osaf/services/saf/amf/amfnd/err.cc | 80 +++++++++++++++++++-------------------
1 files changed, 40 insertions(+), 40 deletions(-)
If any error escalates to component/su failover during headless, amfnd
reboot node.
The issue is other healthy SUs get affected by this reboot, and this degrades
the availability characteristic that AMF supports.
The patch allow component/su failover during headless, but supports it
partially (mark comp/su as failed) since failover to another comp/su
requires amfd's presence.
diff --git a/osaf/services/saf/amf/amfnd/err.cc
b/osaf/services/saf/amf/amfnd/err.cc
--- a/osaf/services/saf/amf/amfnd/err.cc
+++ b/osaf/services/saf/amf/amfnd/err.cc
@@ -77,8 +77,6 @@ static uint32_t avnd_err_restart_esc_lev
static uint32_t avnd_err_restart_esc_level_1(AVND_CB *, AVND_SU *,
AVND_ERR_ESC_LEVEL *, AVSV_ERR_RCVR *);
static uint32_t avnd_err_restart_esc_level_2(AVND_CB *, AVND_SU *,
AVND_ERR_ESC_LEVEL *, AVSV_ERR_RCVR *);
-static void cleanup_all_comps_and_reboot(AVND_CB *cb);
-
/* LSB Changes. Strings to represent source of component Error */
static const char *g_comp_err[] = {
@@ -802,9 +800,23 @@ uint32_t avnd_err_rcvr_comp_failover(AVN
rc = avnd_di_oper_send(cb, failed_comp->su,
AVSV_ERR_RCVR_SU_FAILOVER);
// if headless, we have to perform the 'failover' without amfd
+ // for now, just terminate all components in the SU
if (cb->is_avd_down == true) {
- // SU failover results in a node failfast if headless
(not nice)
- cleanup_all_comps_and_reboot(cb);
+ AVND_COMP *comp;
+
+ LOG_NO("Terminating components of '%s'(abruptly &
unordered)",su->name.value);
+ for (comp =
m_AVND_COMP_FROM_SU_DLL_NODE_GET(m_NCS_DBLIST_FIND_FIRST(&su->comp_list));
+ comp;
+ comp =
m_AVND_COMP_FROM_SU_DLL_NODE_GET(m_NCS_DBLIST_FIND_NEXT(&comp->su_dll_node))) {
+ if (comp->su->su_is_external)
+ continue;
+
+ rc = avnd_comp_clc_fsm_run(cb, comp,
AVND_COMP_CLC_PRES_FSM_EV_CLEANUP);
+ if (NCSCC_RC_SUCCESS != rc) {
+ LOG_ER("'%s' termination failed",
comp->name.value);
+ goto done;
+ }
+ }
}
}
@@ -854,11 +866,7 @@ uint32_t avnd_err_rcvr_su_failover(AVND_
}
avnd_su_pres_state_set(cb, comp->su,
SA_AMF_PRESENCE_TERMINATING);
}
-
done:
- if (cb->is_avd_down == true) {
- cleanup_all_comps_and_reboot(cb);
- }
TRACE_LEAVE2("%u", rc);
return rc;
@@ -962,7 +970,30 @@ done:
// TODO - try to see if we can avoid a reboot & terminate components
more gracefully
// if headless, reboot as we can't perform a switchover without amfd
if (cb->is_avd_down == true) {
- cleanup_all_comps_and_reboot(cb);
+ /* Unordered cleanup of all local application components */
+ for (comp = (AVND_COMP *)ncs_patricia_tree_getnext(&cb->compdb,
(uint8_t *)nullptr);
+ comp != nullptr;
+ comp = (AVND_COMP *)
ncs_patricia_tree_getnext(&cb->compdb, (uint8_t *)&comp->name)) {
+
+ if (comp->su->is_ncs || comp->su->su_is_external)
+ continue;
+
+ rc = avnd_comp_clc_fsm_run(cb, comp,
AVND_COMP_CLC_PRES_FSM_EV_CLEANUP);
+ if (rc != NCSCC_RC_SUCCESS) {
+ LOG_ER("'%s' termination failed",
comp->name.value);
+ opensaf_reboot(avnd_cb->node_info.nodeId,
+ (char
*)avnd_cb->node_info.executionEnvironment.value,
+ "Component
termination failed at node switchover");
+ LOG_ER("Exiting (due to comp term failed) to
aid fast node reboot");
+ exit(1);
+ }
+ }
+
+ opensaf_reboot(avnd_cb->node_info.nodeId,
+ (char *)avnd_cb->node_info.executionEnvironment.value,
+ "Can't perform node switchover while controllers are
down. Recovery is node failfast.");
+ LOG_ER("Exiting to aid fast node reboot");
+ exit(1);
}
TRACE_LEAVE2("%u", rc);
@@ -1598,34 +1629,3 @@ bool is_no_assignment_due_to_escalations
TRACE_LEAVE2("false");
return false;
}
-
-void cleanup_all_comps_and_reboot(AVND_CB *cb)
-{
- AVND_COMP *comp;
- uint32_t rc = NCSCC_RC_SUCCESS;
-
- /* Unordered cleanup of all local application components */
- for (comp = (AVND_COMP *)ncs_patricia_tree_getnext(&cb->compdb,
(uint8_t *)nullptr);
- comp != nullptr;
- comp = (AVND_COMP *) ncs_patricia_tree_getnext(&cb->compdb,
(uint8_t *)&comp->name)) {
-
- if (comp->su->is_ncs || comp->su->su_is_external)
- continue;
-
- rc = avnd_comp_clc_fsm_run(cb, comp,
AVND_COMP_CLC_PRES_FSM_EV_CLEANUP);
- if (rc != NCSCC_RC_SUCCESS) {
- LOG_ER("'%s' termination failed", comp->name.value);
- opensaf_reboot(avnd_cb->node_info.nodeId,
- (char
*)avnd_cb->node_info.executionEnvironment.value,
- "Component termination
failed at node switchover");
- LOG_ER("Exiting (due to comp term failed) to aid fast
node reboot");
- exit(1);
- }
- }
-
- opensaf_reboot(avnd_cb->node_info.nodeId,
- (char *)avnd_cb->node_info.executionEnvironment.value,
- "Can't perform recovery while controllers are down. Recovery is
node failfast.");
- LOG_ER("Exiting to aid fast node reboot");
- exit(1);
-}
------------------------------------------------------------------------------
Site24x7 APM Insight: Get Deep Visibility into Application Performance
APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month
Monitor end-to-end web transactions and take corrective actions now
Troubleshoot faster and improve end-user experience. Signup Now!
http://pubads.g.doubleclick.net/gampad/clk?id=272487151&iu=/4140
_______________________________________________
Opensaf-devel mailing list
Opensaf-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/opensaf-devel
