Considering only Node lock case in the ticket.
Full implementation with faults and with other admin operation will be taken up
in future.
---
** [tickets:#90] In nway model, quiesced callbacks are generated for sponser SI
first then dependent SIs after payload node lock**
**Status:** review
**Created:** Mon May 13, 2013 04:54 AM UTC by Nagendra Kumar
**Last Updated:** Fri Sep 06, 2013 12:39 PM UTC
**Owner:** Praveen
Migrated from http://devel.opensaf.org/ticket/2574
Cgset : 3406
In nway model, after payload node lock observed that quiesced callbacks are
received first for sponser SIs then the dependent SIs
Configuration:-
================
Nway model
5 Node setup
2SUs, 1 component per SU with same compType
4SIs, 1CSIs per SI, No SIRankedSU configured.
saAmfSGMaxActiveSIsperSU=3
saAmfSGMaxStandbySIsperSU=3
saAmfSIPrefStandbyAssignments=2
SU1 spawned on PL-3
SU2 spawned on PL-4
SI-SI dependency configured as shown below :-
SI1 sponser to SI2 with tolerance timer 1min
SI2 sponser to SI3 with tolerance timer 1min
SI3 sponser to SI4 with tolerance timer 1min
SI1 -> SI2 -> SI3 -> SI4
Steps:-
========
1. Performed unlock-in and unlock of SU1 and SU2.
Initial assignments were as shown below:
SU1 : SI1 SI2 SI3 Active and SI4 standby
SU2 : SI1 SI2 SI3 Standby and SI4 Active
2. Do switchover
3. PL-3 node lock on which SU1 was running.
amf-adm lock safAmfNode=PL-3,safAmfCluster=myAmfCluster
After this assignments were as shown below:-
SU2 - SI1 SI2 SI4 active and SI3 standby
SU1 - No assignments
SC-1 standby
SC-2 active
Here observed that quiesced callbacks first got to the sponser SIs and then
dependent SIs [problem]
Component logs after step 3:-
==========================
3 8 18:43:18.528309
safComp=COMP1,safSu=SU1,safSg=SG,safApp=testNwayApp DataReceived? {'DATA':
'CSISET_QUIESCED_CALLBACK', 'csiName':
'safCsi=CSI1,safSi=SI1,safApp=testNwayApp', 'prevCsiAssigned':
['safCsi=CSI1,safSi=SI1,safApp=testNwayApp',
'safCsi=CSI1,safSi=SI2,safApp=testNwayApp',
'safCsi=CSI1,safSi=SI3,safApp=testNwayApp',
'safCsi=CSI1,safSi=SI4,safApp=testNwayApp']}
3 8 18:43:18.533090 safComp=COMP1,safSu=SU1,safSg=SG,safApp=testNwayApp
DataReceived? {'DATA': 'CSISET_QUIESCED_CALLBACK', 'csiName':
'safCsi=CSI1,safSi=SI2,safApp=testNwayApp', 'prevCsiAssigned':
['safCsi=CSI1,safSi=SI1,safApp=testNwayApp',
'safCsi=CSI1,safSi=SI2,safApp=testNwayApp',
'safCsi=CSI1,safSi=SI3,safApp=testNwayApp',
'safCsi=CSI1,safSi=SI4,safApp=testNwayApp']}
3 8 18:43:18.535694 safComp=COMP1,safSu=SU1,safSg=SG,safApp=testNwayApp
DataReceived? {'DATA': 'CSISET_QUIESCED_CALLBACK', 'csiName':
'safCsi=CSI1,safSi=SI3,safApp=testNwayApp', 'prevCsiAssigned':
['safCsi=CSI1,safSi=SI1,safApp=testNwayApp',
'safCsi=CSI1,safSi=SI2,safApp=testNwayApp',
'safCsi=CSI1,safSi=SI3,safApp=testNwayApp',
'safCsi=CSI1,safSi=SI4,safApp=testNwayApp']}
3 8 18:43:18.548191 safComp=COMP1,safSu=SU2,safSg=SG,safApp=testNwayApp
DataReceived? {'DATA': 'CSISET_ACTIVE_CALLBACK', 'csiName':
'safCsi=CSI1,safSi=SI1,safApp=testNwayApp', 'prevCsiAssigned':
