Observing amfnd crash if a PI comp fails in CSI SET callback during fresh 
Sep 22 17:48:10 SC-1 root: INSTANTIATED : 
safComp=AmfDemo1,safSu=SU2,safSg=AmfDemo,safApp=AmfDemo1, PID :30566
Sep 22 17:48:10 SC-1 osafamfnd[29154]: clc.cc:1170: avnd_comp_clc_st_chng_prc: 
Assertion 'csi' failed.
Sep 22 17:48:10 SC-1 osafclmd[29132]: AL AMF Node Director is down, terminate 
this process
Sep 22 17:48:10 SC-1 osafclmd[29132]: exiting for shutdown

(gdb) bt
\#0  0x00007fe6546d6cc9 in __GI_raise (sig=sig@entry=6) at 
\#1  0x00007fe6546da0d8 in __GI_abort () at abort.c:89
\#2  0x00007fe655835b2e in __osafassert_fail (__file=<optimized out>, 
__line=<optimized out>,
    __func=<optimized out>, __assertion=<optimized out>) at sysf_def.c:281
\#3  0x000000000040cf89 in avnd_comp_clc_st_chng_prc (cb=cb@entry=0x666940 
<_avnd_cb>, comp=comp@entry=0x129a930,
    at clc.cc:1170
\#4  0x000000000040f590 in avnd_comp_clc_fsm_run (cb=cb@entry=0x666940 
<_avnd_cb>, comp=comp@entry=0x129a930,
    ev=AVND_COMP_CLC_PRES_FSM_EV_TERM_SUCC) at clc.cc:906
\#5  0x000000000040fe4a in avnd_evt_clc_resp_evh (cb=0x666940 <_avnd_cb>, 
evt=0x7fe63c0008c0) at clc.cc:414
\#6  0x000000000042651f in avnd_evt_process (evt=0x7fe63c0008c0) at main.cc:626
\#7  avnd_main_process () at main.cc:577
\#8  0x0000000000405a73 in main (argc=2, argv=0x7ffdfdd98368) at main.cc:202


** [tickets:#2047] amf: SG unstable when NPI comp in PI SU moves to TERM_FAILED 
state during fresh assignments.**

**Status:** assigned
**Milestone:** 5.0.2
**Created:** Mon Sep 19, 2016 11:31 AM UTC by Praveen
**Last Updated:** Tue Sep 20, 2016 05:50 PM UTC
**Owner:** Praveen

 (21.9 kB; application/x-compressed)

Conf: 2N model, one NPI and one PI comp in SU.
Steps to reproduce:
1)Add application using immcfg command.
2)Lock SG.
3)Unlock-in and unlock SUs.
4)Make provisions so that instantiation and clean up scripts of NPI comp 
returns with non-zero status.
5)Unlock SG.
When SG is unlocked, AMFND initiates active assignments by issuing callback to 
PI comp and by instantiating NPI component. After instantiation failure of NPI 
comp, AMFND tries to clean up it. Cleanup fails. AMFND marks comp and SU in 
TERM_FAILED state and terminates PI comp also, but AMFND neither responds to 
AMFD for the completion of assignment nor it sends any recovery request. 
Because of this SG remains unstable in REALIGN state.In this state, no admin 
operation is allowed.
Attached are traces and configuration.
Even though issue seems to be similar to #538, it is different in one aspect. 
In #538, SU moves to TERM_FAILED state and there is possibiltiy of 
failover/switchover as standby assignments are present.
In the present case, it happened during initial assignments and thus there is 
no standby to switchover/failover to.


Sent from sourceforge.net because opensaf-tickets@lists.sourceforge.net is 
subscribed to https://sourceforge.net/p/opensaf/tickets/

To unsubscribe from further messages, a project admin can change settings at 
https://sourceforge.net/p/opensaf/admin/tickets/options.  Or, if this is a 
mailing list, you can unsubscribe from the mailing list.
Opensaf-tickets mailing list

Reply via email to