Single-ended recovery mode

Healthcare & Life Sciences Parabricks
1

when i use fq2bam_meth, some of my fastq file can’t align using parabrick, they are in Single-ended recovery mode, please help me

2

[PB Debug 2024-Nov-19 00:02:29][src/main.cpp:238] Read in 1 fq files
[PB Info 2024-Nov-19 00:02:29] Mode = single-ended-gpu
[PB Info 2024-Nov-19 00:02:29] Running with 1 GPU(s), using 2 stream(s) per device with 16 worker threads per GPU
[PB Info 2024-Nov-19 00:02:39] # 0 0 0 0 bases/GPU/minute: 0.0
[PB Debug 2024-Nov-19 00:02:40][src/internal/bwa_lib_context.cu:214] handle is created and associated with GPU ID: 0
[PB Debug 2024-Nov-19 00:02:40][src/internal/bwa_lib_context.cu:214] handle is created and associated with GPU ID: 0
[PB Debug 2024-Nov-19 00:02:46][src/internal/seed_gpu.cu:522] Using new-er seeding implementation
[PB Debug 2024-Nov-19 00:02:46][src/internal/seed_gpu.cu:535] Phase 1: allocatedSize %d3290643072
[PB Debug 2024-Nov-19 00:02:46][src/internal/chain_gpu.cu:41] Phase 2 allocated size 2455466512
[PB Debug 2024-Nov-19 00:02:46][src/internal/seed_gpu.cu:522] Using new-er seeding implementation
[PB Debug 2024-Nov-19 00:02:46][src/internal/reg_gpu.cu:119] Partial %d800262144
[PB Debug 2024-Nov-19 00:02:46][src/internal/seed_gpu.cu:535] Phase 1: allocatedSize %d3290643072
[PB Debug 2024-Nov-19 00:02:46][src/internal/chain_gpu.cu:41] Phase 2 allocated size 2455466512
[PB Debug 2024-Nov-19 00:02:46][src/internal/reg_gpu.cu:119] Partial %d800262144
[PB Debug 2024-Nov-19 00:02:46][src/internal/reg_gpu.cu:134] Left/Right %d2720525056
[PB Debug 2024-Nov-19 00:02:46][src/internal/reg_gpu.cu:134] Left/Right %d2720525056
[PB Debug 2024-Nov-19 00:02:46][src/internal/reg_gpu.cu:138] RMAX %d3040525056
[PB Debug 2024-Nov-19 00:02:46][src/internal/reg_gpu.cu:138] RMAX %d3040525056
[PB Debug 2024-Nov-19 00:02:46][src/internal/reg_gpu.cu:145] Buf %d3706370816
[PB Debug 2024-Nov-19 00:02:46][src/internal/reg_gpu.cu:145] Buf %d3706370816
[PB Debug 2024-Nov-19 00:02:46][src/internal/reg_gpu.cu:173] Phase 3 allocatedSize %d3989748544[PB Debug 2024-Nov-19 00:02:46][src/internal/reg_gpu.cu:173] Phase 3 allocatedSize %d3989748544

[PB Debug 2024-Nov-19 00:02:46][src/internal/seed_gpu.cu:522] Using new-er seeding implementation
[PB Debug 2024-Nov-19 00:02:46][src/internal/seed_gpu.cu:535] Phase 1: allocatedSize %d3290643072
[PB Debug 2024-Nov-19 00:02:46][src/internal/seed_gpu.cu:522] Using new-er seeding implementation
[PB Debug 2024-Nov-19 00:02:46][src/internal/seed_gpu.cu:535] Phase 1: allocatedSize %d3290643072
[PB Debug 2024-Nov-19 00:02:46][src/internal/chain_gpu.cu:41] [PB Debug 2024-Nov-19 00:02:46][src/internal/chain_gpu.cu:41] Phase 2 allocated size 2455466512
Phase 2 allocated size 2455466512
[PB Debug 2024-Nov-19 00:02:46][src/internal/reg_gpu.cu:119] Partial %d800262144
[PB Debug 2024-Nov-19 00:02:46][src/internal/reg_gpu.cu:134] Left/Right %d2720525056
[PB Debug 2024-Nov-19 00:02:46][src/internal/reg_gpu.cu:138] RMAX %d3040525056
[PB Debug 2024-Nov-19 00:02:46][src/internal/reg_gpu.cu:145] Buf %d3706370816
[PB Debug 2024-Nov-19 00:02:46][src/internal/reg_gpu.cu:173] Phase 3 allocatedSize %d3989748544
[PB Debug 2024-Nov-19 00:02:46][src/internal/reg_gpu.cu:119] Partial %d800262144
[PB Debug 2024-Nov-19 00:02:46][src/internal/reg_gpu.cu:134] Left/Right %d2720525056
[PB Debug 2024-Nov-19 00:02:46][src/internal/reg_gpu.cu:138] RMAX %d3040525056
[PB Debug 2024-Nov-19 00:02:46][src/internal/reg_gpu.cu:145] Buf %d3706370816
[PB Debug 2024-Nov-19 00:02:46][src/internal/reg_gpu.cu:173] Phase 3 allocatedSize %d3989748544
[PB Debug 2024-Nov-19 00:02:46][src/internal/pb_fmindex.cu:1347] Did not pass one of the seeding asserts on device: returning from GPU kernels early.
[PB Info 2024-Nov-19 00:02:46] Single-ended recovery mode for batch with 0 reads before itself
[PB Debug 2024-Nov-19 00:02:47][src/internal/pb_fmindex.cu:1347] Did not pass one of the seeding asserts on device: returning from GPU kernels early.
[PB Info 2024-Nov-19 00:02:47] Single-ended recovery mode for batch with 131072 reads before itself
[PB Debug 2024-Nov-19 00:02:47][src/internal/pb_fmindex.cu:1347] Did not pass one of the seeding asserts on device: returning from GPU kernels early.
[PB Info 2024-Nov-19 00:02:47] Single-ended recovery mode for batch with 65536 reads before itself
[PB Debug 2024-Nov-19 00:02:47][src/internal/pb_fmindex.cu:1347] Did not pass one of the seeding asserts on device: returning from GPU kernels early.
[PB Info 2024-Nov-19 00:02:47] Single-ended recovery mode for batch with 262144 reads before itself
[PB Debug 2024-Nov-19 00:02:47][src/internal/pb_fmindex.cu:1347] Did not pass one of the seeding asserts on device: returning from GPU kernels early.
[PB Info 2024-Nov-19 00:02:47] Single-ended recovery mode for batch with 196608 reads before itself
[PB Debug 2024-Nov-19 00:02:47][src/internal/pb_fmindex.cu:1347] Did not pass one of the seeding asserts on device: returning from GPU kernels early.
[PB Debug 2024-Nov-19 00:02:47][src/internal/pb_fmindex.cu:1347] Did not pass one of the seeding asserts on device: returning from GPU kernels early.
[PB Info 2024-Nov-19 00:02:47] [PB Info 2024-Nov-19 00:02:47] Single-ended recovery mode for batch with 327680 reads before itselfSingle-ended recovery mode for batch with 393216 reads before itself