['safCsi=CSI1,safSi=SI4,safApp=testNwayApp',
'safCsi=CSI1,safSi=SI1,safApp=testNwayApp',
'safCsi=CSI1,safSi=SI2,safApp=testNwayApp',
'safCsi=CSI1,safSi=SI3,safApp=testNwayApp']}
3 8 18:43:18.564020 safComp=COMP1,safSu=SU1,safSg=SG,safApp=testNwayApp
DataReceived? {'DATA': 'CSIREMOVE_CALLBACK', 'csiName': , 'prevCsiAssigned':
['safCsi=CSI1,safSi=SI1,safApp=testNwayApp',
'safCsi=CSI1,safSi=SI2,safApp=testNwayApp',
'safCsi=CSI1,safSi=SI3,safApp=testNwayApp',
'safCsi=CSI1,safSi=SI4,safApp=testNwayApp']}
3 8 18:43:18.590242 safComp=COMP1,safSu=SU2,safSg=SG,safApp=testNwayApp
DataReceived? {'DATA': 'CSISET_ACTIVE_CALLBACK', 'csiName':
'safCsi=CSI1,safSi=SI2,safApp=testNwayApp', 'prevCsiAssigned':
['safCsi=CSI1,safSi=SI4,safApp=testNwayApp',
'safCsi=CSI1,safSi=SI1,safApp=testNwayApp',
'safCsi=CSI1,safSi=SI2,safApp=testNwayApp',
'safCsi=CSI1,safSi=SI3,safApp=testNwayApp',
'safCsi=CSI1,safSi=SI1,safApp=testNwayApp']}
4. Further performed lock of sponser SI1 followed by controller
switchover.
assignments were as shown below:-
SU2 - SI4 SI2 active and SI3 standby
5. Performed lock of PL-4 on which SU2 was spawned but operation failed with
below error.
# amf-adm lock safAmfNode=PL-4,safAmfCluster=myAmfCluster
error - command timed out (alarm)
safAmfNode=PL-3,safAmfCluster=myAmfCluster
saAmfNodeAdminState=LOCKED(2)
saAmfNodeOperState=ENABLED(1)
safAmfNode=PL-4,safAmfCluster=myAmfCluster
saAmfNodeAdminState=UNLOCKED(1)
saAmfNodeOperState=ENABLED(1)
/var/log/messages of SC-1 was printing the below messages:-
===
Mar 8 18:44:01 linux-xc76 osafamfd[11121]: SG of SU on this node not in STABLE
state (safSu=SU2,safSg=SG,safApp=testNwayApp)
Mar 8 18:44:02 linux-xc76 osafamfd[11121]: SG of SU on this node not in STABLE
state (safSu=SU2,safSg=SG,safApp=testNwayApp)
Mar 8 18:44:03 linux-xc76 osafamfd[11121]: SG of SU on this node not in STABLE
state (safSu=SU2,safSg=SG,safApp=testNwayApp)
Mar 8 18:44:04 linux-xc76 osafamfd[11121]: SG of SU on this node not in STABLE
state (safSu=SU2,safSg=SG,safApp=testNwayApp)
Mar 8 18:44:05 linux-xc76 osafamfd[11121]: SG of SU on this node not in STABLE
state (safSu=SU2,safSg=SG,safApp=testNwayApp)
Mar 8 18:44:06 linux-xc76 osafamfd[11121]: SG of SU on this node not in STABLE
state (safSu=SU2,safSg=SG,safApp=testNwayApp)
Mar 8 18:44:07 linux-xc76 osafamfd[11121]: SG of SU on this node not in STABLE
state (safSu=SU2,safSg=SG,safApp=testNwayApp)
Mar 8 18:44:08 linux-xc76 osafamfd[11121]: SG of SU on this node not in STABLE
state (safSu=SU2,safSg=SG,safApp=testNwayApp)
Mar 8 18:44:09 linux-xc76 osafamfd[11121]: SG of SU on this node not in STABLE
state (safSu=SU2,safSg=SG,safApp=testNwayApp)
Mar 8 18:44:10 linux-xc76 osafamfd[11121]: SG of SU on this node not in STABLE
state (safSu=SU2,safSg=SG,safApp=testNwayApp)
Mar 8 18:44:11 linux-xc76 osafamfd[11121]: SG of SU on this node not in STABLE
state (safSu=SU2,safSg=SG,safApp=testNwayApp)
And finally the dependent SIs (SI4, SI2 and SI3 ) assigned to SU2 holds the
assignments for ever. They did not move to unassigned state.
