Forum Discussion
Nimbostratusfailover config
I've setup 2 new units (Active/Standby)
PS: I do not have fail-back enabled on both units.
When a port channel fails on Active unit Standby goes to Active for example, 5 mins later port channel comes up, I don't want the Standby (old Active) to go back into Active. I want the Active to remain Active. Why does this keep flipping back? i.e - port channel fails flips over to Standby which is now Active STAY THERE! even if the Standby unit (old Active) port channel comes backup I want this to remain in STANDBY mode. Otherwise I'm going to get another outage/blip with these Units flipping
How can this be achieved, I've read the docs but still confusing to me. STOP FLIPPING BACK, stay on the other unit please even you are available again.......
Many thanks, anyone
12 Replies
portoalegrei.e - like preempt command under HSRP (Cisco) - stay there!- nitass
Employee
it should not failback (because auto failback is disabled). have you checked /var/log/ltm? what did it say? 
Ajmal_2608This will disable the preemption
tmsh modify sys failover force-active disabled
tmsh modify sys failover force-standby disabled
- portoalegre
Nitass
ar 2 18:50:59 unit-f5 notice sod[5743]: 010c0053:5: Active for traffic group /Common/traffic-group-1.
Mar 2 18:50:59 unit-f5 notice sod[5743]: 010c0019:5: Active
Mar 2 18:51:01 unit-f5 notice logger: /usr/bin/tmipsecd --tmmcount 4 ==> /usr/bin/bigstart start racoon
Mar 2 18:51:07 unit-f5 notice sod[5743]: 010c0045:5: Leaving active, group score 40 peer group score 40.
Mar 2 18:51:07 unit-f5 notice sod[5743]: 010c0052:5: Standby for traffic group /Common/traffic-group-1.
Mar 2 18:51:07 unit-f5 notice sod[5743]: 010c0018:5: Standby
Mar 2 18:51:07 unit-f5 notice logger: /usr/bin/tmipsecd --tmmcount 4 ==> /usr/bin/bigstart stop racoon
Mar 2 18:51:13 unit-f5 err tmm[8143]: 01340002:3: HA Connection with peer 10.170.135.36:1028 lost.
Mar 2 18:51:13 unit-f5 err tmm1[8144]: 01340002:3: HA Connection with peer 10.170.135.36:1028 lost.
Mar 2 18:51:13 unit-f5 err tmm3[8146]: 01340002:3: HA Connection with peer 10.170.135.36:1028 lost.
Mar 2 18:51:13 unit-f5 err tmm[8143]: 01340002:3: HA Connection with peer 10.166.142.13:1028 lost.
Mar 2 18:51:13 unit-f5 err tmm1[8144]: 01340002:3: HA Connection with peer 10.166.142.13:1028 lost.
Mar 2 18:51:13 unit-f5 err tmm2[8145]: 01340002:3: HA Connection with peer 10.170.135.36:1028 lost.
Mar 2 18:51:13 unit-f5 err tmm2[8145]: 01340002:3: HA Connection with peer 10.166.142.13:1028 lost.
Mar 2 18:51:13 unit-f5 err tmm3[8146]: 01340002:3: HA Connection with peer 10.166.142.13:1028 lost.
Mar 2 18:51:14 unit-f5 notice sod[5743]: 010c0053:5: Active for traffic group /Common/traffic-group-1.
Mar 2 18:51:14 unit-f5 notice sod[5743]: 010c0019:5: Active - nitass
the problem is the last log message, isn't it?
Mar 2 18:51:14 unit-f5 notice sod[5743]: 010c0053:5: Active for traffic group /Common/traffic-group-1.
Mar 2 18:51:14 unit-f5 notice sod[5743]: 010c0019:5: Active
you are using ha-group, aren't you? if so, can you post ha-group configuration? - portoalegre
Stanby unit
root@(standby-f5)(cfg-sync In Sync)(Standby)(/Common)(tmos) show sys ha-group detail
---------------------------
Sys::HA Group: Port-Channel
---------------------------
State enabled
Active Bonus 0
Score 40
---------------------------------------------------
| Sys::HA Group Trunk: Port-Channel:Port-channel104
---------------------------------------------------
| Threshold 0
| Percent Up 100
| Weight 40
| Score Contribution 40Active Unit
[root@-f5:Active:In Sync] config tmsh show sys ha-group detail
---------------------------
Sys::HA Group: Port-Channel
---------------------------
State enabled
Active Bonus 0
Score 40
---------------------------------------------------
| Sys::HA Group Trunk: Port-Channel:Port-channel104
---------------------------------------------------
| Threshold 0
| Percent Up 100
| Weight 40
| Score Contribution 40 - nitasscan you try to configure active bonus?
Specifying an active bonus
http://support.f5.com/kb/en-us/products/big-ip_ltm/manuals/product/tmos_management_guide_10_1/tmos_high_avail.html1026652 - portoalegreok, both units have a weight of 40 (4 x ports trunked), I changed the Active bonus to 5 on each unit. When I failed the Active unit (forced Standby), the Standby unit went into Active mode and stayed there.
However I did notice the traffic scores are the same, i.e - 12000 vs 4000. Why does not this change and what makes up these scores? Does anyone know. For me the scores should be changed after failing over but does not. i.e
)(Active)(/Common)(tmos) sh cm traffic-group all-properties
---------------------------------------------------------------------------------------------------------
CM::Traffic-Group
Name Device Status Next Auto Failback Failover
Active Time (in sec) Score
---------------------------------------------------------------------------------------------------------
traffic-group-1 unitA standby true 60 12258
traffic-group-1 unitB active - 60 4224
traffic-group-local-only - - - 60 - - portoalegreIt should be noted with the above, I changed the Active bonus to 10 on the Active and Standby to 0 firstly with both weight of 40, the failover did not work. Unit A went Active>Standby>then Active again. Specifing a Active bonus of 5 solved the problem.
- nitassHowever I did notice the traffic scores are the same, i.e - 12000 vs 4000. Why does not this change and what makes up these scores? Does anyone know. For me the scores should be changed after failing over but does not.traffic group score is calculated from traffic group name and management ip. it is used as tie-break when two units are active at the same time.
