Migrated from http://devel.opensaf.org/ticket/2834.
Changeset : 3784 4.2.2 RC1
Found on SLES 32 Bit Setup
2 Controllers and 2 Payloads
Model - 2n
SUs are hosted on payloads.
Scenario:
Created Dependency like SI1 as sponsor and SI2 as dependent with tolerance time as 60 sec.
Lock SI1
SI1 has been unassigned.
Waited for 56 sec
Did "pkill osafamfd" on Active Controller and Active controller went for reboot
Controller node joined successfully after reboot.
SI2 unassignment never happened.
SI assignment status is showing as:
safSi=SI1,safApp=App1
saAmfSIAdminState=LOCKED(2)
saAmfSIAssignmentState=UNASSIGNED(1)
safSi=SI2,safApp=App1
saAmfSIAdminState=UNLOCKED(1)
saAmfSIAssignmentState=FULLY_ASSIGNED(2)
safSi=SI3,safApp=App1
saAmfSIAdminState=UNLOCKED(1)
saAmfSIAssignmentState=FULLY_ASSIGNED(2)
safSi=SI4,safApp=App1
saAmfSIAdminState=UNLOCKED(1)
saAmfSIAssignmentState=FULLY_ASSIGNED(2)
Expected behavior is when tolerance timer is expires during controller failover, tolerance timer should restart after failover is done.
Attached /var/log/messages of all nodes
logs
Traces of amfd and amfnd
This issue is reproducible on the change set 6484:dff285e6e84a.
Attached are the AMF traces and configuration after reproducing it.
changeset: 6841:0198c81ad4ad
branch: opensaf-4.5.x
parent: 6838:9ff3ee6e28df
user: praveen.malviya@oracle.com
date: Fri Sep 18 10:13:36 2015 +0530
summary: amfd: act on dep SIs if tol timer expires during controller fail-over [#427]
changeset: 6842:e9b051d8a81e
branch: opensaf-4.6.x
parent: 6839:964e043fa545
user: praveen.malviya@oracle.com
date: Fri Sep 18 10:13:52 2015 +0530
summary: amfd: act on dep SIs if tol timer expires during controller fail-over [#427]
changeset: 6843:6a62544eb8ac
tag: tip
parent: 6840:a2c9e8a31cb4
user: praveen.malviya@oracle.com
date: Fri Sep 18 10:14:03 2015 +0530
summary: amfd: act on dep SIs if tol timer expires during controller fail-over [#427]
Related
Tickets:
#427