Problem when processing big batches of images (512*512)

[OBrink]

• @Kohulan bumped into this problem when generating large amounts of 512*512 images.
• We have recently produced a dataset of > 500,000,000 299*299 images with markush structures without big problems. We had a problem with dying threads where we did not really find the root. It was "solved" by regularly checking for dead processes and terminating them. It might not have been the most elegant workaround, but we got our training data.
• During the generation of large amounts of 512*512 images, the threads simply die and everything stops (correct me, if I am wrong here, @Kohulan)
• @JonasSchaub: I'm just tagging you because I suspect that you might have seen a lot more problems with the Java Runtime Environment than Kohulan and me. When you look at the error below, does this tell you anything?

#
# A fatal error has been detected by the Java Runtime Environment:
#
#  SIGSEGV (0xb) at pc=0x00002ad088d08a84, pid=252544, tid=0x00002ad017f80ec0
#
# JRE version: OpenJDK Runtime Environment (8.0_191-b12) (build 1.8.0_191-b12)
# Java VM: OpenJDK 64-Bit Server VM (25.191-b12 mixed mode linux-amd64 compressed oops)
# Problematic frame:
# C  [_jpype.cpython-38-x86_64-linux-gnu.so+0x3ea84]  JPClass::invoke(JPJavaFrame&, _jobject*, _jclass*, _jmethodID*, jvalue*)+0x74
#
# Failed to write core dump. Core dumps have been disabled. To enable core dumping, try "ulimit -c unlimited" before starting Java again
#
# If you would like to submit a bug report, please visit:
#   http://bugreport.java.com/bugreport/crash.jsp
# The crash happened outside the Java Virtual Machine in native code.
# See problematic frame for where to report the bug.
#

---------------  T H R E A D  ---------------

Current thread (0x00005601104c6800):  JavaThread "main" [_thread_in_native, id=252544, stack(0x00007fff9291e000,0x00007fff92a1e000)]

siginfo: si_signo: 11 (SIGSEGV), si_code: 1 (SEGV_MAPERR), si_addr: 0x0000000000000000

Registers:
RAX=0x0000000000000000, RBX=0x00007fff92a19c10, RCX=0x0000000000000000, RDX=0x00002ad08a5c06b0
RSP=0x00007fff92a19a70, RBP=0x00007fff92a19c20, RSI=0x0000000000000000, RDI=0x00005601104c6800
R8 =0x00002ad0a288b2f0, R9 =0x0000000000000000, R10=0x0000000000000000, R11=0x0000000000000000
R12=0x00007fff92a19a90, R13=0x000056011446ba30, R14=0x0000000000000000, R15=0x00005601149ffcc0
RIP=0x00002ad088d08a84, EFLAGS=0x0000000000010206, CSGSFS=0x0000000000000033, ERR=0x0000000000000004
  TRAPNO=0x000000000000000e

flags		: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc aperfmperf eagerfpu pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch epb cat_l3 cdp_l3 intel_ppin intel_pt ssbd mba ibrs ibpb stibp tpr_shadow vnmi flexpriority ept vpid fsgsbase tsc_adjust bmi1 hle avx2 smep bmi2 erms invpcid rtm cqm mpx rdt_a avx512f avx512dq rdseed adx smap clflushopt clwb avx512cd avx512bw avx512vl xsaveopt xsavec xgetbv1 cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local dtherm ida arat pln pts hwp hwp_act_window hwp_epp hwp_pkg_req pku ospke spec_ctrl intel_stibp flush_l1d
bogomips	: 4605.66
clflush size	: 64
cache_alignment	: 64
address sizes	: 46 bits physical, 48 bits virtual
power management:



Memory: 4k page, physical 196687540k(129534532k free), swap 12582908k(12580860k free)

vm_info: OpenJDK 64-Bit Server VM (25.191-b12) for linux-amd64 JRE (1.8.0_191-b12), built on Nov 19 2018 16:07:16 by "mockbuild" with gcc 4.8.5 20150623 (Red Hat 4.8.5-36)

time: Fri Sep 30 06:40:12 2022
elapsed time: 1794 seconds (0d 0h 29m 54s)

[Kohulan]

The complete log could be found here:
hs_err_pid25415.log

[JonasSchaub]

All I can tell you about this (which you might already know) is that a segmentation fault error (SIGSEGV, see line 2) indicates a memory problem, see e.g. here: https://stackoverflow.com/questions/1564372/what-causes-a-sigsegv
Could it simply be that you are running out of memory?

[JonasSchaub]

Can you try running it with more heap space allocated to the JRE? Suggestion by @FelixBaensch

[OBrink]

Will look into it, thank you for the suggestion!

[OBrink]

Possibly related:
jpype-project/jpype#81

[Kohulan]

We should try running it by increasing the heap space? Maybe we can set it using jpype?

[OBrink]

@Kohulan - I have set the JVM heap size to 4 GB. When starting the next big batch training data generation, we will see if this fixes it.