REQUIRED for all errors and issues:
a) GATK version used:  v4.2.6.1
b) Exact command used:

$GATK --java-options "-Xmx98G -XX:-UseGCOverheadLimit -XX:+UseParallelGC -XX:ParallelGCThreads=18"  ModelSegments \
--allelic-counts T/chr2_T.allelicCounts.tsv \
--denoised-copy-ratios OUTPUT_GATK_CALLER/chr2_T.denoisedCR.tsv \
--output-prefix chr2_T \
--genotyping-homozygous-log-ratio-threshold -15 \
--number-of-changepoints-penalty-factor 0.1 \
--kernel-variance-allele-fraction 0.025 \
--minimum-total-allele-count-normal 5 \
--kernel-variance-copy-ratio 0 \
--window-size 256 \
--tmp-dir $TEMP \
--number-of-smoothing-iterations-per-fit 2 \
--normal-allelic-counts N/chr2_N.allelicCounts.tsv \
-O OUTPUT_GATK_CALLER

c) Entire program log: 08:25:04.687 INFO  NativeLibraryLoader - Loading libgkl_compression.so from jar:file:/save/user/gatk-4.2.6.1/gatk-package-4.2.6.1-local.jar!/com/intel/gkl/native/libgkl_compression.so
08:25:04.810 INFO  ModelSegments - ------------------------------------------------------------
08:25:04.810 INFO  ModelSegments - The Genome Analysis Toolkit (GATK) v4.2.6.1
08:25:04.811 INFO  ModelSegments - For support and documentation go to https://software.broadinstitute.org/gatk/
08:25:04.811 INFO  ModelSegments - Executing as sassou@n016 on Linux v4.18.0-553.89.1.el8_10.x86_64 amd64
08:25:04.811 INFO  ModelSegments - Java runtime: OpenJDK 64-Bit Server VM v1.8.0_472-b08
08:25:04.811 INFO  ModelSegments - Start Date/Time: March 10, 2026 8:25:04 AM CET
08:25:04.811 INFO  ModelSegments - ------------------------------------------------------------
08:25:04.811 INFO  ModelSegments - ------------------------------------------------------------
08:25:04.811 INFO  ModelSegments - HTSJDK Version: 2.24.1
08:25:04.812 INFO  ModelSegments - Picard Version: 2.27.1
08:25:04.812 INFO  ModelSegments - Built for Spark Version: 2.4.5
08:25:04.812 INFO  ModelSegments - HTSJDK Defaults.COMPRESSION_LEVEL : 2
08:25:04.812 INFO  ModelSegments - HTSJDK Defaults.USE_ASYNC_IO_READ_FOR_SAMTOOLS : false
08:25:04.812 INFO  ModelSegments - HTSJDK Defaults.USE_ASYNC_IO_WRITE_FOR_SAMTOOLS : true
08:25:04.812 INFO  ModelSegments - HTSJDK Defaults.USE_ASYNC_IO_WRITE_FOR_TRIBBLE : false
08:25:04.812 INFO  ModelSegments - Deflater: IntelDeflater
08:25:04.812 INFO  ModelSegments - Inflater: IntelInflater
08:25:04.812 INFO  ModelSegments - GCS max retries/reopens: 20
08:25:04.812 INFO  ModelSegments - Requester pays: disabled
08:25:04.812 INFO  ModelSegments - Initializing engine
08:25:04.812 INFO  ModelSegments - Done initializing engine
08:25:04.812 INFO  ModelSegments - Used memory (MB) after initializing engine: 218
08:25:04.826 INFO  ModelSegments - Reading file (GATK_CNV_CALLER/chr2_T.denoisedCR.tsv)...
08:25:05.269 INFO  ModelSegments - Reading file (GATK_CNV_CALLER/2/chr2_T.allelicCounts.tsv)...
08:41:06.942 INFO  ModelSegments - Reading file (GATK_CNV_CALLER/N/chr2_N.allelicCounts.tsv)...
Using GATK jar /save/user/gatk-4.2.6.1/gatk-package-4.2.6.1-local.jar
Running:
    java -Dsamjdk.use_async_io_read_samtools=false -Dsamjdk.use_async_io_write_samtools=true -Dsamjdk.use_async_io_write_tribble=false -Dsamjdk.compression_level=2 -Xmx98G -XX:-UseGCOverheadLimit -XX:+UseParallelGC -XX:ParallelGCThreads=
18 -jar /save/user/gatk-4.2.6.1/gatk-package-4.2.6.1-local.jar ModelSegments --allelic-counts GATK_CNV_CALLER/T/chr2_T.allelicCounts.tsv --denoised-copy-ratios GATK_CNV_CALLER/chr2_T.denoisedCR.tsv --output-prefix chr2_T --genotyp
ing-homozygous-log-ratio-threshold -15 --number-of-changepoints-penalty-factor 0.1 --kernel-variance-allele-fraction 0.025 --minimum-total-allele-count-normal 5 --kernel-variance-copy-ratio 0 --window-size 256 --tmp-dir /home/work/NGS/WGS/tmp --number-of-smoothing-iterations-per-fit 2 --normal-allelic-counts GATK_CNV_CALLER/N/chr2_N.allelicCounts.tsv -O GATK_CNV_CALLER

[2026-03-10T08:55:44.120] error: Detected 1 oom_kill event in StepId=32551182.batch. Some of the step tasks have been OOM Killed.

size of the input files:

chr2_N.allelicCounts.tsv --> 5.3G
chr2_T.allelicCounts.tsv --> 5.3G
chr2_T.denoisedCR.tsv --> 7.4M

command crash due to out of memory call.

So how to avoid this type of failure ? I work on WGS CNV analysis with big input files, so i have splitted the process to work only per chromosome.

thank you --