[kepler-dev] Workflows non-terminating in nightly build

Kevin Ruland kruland at ku.edu
Wed Feb 8 07:18:21 PST 2006



Dan,

Are you referring to these exceptions:

Workflow workflows/eco/eml-simple-plot.moml FAILED: 
> [java] java.sql.SQLException: Table not found: CACHECONTENTTABLE in statement [insert into cacheContentTable (name, lsid, date, file, expiration) values ( ?, ?, ?, ?, ? )] [java] java.lang.ExceptionInInitializerError [java] Caused by: java.lang.NullPointerException [java] java.lang.ExceptionInInitializerError [java] Caused by: java.lang.NullPointerException [java] java.lang.IllegalStateException: zip file closed
> 

etc.

There is something weird and wacky going on.  It looks like either the .kepler cache isn't getting flushed or the config.xml isn't up to date.  I'll see if Chad won't bless me with ~kepler access to verify.

Kevin



Dan Higgins wrote:
> Incidently, workflows with eml2-datasources are still failling when run 
> with ptexecute (but at least the nightly build script is reporting this 
> and continuing)
>
> Dan
>
> --
>
> Matt Jones wrote:
>
>   
>> Maybe you should use the test actors to test for valid results rather 
>> than just running a larger workflow like PIW.  There's really no way 
>> to tell what went wrong if PIW etc. fails in its current 
>> configuration. Using the test actors generates specific exceptions 
>> when the test fails and so it lets you pinpoint the critical stuff and 
>> test if the right values are being produced for known inputs. I think 
>> that would probably fix many of these issues.
>>
>> Matt
>>
>> Dan Higgins wrote:
>>
>>     
>>> Hi Efrat,
>>>    It looks like your workflow 'workflows/srb/srbPhysLoc.xml is 
>>> 'hanging' in the nightly build (and when I try to run it locally). I 
>>> would guess that it should run in a few seconds, but it is still 
>>> 'executing' after ~ 1hr. Can you take a look at it. In the mean time 
>>> I am going to remove it from the nightly test workflow list (?).
>>>
>>> To Kepler-dev:
>>>    Since we apparently have had several problems with workflows that 
>>> do not terminate, we may have to consider including timeouts in our 
>>> workflows ?  (To handle the cases where actors are not executing 
>>> locally.) Someone may not always be looking and be able to close an 
>>> executing workflow
>>>
>>> Dan
>>>
>>> ---
>>>
>>> Chad Berkley wrote:
>>>
>>>       
>>>> We had that problem before.  Sometimes the PIW WF doesn't finish.  
>>>> I  can't remember what i did to fix it the last time though.
>>>>
>>>> chad
>>>>
>>>> On Feb 7, 2006, at 10:22 AM, Dan Higgins wrote:
>>>>
>>>>         
>>>>> Hi Ilkay,
>>>>>    I have been investigating why our nightly build has not been  
>>>>> working correctly and discovered that we had a bunch of 'hung'  
>>>>> kepler processes. Apparently this is occurring when the nightly  
>>>>> build script is trying to execute the SPA PIW workflow. It looks  
>>>>> like that workflow may never terminate! Could you check it please.  
>>>>> In the mean time, I am going to try removing it from the list of  
>>>>> workflows to test (for the time being, anyway).
>>>>>
>>>>>    Incidently, in case you are wondering why this problem just 
>>>>> came  up, I thing it is because we previously did not have X11 set 
>>>>> up on  the server. It was just added, and now the server actually 
>>>>> tried to  run the test workflows! (for the fiirst time in a long 
>>>>> while).
>>>>>
>>>>> Dan
>>>>>
>>>>> -- 
>>>>> *******************************************************************
>>>>> Dan Higgins                                  higgins at nceas.ucsb.edu
>>>>> http://www.nceas.ucsb.edu/    Ph: 805-893-5127
>>>>> National Center for Ecological Analysis and Synthesis (NCEAS)  
>>>>> Marine Science Building - Room 3405
>>>>> Santa Barbara, CA 93195
>>>>> *******************************************************************
>>>>>
>>>>>
>>>>>
>>>>>           
>>>       
>
>
>   



More information about the Kepler-dev mailing list