No clue, I am not a hardware expert. Removing memory extensions one by one (or binary-searching for the faulty one)?
Dawid On Wed, Jan 11, 2012 at 10:47 AM, Frank Moss <frankmoss1...@gmail.com> wrote: > The same with IBM J9. The dump file is attached. > > It seems to be HW related. Recently, we have added more RAM. We actually > removed old memory bricks and put a new set. Before increasing the RAM, I > kept getting OutOfMemoryError. > > How do I check what exactly the problem is and with what hardware? I run the > memory check and it did not detect any errors on OpenSuse 11.3, 64bit. > > > > On Wed, Jan 11, 2012 at 10:06 AM, Dawid Weiss <dawid.we...@gmail.com> wrote: >> >> Opps, yes, sorry -- I only quickly looked at the invocation line on >> stack overflow and overlooked it. -Xms4g shouldn't make any >> difference. >> >> Dawid >> >> On Wed, Jan 11, 2012 at 10:02 AM, Frank Moss <frankmoss1...@gmail.com> >> wrote: >> > 4gb is the initial heap size. Are you thinking about Xss? I will try >> > it >> > as well as the rest of your suggestions and post back the results. >> > >> > Thanks. >> > >> > >> > On Wed, Jan 11, 2012 at 9:56 AM, Dawid Weiss <dawid.we...@gmail.com> >> > wrote: >> > >> >> The dump you're getting indicates a sigserv in a garbage collection. >> >> This isn't unlikely (there are bugs in there as well), but less likely >> >> than a hardware error on your side... at least in my opinion. I would >> >> experiment with the following: >> >> >> >> 1) do you really need a 4gb max stack? Seems weird to me. >> >> 2) collect the dump from jrockit; is it something related to memory >> >> collection? If so this would suggest memory issues again >> >> 3) try ibm j9 and see if you can reproduce the problem. If so, again >> >> -- I'd double check the hardware. >> >> 4) collect more dumps from hotspot (same version) and compare them >> >> against each other; is the problem recurring in the same place >> >> (gc-related)? If so, then run with the -fastdebug VM (unfortunately >> >> you'll probably need to build this one yourself these days) and make >> >> the core dump available to gc mailing list folks, they'll be able to >> >> investigate. >> >> >> >> Dawid >> >> >> >> On Wed, Jan 11, 2012 at 9:50 AM, Frank Moss <frankmoss1...@gmail.com> >> >> wrote: >> >> > Thanks for your quick reply. >> >> > >> >> > As I said, it happens with the latest version of VM, i.e, both >> >> > "jdk1.6.0_30" and "jdk1.7.0_02" >> >> > >> >> > The memory is not an ECC one it seems. >> >> > >> >> > I have a kit of this (3*4GB * 2 = 24GB): >> >> > >> >> > http://www.kingston.com/datasheets/KHX1600C9S3K2_8GX.pdf >> >> > >> >> > On Wed, Jan 11, 2012 at 9:33 AM, Dawid Weiss <dawid.we...@gmail.com> >> >> wrote: >> >> > >> >> >> This is a fairly old VM you're running with, but if you get the same >> >> >> error with jrockit then I would assume it's a hardware problem. Any >> >> >> other software crashing? Do you have ECC-memory? Have you tried >> >> >> IBM's >> >> >> J9 VM as well maybe? >> >> >> >> >> >> Dawid >> >> >> >> >> >> >> >> >> On Wed, Jan 11, 2012 at 9:28 AM, Frank Moss >> >> >> <frankmoss1...@gmail.com> >> >> >> wrote: >> >> >> > I have a collection of 50 million documents and I hit the SIGSEGV >> >> error. >> >> >> > For every 10000 documents I perform commit. >> >> >> > The logs and the question has been posted to SO here: >> >> >> http://bit.ly/xyZUEG >> >> >> > where I was encouraged to post this question to this mailing list. >> >> >> > >> >> >> > I have tested with different versions of java both in 6.0 and 7.0 >> >> release >> >> >> > lines (including older versions). I get the same error when >> >> >> > running >> >> with >> >> >> > Jrockit jvm. >> >> >> > Memory test was ok. >> >> >> > >> >> >> > Anybody had similar issue? Any tips? >> >> >> > >> >> >> > Thanks. >> >> >> > Regards, >> >> >> > FM >> >> >> >> >> >> >> >> >> --------------------------------------------------------------------- >> >> >> To unsubscribe, e-mail: java-user-unsubscr...@lucene.apache.org >> >> >> For additional commands, e-mail: java-user-h...@lucene.apache.org >> >> >> >> >> >> >> >> >> >> --------------------------------------------------------------------- >> >> To unsubscribe, e-mail: java-user-unsubscr...@lucene.apache.org >> >> For additional commands, e-mail: java-user-h...@lucene.apache.org >> >> >> >> >> >> --------------------------------------------------------------------- >> To unsubscribe, e-mail: java-user-unsubscr...@lucene.apache.org >> For additional commands, e-mail: java-user-h...@lucene.apache.org >> > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: java-user-unsubscr...@lucene.apache.org > For additional commands, e-mail: java-user-h...@lucene.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: java-user-unsubscr...@lucene.apache.org For additional commands, e-mail: java-user-h...@lucene.apache.org
