Hello,
This issue got resolved & this is how. They ran out of Memory !!!
May 6 02:33:42 nysmtpbb3 tmpfs: [ID 518458 kern.warning] WARNING: /tmp:
File system full, swap space limit exceeded
May 6 02:33:46 nysmtpbb3 genunix: [ID 470503 kern.warning] WARNING:
Sorry, no swap space to grow stack for pid 5991 (centasvc_file_up)
May 6 02:33:51 nysmtpbb3 genunix: [ID 603404 kern.notice] NOTICE:
core_log: centasvc_file_up[5991] core dumped: /bb/cores/core.centasvc_file
_up.5991.1210055627
Rod Evans had provided his input earlier & actually pointed that
customer should be looking for system messages as below
> $c
ld.so.1`_elf_lookup_filtee+8(ffbfe098, ffbfe128, ffbfe124, 1, fba16764,
80000)
ld.so.1`elf_lookup_filtee+0x90(ffbfe098, ffbfe128, ffbfe124, 1, 0, 36654)
which is a result of libnsl.so.1 looking up "open" and presently inspecting
libpthread, which is a filter on libc.
The instruction that causes the core seems to be:
ld.so.1`_elf_lookup_filtee: save %sp, -0xc0, %sp
ld.so.1`_elf_lookup_filtee+4: mov %o7, %i4
ld.so.1`_elf_lookup_filtee+8: st %i2, [%sp + 0x6c]
^^^^^^^^^^^^^^^^^^^^^^^^^^^
which is writing to the stack:
FFBFC000 16K rw--- [ stack ]
> <sp=X
ffbfde58
> ffbfde58+6c/XXXX
0xffbfdec4: 0 0 0 0
There's no reason this should have dumped core as far as I know.
Do you have more core files? Are they all the same? Were there any
system console messages that could be associated to the same time that
the core was produced?
Thanks,
Nishath
Nishath Fathima
TSE- Java/Compilers
781-442-1025
Working Hrs
6am - 3pm est
Nishath Fathima wrote:
> Hello,
> I am troubleshooting a Bloomberg case & am looking for help with
> resolving this issue as I could not find anything in sunsolve.
> This is a C++/Fortran application which crashes at runtime while
> looking up a elf symbol on a solaris 10 box.
> This has only occurred twice, or so, against 2 different tasks. The
> walkback for each looked similar. The task from the corefile runs
> 600,000 to 700,000 times daily, and that it ran successfully both
> before and after this crash occurred. There were no changes to the
> system.
> I have attached the pstack, pmap & pldd output. Additional information
> can be found under /net/cores.central//cores/65911320 ( if you have
> access to )
> Any pointers is areally appreciated.
>
> Please email me directly as I am not in this alias
> Best Regards,
> Nishath
>
