Notifier dumps core repeatedly when unable to resolve SNMP host address

  • Article ID:100028408
  • Modified Date:
  • Product(s):


In the VCS notifier configuration, If the address used for the SNMP host is incorrect or otherwise un-resolvable to an IP address the notifier agent may dump core on every monitor cycle.

Error Message

Stack from notifier process core will look like below.

bash-3.2# mdb core_testhost_notifier_0_0_1351696573_18365
Loading modules: [ ]

> ::status
debugging core file of notifier (32-bit) from testhost
file: /opt/VRTSvcs/bin/notifier
initial argv: /opt/VRTSvcs/bin/notifier -v p=14141,l=30 -n p=14144 -s m=monpp5,l=Error,c=publ
threading model: multi-threaded
status: process terminated by SIGSEGV (Segmentation Fault)

> ::stack`freeaddrinfo+4(0, 55628, ff21fb3c, 11b060, 14a050, 0) 
__0FPsnmp_sess_closePv+0xac(18c770, 1, ff223a00, 0, 0, ff223a00)
__0FOsnmp_sess_openP6Msnmp_session+0x75c(feef86a4, 0, 0, 0, 0, 0)
__0FJsnmp_openP6Msnmp_session+0x28(feef86a4, 0, 4e, 0, 80808080, 1010101)
__0FOsnmp_trap_sendP6FVElemP6RSNMP_manager_type+0x2804(18d888, 182490, 183060, 0, 3e8, 0)
notifier_srv+0xe84(0, feefc000, 0, 0, 422f0, 0)`_lwp_start(0, 0, 0, 0, 0, 0)


VCS notifier process uses freeaddrinfo() function when trying to start a SNMP session. When there is a SNMP configuration issue, the function is called with NULL value which causes the notifier process to core dump. 

This issue occurs in clusters running on Solaris operating system with SNMP configuration issues.


Use an IP address or resolvable hostname in the  SNMP configuration to avoid this issue.

Note: This has been resolved on 5.1SP1RP4 and 6.1 and above patch levels. 

Was this content helpful?

Get Support