll was run.

Here is a table with a summary of failures found including comments on these:

Counting the number of distinct errors, total number of error occurences and platforms used in 10.3.1.1 testing we find:

 

10.2.2.1

10.3.1.0

10.3.1.1

Distinct
errors

9

6

10

1.4

3 / 5 / 21

4 / 4 / 21

6 / 23 / 21

1.5

2 / 2 / 22

3 / 5 / 22

6 / 30 / 22

1.6

5 / 5 / 21

1 / 3 / 21

3 / 29 / 21

Legend: Distinct errors / Sum errors / Number of platforms.

The DOTS/ATCJ4(previously called ATCJ2) part of the Large DataVolume test was run for 48 hrs on a database with ATCJ4 tables of 7, 299, 232 million tuples, using 183 GB raw disk space. No issues seen.

JDBC 3.0 Compatibility Test Suite (CTS) (part of the J2EE TCK) was run on a number of different platforms. See SunCTS for details.

The compatibility test (CompatibilityCombinations/JDBC) was not run on all platforms except for Windows. An environmental issue causes these to fail.
A LongTermTest was started. See LongTermTest for details. No derby-failures seen.
The upgrade test is now run as part of suitesAll.

Report generated: Fri July 13 13:25:23 MEST 2007

JIRA/Id

JVM/OS

Comment

Seen in XX

1585 suitesAll: org.apache.derbyTesting.functionTests.tests.lang.ProcedureInTriggerTest

jvm1.4: []
jvm1.5: [lin linN-1]
jvm1.6: []

 

10.3.1.0
10.2.2.1
10.2.2.0
10.2.1.7
10.2.1.6

1750 derbyall: stress/stress.multi

jvm1.4: [sol]
jvm1.5: []
jvm1.6: []

...running last checks via final.sql
...timed out trying to kill all testers, ....

10.3.1.0

2290 lang/logStream.java:

jvm1.4: [vista]
jvm1.5: []
jvm1.6: []

 

10.2.2.1
10.3.1.0

2434 suitesAll:org.apache.derbyTesting.functionTests.tests.jdbcapi.JDBCHarnessJavaTest:

jvm1.4: [vista]
jvm1.5: [vista w2000 w2003]
jvm1.6: []

 

10.3.1.0

2636 ImportExportProcedureTestjunit? .framework.AssertionFailedError:

jvm1.4: []
jvm1.5: [w2000]
jvm1.6: []

extinout\t1.dat

 

2842 derbynetclientmats/derbynetmats, derbynetmats/DerbyNet

jvm1.4: [vista w2003 winxp]
jvm1.5: [hpux vista w2003 winxp]
jvm1.6: [vista w2000 w2003 winxp]

Could not connect to Derby Network Server on host xxxFILTERED_HOSTNAMExxx, port 9393: Connection refused
Could not connect to Derby Network Server on host xxxFILTERED_HOSTNAMExxx, port 9393: Connection refused: connect

 

2843 derbynet/testProperties.java:

jvm1.4: []
jvm1.5: []
jvm1.6: [sles]

 

 

JIRA-2871 org.apache.derbyTesting.functionTests.tests.jdbcapi.XATransactionTest:

jvm1.4: [sparcN-2]
jvm1.5: []
jvm1.6: []

testXATransactionTimeout(org.apache.derbyTesting.functionTests.tests.jdbcapi.XATransactionTest)org.apache.derby.client.am.XaException: XAER_PROTO : Error executing a XAResource.commit(), server returned XAER_PROTO.

10.3.1.0

2899 statementJdbc20:

jvm1.4: []
jvm1.5: [w2000]
jvm1.6: []

ComparisonFailure: Output at line 3 expected:
... parameter value '-1[,]000' for Statement.s...
but was:
... parameter value '-1[?]000' for Statement.s...

 

2915 noConnectionAfterHardUpgrade

jvm1.4: [all except windows]
jvm1.5: [all except windows]
jvm1.6: [all]

 

 

The following are NOT Derby failures.Not counted above!

???? derbynetmats/DerbyNet

jvm1.4: []
jvm1.5: [zone1 zone2]
jvm1.6: [zone2 zone3]

derby.drda.host=xxxFILTERED_HOSTNAMExxx
derby.drda.host=127.0.0.1

 

???? derbynetmats/DerbyNet

jvm1.4: []
jvm1.5: []
jvm1.6: [zone2 zone3]

com.ibm.db2.jcc.c.SqlException: java.net.ConnectException :
Error opening socket to server xxxFILTERED_HOSTNAMExxx on port 1527 with message : Connection refused
derby.drda.host=xxxFILTERED_HOSTNAMExxx
derby.drda.host=127.0.0.1
derby.drda.traceAll=false
derby.drda.traceAll=true

 

???? derbynet.ShutDownDBWhenNSShutsDownTest

jvm1.4: []
jvm1.5: [zones]
jvm1.6: []

java.sql.SQLException: java.net.SocketException :
Error connecting to server localhost on port 1527 with message Address already in use.

 

Conclusion

We do not regard any of the found problems as showstoppers for a next release, i.e. none of these failures have to be fixed before a next release of JavaDB can be created. But it would be nice to get JIRA-2842 and JIRA-2915 fixed, to get rid of many failures. In addition we should consider have a closer look at zones-testing and the failures reported here.