safSISU=safSu=SU2\,safSg=SG\,safApp=testNwayApp,safSi=SI4,safApp=testNwayApp
saAmfSISUHAState=ACTIVE(1)
safSISU=safSu=SU2\,safSg=SG\,safApp=testNwayApp,safSi=SI2,safApp=testNwayApp
saAmfSISUHAState=ACTIVE(1)
safSISU=safSu=SU2\,safSg=SG\,safApp=testNwayApp,safSi=SI3,safApp=testNwayApp
saAmfSISUHAState=STANDBY(2)
linux-xc76:/opt/goahead/tetware/framework #
linux-xc76:/opt/goahead/tetware/framework #
linux-xc76:/opt/goahead/tetware/framework # amf-state si
safSi=NoRed?1,safApp=OpenSAF
saAmfSIAdminState=UNLOCKED(1)
saAmfSIAssignmentState=FULLY_ASSIGNED(2)
safSi=NoRed?2,safApp=OpenSAF
saAmfSIAdminState=UNLOCKED(1)
saAmfSIAssignmentState=FULLY_ASSIGNED(2)
safSi=NoRed?3,safApp=OpenSAF
saAmfSIAdminState=UNLOCKED(1)
saAmfSIAssignmentState=FULLY_ASSIGNED(2)
safSi=NoRed?4,safApp=OpenSAF
saAmfSIAdminState=UNLOCKED(1)
saAmfSIAssignmentState=FULLY_ASSIGNED(2)
safSi=NoRed?5,safApp=OpenSAF
saAmfSIAdminState=UNLOCKED(1)
saAmfSIAssignmentState=FULLY_ASSIGNED(2)
safSi=SC-2N,safApp=OpenSAF
saAmfSIAdminState=UNLOCKED(1)
saAmfSIAssignmentState=FULLY_ASSIGNED(2)
safSi=SI1,safApp=testNwayApp
saAmfSIAdminState=LOCKED(2)
saAmfSIAssignmentState=UNASSIGNED(1)
safSi=SI2,safApp=testNwayApp
saAmfSIAdminState=UNLOCKED(1)
saAmfSIAssignmentState=PARTIALLY_ASSIGNED(3)
safSi=SI3,safApp=testNwayApp
saAmfSIAdminState=UNLOCKED(1)
saAmfSIAssignmentState=PARTIALLY_ASSIGNED(3)
safSi=SI4,safApp=testNwayApp
saAmfSIAdminState=UNLOCKED(1)
saAmfSIAssignmentState=PARTIALLY_ASSIGNED(3)
Change History
Changed 14 months ago by ravisekhar ¶
■version changed from 4.2.0 to 4.1.0
■milestone changed from 4.2.1 to future_releases
un-assignment order(dependency —> sponsor) based on dependency is not yet
implemented in NWay
Changed 14 months ago by nagendra ¶
■type changed from defect to enhancement
---
Sent from sourceforge.net because [email protected] 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.------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60133471&iu=/4140/ostg.clktrk
_______________________________________________
Opensaf-tickets mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/opensaf-tickets