application crashing with messages 'Resources lost(coredump)'

[FRESNEDA Christine]

Hello,
We have an application crashing at customer site with message:
Resources lost(coredump)

Could you tell me what it means, please?

Here are the proc system paremeters:
> proc:
> max_proc_per_user = 1024
> max_threads_per_user = 8192
> per_proc_stack_size = 33554432
> max_per_proc_stack_size = 33554432
> per_proc_data_size = 2147483648
> max_per_proc_data_size = 2147483648
> max_per_proc_address_space = 4294967296
> per_proc_address_space = 4294967296

The stack shows the following:

>0 0x3ff805c1af8 in __nxm_thread_kill(...) in /usr/shlib/libpthread.so
#1 0x3ff805ba150 in pthread_kill(...) in /usr/shlib/libpthread.so
#2 0x3ff80140464 in __tis_raise(...) in /usr/shlib/libc.so
#3 0x3ff801d25f8 in abort(...) in /usr/shlib/libc.so
#4 0x3ff81f34720 in terminate(...) in /usr/lib/cmplrs/cxx/libcxx.so
#5 0x3ff81f33bb4 in __cxx_throw_setup(...) in /usr/lib/cmplrs/cxx/libcxx.so
#6 0x3ff81f31118 in operator new(...) in /usr/lib/cmplrs/cxx/libcxx.so
#7 0x3ffbffe51c0 in append(...) in /usr/opt/temip/lib/libtls75-m.so
#8 0x3ffbffe61b0 in UnknownProcedure19FromFile143(...) in /usr/opt/temip/lib/libtls75-m.so
#9 0x3ffbffcc4f0 in insertIndex(...) in /usr/opt/temip/lib/libtls75-m.so #10 0x3ffbffe0ed4 in insert(...) in /usr/opt/temip/lib/libtls75-m.so
#11 0x3ffbffe2bc0 in add(...) in /usr/opt/temip/lib/libtls75-m.so
#12 0x3ffbffa4cb0 in recursiveSaveOn(...) in /usr/opt/temip/lib/libtls75-m.so
#13 0x3ffbffa3114 in saveObjToRWFile(...) in /usr/opt/temip/lib/libtls75-m.so
#14 0x3ffbffbdd30 in apply(...) in /usr/opt/temip/lib/libtls75-m.so
#15 0x3ffbffbeae0 in UnknownProcedure11FromFile98(...) in /usr/opt/temip/lib/libtls75-m.so
#16 0x3ffbffa2ce8 in saveGuts(...) in /usr/opt/temip/lib/libtls75-m.so
#17 0x3ffbffa4e40 in recursiveSaveOn(...) in /usr/opt/temip/lib/libtls75-m.so
#18 0x300001fab10 in saveGuts(...) in /usr/opt/temip/lib/libtfc_wp.so
#19 0x3ffbffa4e40 in recursiveSaveOn(...) in /usr/opt/temip/lib/libtls75-m.so #20 0x3ffbffa73d4 in operator <<(...) in /usr/opt/temip/lib/libtls75-m.so
#21 0x300001fb690 in SaveDictionaryContext(...) in /usr/opt/temip/lib/libtfc_wp.so
#22 0x120003894 in /usr/opt/temip/bin/tfc_cache_dict_builder
#23 0x120003248 in /usr/opt/temip/bin/tfc_cache_dict_builder

[Johan Brusche]

Looking at the stack this seems to be a TeMIP application, and probably not much expertise in that area is watching this forum.

TeMIP sits under the 'OpenView' umbrella, maybe posting in the management forum: http://forums1.itrc.hp.com/service/forums/familyhome.do?familyId=121 , migth trigger more response.

__ Johan./

[FRESNEDA Christine]

I was asking in fact for the meaning of "Resources lost".
Could you explain me what it means exactly?
How can I determine which resource I should investigate?
I know it is a TeMIP application I work in the 3 level TeMIP support group.
Christine

[Venkatesh BL]

Typically, core dump happens when the program tries to access a memory location that is wasn't supposed to.

'Resources' here, I think, referes to the program itself...

[Johan Brusche]

Christine,

This message can come from the ksh shell from which the program was launched. You might get a different message when running it from within the csh shell.

Potentially "UnknownProcedure19FromFile143" in the stack could be the real issue.

Under 'resources' one could understand, max per proc dataspace, max per proc threads, etc.
Are you running this as root ? If not, does it make a difference if you use "ulimit" ?

Rgds,
Christine.

[FRESNEDA Christine]

Yes I run this tool as root.
Christine

[Ann Majeske]

It looks like "Resources lost" is the message printed by ksh when the process receives an abort signal also known as SIGABRT. Apparently SIGLOST is an alternate name for this signal. Note that #3 in your stack is abort(), the routine that generates the SIGABRT signal.