java.lang.OutOfMemoryError: Java heap space error

[thomasyu888]

The distribution of variants are:

76410
15354
13084
226854
17616
47908
6115
326934
3399
324668
55294
49259
268895
123123
889
35805
10275
2122

Some of these are combinations of VCFs, but the largest MAF is 326934 rows. I am currently using a t3.medium instance which is 2CPU and 4GB mem.

[sheridancbio]

👍 thanks for the report

[sheridancbio]

Initial testing shows that for a MAF with 379772 records, the memory usage seen (on dashi-dev) was 37652044 bytes virtual memory (~37GB) and 6463572 bytes resident memory (~6.2G)

[ao508]

@sheridancbio @thomasyu888 is this issue okay to close for now since Tom will be increasing his node size or is there still work to do?

[sheridancbio]

t3.medium node was insufficient for mskimpact maf
not sure if a success has happened yet and what size node was used.

[ao508]

Follow up question - did anyone attempt to run the annotator with the jvm argument to increase the heap size?

[thomasyu888]

I just encountered this error with a mutation file that is 240 MB, and 226870 rows. How do I increase the heap size?

I am now on a t3.xlarge machine. 4 cores and 16 GB ram.

Annotator logs show:

2020-04-30 04:13:30 [main] INFO  org.cbioportal.annotator.internal.GenomeNexusImpl - 	On record 211000 out of 226869, annotation 93% complete
2020-04-30 04:40:10 [main] ERROR org.springframework.batch.core.step.AbstractStep - Encountered an error executing step step in job annotationJob

[thomasyu888]

Adding java -Xmx48g resolved this issue.

[inodb]

Prolly good to add this to the documentation somewhere