We've moved!
This site is now read-only. You can find our new documentation site and support forum for posting questions here.
Be sure to read our welcome blog!

IO Error while running MergeBamAlignment on Terra

Hello,

I'm currently running a sample through snapshot 8 of the "processing-for-variant-discovery-gatk4" workflow on Terra, and I'm consistently hitting the following error on the MergeBamAlignment step:

INFO 2019-07-02 03:36:35 AbstractAlignmentMerger Merged 375,000,000 records. Elapsed time: 02:27:06s. Time for last 1,000,000: 23s. Last read position: chr8:137,530,448
INFO 2019-07-02 03:36:35 AbstractAlignmentMerger 24457599 Reads have been unmapped due to being suspected of being Cross-species contamination.
[Tue Jul 02 03:36:52 UTC 2019] picard.sam.MergeBamAlignment done. Elapsed time: 147.42 minutes.
Runtime.totalMemory()=7337410560
To get help, see http://broadinstitute.github.io/picard/index.html#GettingHelp
htsjdk.samtools.util.RuntimeIOException: Write error; BinaryCodec in writemode; streamed file (filename not available)
at htsjdk.samtools.util.BinaryCodec.writeBytes(BinaryCodec.java:222)
at htsjdk.samtools.util.BlockCompressedOutputStream.writeGzipBlock(BlockCompressedOutputStream.java:429)
at htsjdk.samtools.util.BlockCompressedOutputStream.deflateBlock(BlockCompressedOutputStream.java:392)
at htsjdk.samtools.util.BlockCompressedOutputStream.write(BlockCompressedOutputStream.java:291)
at htsjdk.samt

A few notes:
1. I've tried increasing the java_opt and mem_size inputs of MergeBamAlignment to "-Xms7000m" and "7500 MB" respectively. Unsurprisingly, this didn't work, as the error seems to be in writing the output of MergeBamAlignment.
2. I've shared the workspace "pd-wgs-workspace" in the project "pd-wgs-project" with the FireCloud support group. The offending sample is "HDF111iPS501_D0_r1".
3. Attached is the full stderr log.

Thank you!
Lee

Answers

Sign In or Register to comment.