Resend patch #1 with attached file.
On 08/11/16 18:47, minh chau wrote:
Resend patch #1
On 08/11/16 14:23, Minh Hon Chau wrote:
osaf/services/saf/amf/amfd/cluster.cc | 2 +-
osaf/services/saf/amf/amfd/sgproc.cc | 24 ++++++++++++++++++++----
2 files changed, 21 insertions(+), 5 deletions(-)
At the state of AVD_INIT_DONE and standby controller has middleware SUs
assigned, SC failover will leave transient assignments from headless
stage
unrecovered.
The problem is when standy controller becomes new active one, new AMFD
does not continue the headless recovery which is initiated under cluster
initiation timeout.
The patch continues the cluster initiation phase at new active AMFD
after
SC failover.
diff --git a/osaf/services/saf/amf/amfd/cluster.cc
b/osaf/services/saf/amf/amfd/cluster.cc
--- a/osaf/services/saf/amf/amfd/cluster.cc
+++ b/osaf/services/saf/amf/amfd/cluster.cc
@@ -58,7 +58,7 @@ void avd_cluster_tmr_init_evh(AVD_CL_CB
saflog(LOG_NOTICE, amfSvcUsrName, "Cluster startup timeout,
assigning SIs to SUs");
osafassert(evt->info.tmr.type == AVD_TMR_CL_INIT);
- LOG_NO("Cluster startup is done");
+ LOG_NO("Cluster initiated");
if (avd_cluster->saAmfClusterAdminState !=
SA_AMF_ADMIN_UNLOCKED) {
LOG_WA("Admin state of cluster is locked");
diff --git a/osaf/services/saf/amf/amfd/sgproc.cc
b/osaf/services/saf/amf/amfd/sgproc.cc
--- a/osaf/services/saf/amf/amfd/sgproc.cc
+++ b/osaf/services/saf/amf/amfd/sgproc.cc
@@ -1566,17 +1566,33 @@ void avd_su_si_assign_evh(AVD_CL_CB *cb,
}
done:
- /* Free the messages */
- avsv_dnd_msg_free(n2d_msg);
- evt->info.avnd_msg = nullptr;
-
if (su != nullptr) {
if (su->sg_of_su->sg_ncs_spec == false) {
if (su->sg_of_su->any_assignment_absent() == true) {
su->sg_of_su->failover_absent_assignment();
}
+ } else {
+ // Running at this code line, that means 2N-MW SU has
been assigned
+ // to ACTIVE (from STANDBY).
+ if (n2d_msg->msg_info.n2d_su_si_assign.msg_act ==
AVSV_SUSI_ACT_MOD &&
+ n2d_msg->msg_info.n2d_su_si_assign.ha_state ==
SA_AMF_HA_ACTIVE) {
+ // The cb->init_state must be AVD_INIT_DONE or
AVD_APP_STATE
+ // If AVD_INIT_DONE, there was a SC failover during
cluster
+ // instantiation phase in cluster (after all NCS SU
is assigned)
+ // If AVD_APP_STATE, this should be come from 2N-MW
SI swap
+ if (cb->init_state >= AVD_INIT_DONE) {
+ if (cluster_su_instantiation_done(cb, nullptr)
== true) {
+ cluster_startup_expiry_event_generate(cb);
+ } else {
+ m_AVD_CLINIT_TMR_START(cb);
+ }
+ }
+ }
}
}
+ /* Free the messages */
+ avsv_dnd_msg_free(n2d_msg);
+ evt->info.avnd_msg = nullptr;
TRACE_LEAVE();
}
AMFD: Fix SC failover during headless sync at INIT_DONE state [#2162]
At the state of AVD_INIT_DONE and standby controller has middleware SUs
assigned, SC failover will leave transient assignments from headless stage
unrecovered.
The problem is when standy controller becomes new active one, new AMFD
does not continue the headless recovery which is initiated under cluster
initiation timeout.
The patch continues the cluster initiation phase at new active AMFD after
SC failover.
diff --git a/osaf/services/saf/amf/amfd/cluster.cc b/osaf/services/saf/amf/amfd/cluster.cc
--- a/osaf/services/saf/amf/amfd/cluster.cc
+++ b/osaf/services/saf/amf/amfd/cluster.cc
@@ -58,7 +58,7 @@ void avd_cluster_tmr_init_evh(AVD_CL_CB
saflog(LOG_NOTICE, amfSvcUsrName, "Cluster startup timeout, assigning SIs to SUs");
osafassert(evt->info.tmr.type == AVD_TMR_CL_INIT);
- LOG_NO("Cluster startup is done");
+ LOG_NO("Cluster initiated");
if (avd_cluster->saAmfClusterAdminState != SA_AMF_ADMIN_UNLOCKED) {
LOG_WA("Admin state of cluster is locked");
diff --git a/osaf/services/saf/amf/amfd/sgproc.cc b/osaf/services/saf/amf/amfd/sgproc.cc
--- a/osaf/services/saf/amf/amfd/sgproc.cc
+++ b/osaf/services/saf/amf/amfd/sgproc.cc
@@ -1566,17 +1566,33 @@ void avd_su_si_assign_evh(AVD_CL_CB *cb,
}
done:
- /* Free the messages */
- avsv_dnd_msg_free(n2d_msg);
- evt->info.avnd_msg = nullptr;
-
if (su != nullptr) {
if (su->sg_of_su->sg_ncs_spec == false) {
if (su->sg_of_su->any_assignment_absent() == true) {
su->sg_of_su->failover_absent_assignment();
}
+ } else {
+ // Running at this code line, that means 2N-MW SU has been assigned
+ // to ACTIVE (from STANDBY).
+ if (n2d_msg->msg_info.n2d_su_si_assign.msg_act == AVSV_SUSI_ACT_MOD &&
+ n2d_msg->msg_info.n2d_su_si_assign.ha_state == SA_AMF_HA_ACTIVE) {
+ // The cb->init_state must be AVD_INIT_DONE or AVD_APP_STATE
+ // If AVD_INIT_DONE, there was a SC failover during cluster
+ // instantiation phase in cluster (after all NCS SU is assigned)
+ // If AVD_APP_STATE, this should be come from 2N-MW SI swap
+ if (cb->init_state >= AVD_INIT_DONE) {
+ if (cluster_su_instantiation_done(cb, nullptr) == true) {
+ cluster_startup_expiry_event_generate(cb);
+ } else {
+ m_AVD_CLINIT_TMR_START(cb);
+ }
+ }
+ }
}
}
+ /* Free the messages */
+ avsv_dnd_msg_free(n2d_msg);
+ evt->info.avnd_msg = nullptr;
TRACE_LEAVE();
}
------------------------------------------------------------------------------
Developer Access Program for Intel Xeon Phi Processors
Access to Intel Xeon Phi processor-based developer platforms.
With one year of Intel Parallel Studio XE.
Training and support from Colfax.
Order your platform today. http://sdm.link/xeonphi
_______________________________________________
Opensaf-devel mailing list
Opensaf-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/opensaf-devel
