SRMA - java.lang.StackOverflowError

[YannickRenaud]

Hello community,

I’ve been trying for the last week to make SRMA finish a file.

Below is what I’ve done:
1-I started SRMA on a file and it crashed after 3hours giving the following error:

MRecords processsed: 4092848 (70720198-70720295)[Fri Apr 12 14:13:41 EDT 2
013] srma.SRMA done.
Runtime.totalMemory()=7082278912
Exception in thread "main" java.lang.StackOverflowError
at java.util.ArrayList$SubList.removeRange(ArrayList.java:944)
at java.util.ArrayList$SubList.removeRange(ArrayList.java:944)
at java.util.ArrayList$SubList.removeRange(ArrayList.java:944)
…

2-After reading abit, I’ve added it more memory (-Xmx20G). The file now crashes after 5 hours giving the following error:

8551062 (last 1:107972936-107973033)[Mon Apr 15 16:06:58 EDT 2013] srma.SRMA done.
Runtime.totalMemory()=4025548800
Exception in thread "main" java.lang.StackOverflowError
at java.util.ArrayList$SubList.removeRange(ArrayList.java:944)
at java.util.ArrayList$SubList.removeRange(ArrayList.java:944)
…

3-I’ve tried playing with the MAX_QUEUE_SIZE and with –Xss20G but did not see any differences


Is there anything I can do for it to finish my file without splitting it?
The only option I see now is to use SRMA on multiple file using the RANGES option.
Right now it stops at ~3% of the my bam and that would mean I need to split my genome into ~33 files!

Thanks

Yannick