Hi Ilja,

Thanks for noting this. I have not yet encountered this error, but it definitely would be worth trying to find a way to reliably reproduce it (if possible).


From the error stacktrace you linked to, it looks like the problem originates in the UploadWithEmbargoStep of the submission process (at least in this example). If you are able to recreate this error again, it might be interesting to see what is similar between each of the stacktraces. Is it a single submission step to blame? Does the error occur more randomly (in different submission steps)?

Essentially if we can find a similarity between these stacktraces or a way to reliably reproduce the error, it would help us in determining whether this is only encountered with particular configurations/setups, and/or what the root cause may be.

You are welcome to create a ticket for investigating this issue, as that may be a better place to add notes should anyone encounter this issue. If we find it to be a non-issue, we can always close the ticket.

Thanks, and please do let us know if you find ways to reproduce or similarities between the error stacktraces.

- Tim

On 9/22/2016 5:28 AM, Ilja Sidoroff wrote:
I've been testing the latest git master (upcoming DS 6), and since pulling the 
latest changes on Tuesday, I've encountered errors like:

ERROR: prepared statement "S_29" does not exist
(a paste for the full stack trace: http://pastebin.com/2mLKNNHD)

(where "S_29" varies). The errors seem to happen when I use workflow functionality 
and try to submit something, or directly try to submit something to a collection. When the 
error state is on, I cannot start new submissions, upload bitstreams from the workflow, or 
upload bitstreams from administration->edit item - and possibly more. I don't know 
exactly what are the conditions for reproducing this nor I know how to recover from the 
state. I managed recovery once by restarting postgres, but in general it doesn't seem to 
help.

I first encountered this on Tuesday and asked help on irc. Helix84 did some 
googling and one reason this might happen, is some sort of concurrency problem, 
where the original PSTMT gets deleted/flushed out and is not available when 
resumed (unfortunately I don't have the reference or the irc log).

My environment is: Ubuntu 16.04, Java 1.8, Jetty 9, PostgreSQL 9.5. I'll 
continue examining the issue to determine if it is just something in my own 
environment or a genuine bug.

br,

Ilja Sidoroff
University of Eastern Finland


--
Tim Donohue
Technical Lead for DSpace & DSpaceDirect
DuraSpace.org | DSpace.org | DSpaceDirect.org

--
You received this message because you are subscribed to the Google Groups "DSpace 
Technical Support" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to dspace-tech+unsubscr...@googlegroups.com.
To post to this group, send email to dspace-tech@googlegroups.com.
Visit this group at https://groups.google.com/group/dspace-tech.
For more options, visit https://groups.google.com/d/optout.

Reply via email to