[ 
http://issues.apache.org/jira/browse/DERBY-1222?page=comments#action_12374661 ] 

Myrna van Lunteren commented on DERBY-1222:
-------------------------------------------

This is not a test specific issue; I can see a problem when I run with just ij.

The problem is reproduced easily by doing the following on os/390 - zOS:
java -Dderby.ui.codeset=EUC_JP org.apache.derby.tools.ij
[gobbligook removed for jira] connect 'jdbc:derby:tstdb;create=true';
[gobbligook removed for jira] create table t1 (c1 int);
...tons and tons and tons of more gobbligook.

The first bits of gobbligook are apparently what the ij prompt looks like when 
you try to do a derby.ui.codeset of EUC_JP on a system that normally has 
default encoding of Cp1047. 
>From various native2ascii experiments, it turns out what is returned in a 
>loopy way (i.e. it apparently loops forever until the process is killed, 
>gobbling up resources on the machine as it goes) is:
ij > IJ ERROR: IOException: null

There is no stack trace, because I have to kill the process.

When I first tested this test with version 10.1.1.0 on this platform, I did not 
have the changes for DERBY-658 in place, so I had to convert the files using 
native2ascii -encoding Cp1047 -reverse but it looks like that did not work and 
the file UnicodeEscape_JP.sql actually never got converted and the test failed 
with a FileNotFound exception from the harness. So this may not be a regression 
per se. I could convert the UnicodeEscape_JP.sql file this time, however. 

Either way I think the looping is worrysome. 


> test i18n/UnicodeEscape_JP.sql appears to hang on ibm15 z/OS.
> -------------------------------------------------------------
>
>          Key: DERBY-1222
>          URL: http://issues.apache.org/jira/browse/DERBY-1222
>      Project: Derby
>         Type: Bug

>     Versions: 10.1.2.4
>  Environment: zOS
>     Reporter: Myrna van Lunteren

>
> The test i18n/UnicodeEscape.sql appears to hang on z/OS with ibm15, both 
> latest builds of 10.1 and 10.2.
> The test only takes about 3 seconds on a (fast) linux box, but I've had it 
> sit for hours - i.e. past the 1 hour time out of the test harness - on zOS. 
> When I finally kill the ij process kicked off by the test, the system shows 
> indications of major stress, like javacore, and heapdumps, and OutOfMemory 
> errors. Not sure that the OutOfMemory is caused by the whatever trouble is 
> brewing, or that the trouble is caused by the OutOfMemory...
> This could be a test issue or test harness issue, needs further research.
> This happens both with ibm14 and ibm15, with the patch for 658, both 10.1.2.4 
> and 10.2.

-- 
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

Reply via email to