Thanks, Terry:
I will look at those changes. However, I was able to get my OAI to work
for now. I had a two fold problem. I originally had two “bad” records
that were causing the OAI indexing to stop working, but once I had those
identified and taken care of, I ran into the hibernate problems,
George,
I was seeing out of memory errors for the oai import process. If that is
the problem you are encountering, the following PR has resolved the issue
for me.
https://github.com/DSpace/DSpace/pull/2320
Terry
On Mon, Jan 7, 2019 at 1:03 PM George Kozak wrote:
> Hi, Mark:
> In the
Hi, Mark:
In the postgres log file, I am seeing "The postmaster has commanded this
server process to roll back the current transaction and exit, because
another server process exited abnormally and possibly corrupted shared
memory." whenever this fails. I am suspecting maybe an out of memory
Thanks, Mark:
This hibernate stuff is a mystery to me, but I will continue digging to try
and fix my problem.
George Kozak
Cornell University
On Mon, Jan 7, 2019 at 12:07 PM Mark H. Wood wrote:
> I see that this is caused by a connection reset. Is PostgreSQL logging
> anything interesting
I see that this is caused by a connection reset. Is PostgreSQL logging
anything interesting around the time that this happens?
Remember that the DSpace connection limit should be a bit *lower* than the
Pg connection limit, because Pg reserves a few connections for
administration.
I found
I just upgraded to DSpace 6.3 today and I have not yet been able to get the
oai import to complete. ~400,000 items that I need to import. I am
finding that the process keeps running out of memory.
I had did not have problems running this in DSpace 5.x. I will post back
if I discover something.
Tim:
I checked the max connections in postgres and dspace, and there was a
discrepancy, but after fixing that, I am still having the same problem. I
double checked the hibernate configs, and they are the defaults. So, I
have attached the error trace from my oai indexing job. Perhaps, that will
Thanks, Tim:
I will take a look at the postgres configuration. When we upgraded, we,
also, migrated to a new server and maybe there's a problem with the
postgres settings. I didn't think about that, so I am glad you suggested
it.
If that's not it, I will be happy to send a more detailed trace of
Hi George,
That is an odd error. It almost sounds like Hibernate is losing access to
the database (maybe a connection timeout or something?).
Could you check your logs to see if there's more information around *when*
and *how* this error occurs. You said it's occurring at different
Hi, Terry:
I am using jdk1.8.0_74 so java 7 is not the problem. We did upgrade a few
months ago, and ran into trouble immediately with the OAI index. Emilio
Lorenzo and Adan Roman gave me a patch to org.dspace.xoai.app.XOAI.java to
print out records being indexed in verbose mode, and that helped
George,
I am not seeing this error.
Is there any chance that you are using an older version of java (such as
Java 7) when you run the dspace command?
Did you upgrade recently?
On Thu, Jan 3, 2019 at 11:02 AM George Kozak wrote:
> Hi,
> We are using DSpace 6.3 and we are seeing an error when
Hi,
We are using DSpace 6.3 and we are seeing an error when running the dspace
oai import -c. The Import fails with
"org.hibernate.exception.JDBCConnectionException: could not extract
ResultSet". Has anyone else encountered this problem?
--
***
George Kozak
Digital
12 matches
Mail list logo
