pmmgr mem leak or someone else's problem?

[Ken McDonell <kenj@xxxxxxxxxxxxxxxx>]

Triaging some QA failures and noticed this on vm14 x86_64 CentOS6.7 from qa/666

=== valgrind stdout === 
==6948== 132,000 bytes in 165 blocks are definitely lost in loss record 16 of 16
==6948==    at 0x4C28192: operator new[](unsigned long) 
(vg_replace_malloc.c:363)
==6948==    by 0x10FDC0: pmmgr_job_spec::compute_hostid(std::string const&) 
(pmmgr.cxx:249)
==6948==    by 0x1127B0: pmmgr_job_spec::poll() (pmmgr.cxx:427)
==6948==    by 0x114FE8: main (pmmgr.cxx:1461)
==6948==
{
   <insert_a_suppression_name_here>
   Memcheck:Leak
   fun:_Znam
   fun:_ZN14pmmgr_job_spec14compute_hostidERKSs
   fun:_ZN14pmmgr_job_spec4pollEv
   fun:main
}

Is this a new (?) regression in pmmgr, or something we can't avoid (and so 
should be added to the suppressions file)?

I don't recall seeing this before, and it looks like a fair chunk of memory.