[PB Debug 2024-Nov-19 00:02:48][src/internal/pb_fmindex.cu:1347] Did not pass one of the seeding asserts on device: returning from GPU kernels early.
[PB Info 2024-Nov-19 00:02:48] Single-ended recovery mode for batch with 524288 reads before itself
[PB Debug 2024-Nov-19 00:02:48][src/internal/pb_fmindex.cu:1347] Did not pass one of the seeding asserts on device: returning from GPU kernels early.
[PB Info 2024-Nov-19 00:02:48] Single-ended recovery mode for batch with 458752 reads before itself
[PB Debug 2024-Nov-19 00:02:48][src/internal/pb_fmindex.cu:1347] Did not pass one of the seeding asserts on device: returning from GPU kernels early.
[PB Info 2024-Nov-19 00:02:48] Single-ended recovery mode for batch with 589824 reads before itself
[PB Debug 2024-Nov-19 00:02:48][src/internal/pb_fmindex.cu:1347] Did not pass one of the seeding asserts on device: returning from GPU kernels early.
[PB Info 2024-Nov-19 00:02:48] Single-ended recovery mode for batch with 655360 reads before itself
[PB Debug 2024-Nov-19 00:02:48][src/internal/pb_fmindex.cu:1347] Did not pass one of the seeding asserts on device: returning from GPU kernels early.
[PB Info 2024-Nov-19 00:02:48] Single-ended recovery mode for batch with 720896 reads before itself
[PB Debug 2024-Nov-19 00:02:48][src/internal/pb_fmindex.cu:1347] Did not pass one of the seeding asserts on device: returning from GPU kernels early.
[PB Info 2024-Nov-19 00:02:48] Single-ended recovery mode for batch with 786432 reads before itself
[PB Debug 2024-Nov-19 00:02:49][src/internal/pb_fmindex.cu:1347] Did not pass one of the seeding asserts on device: returning from GPU kernels early.
[PB Info 2024-Nov-19 00:02:49] Single-ended recovery mode for batch with 851968 reads before itself
[PB Debug 2024-Nov-19 00:02:49][src/internal/pb_fmindex.cu:1347] Did not pass one of the seeding asserts on device: returning from GPU kernels early.
[PB Info 2024-Nov-19 00:02:49] Single-ended recovery mode for batch with 917504 reads before itself
[PB Info 2024-Nov-19 00:02:49] # 9 0 0 0 bases/GPU/minute: 0.0
[PB Debug 2024-Nov-19 00:02:49][src/internal/pb_fmindex.cu:1347] Did not pass one of the seeding asserts on device: returning from GPU kernels early.
[PB Info 2024-Nov-19 00:02:49] Single-ended recovery mode for batch with 983040 reads before itself
[PB Info 2024-Nov-19 00:02:59] # 10 10 0 0 bases/GPU/minute: 0.0
[PB Info 2024-Nov-19 00:03:09] # 10 10 0 0 bases/GPU/minute: 0.0
[PB Info 2024-Nov-19 00:03:19] # 10 11 0 19182729 bases/GPU/minute: 115096374.0
[PB Info 2024-Nov-19 00:03:29] # 10 11 0 19182729 bases/GPU/minute: 0.0
[PB Info 2024-Nov-19 00:03:39] # 10 11 0 19182729 bases/GPU/minute: 0.0
[PB Info 2024-Nov-19 00:03:49] # 10 10 0 38346594 bases/GPU/minute: 114983190.0

3

when i input a small fastq file, the output has changed:

image
image788×578 113 KB

what means of “chain”? How to avoid this problem？

4

Hi @wangyanni541,

Recovery mode happens because some workload cannot be handled on GPU so we bring them back to CPU to process. But the output file still stays accurate and there should be no worries on that.

Thanks!