Forum Discussion
traffic groups, active/active, and failover
Hey All,
I'm having a bit of difficulty getting this working the way I want. I found plenty of info regarding setting up DSC configs, but very little on how it works with failover. Here's how things are set up:
switch1-----trunk----lb1
| [staging traffic group]
|
switch2-----trunk----lb2
[prod traffic group]
Obviously, I would like for the traffic groups to failover to the other lb if the trunk goes down, and my understanding with our active/standby pairs is that's how it works. However, if I shut down all the trunk interfaces, the lb remains active and does NOT failover the traffic group. I have to manually fail it over. I thought it may have been due to the self-ips not set up for default allowed, but changing that didn't make a difference.
I had to set up an HA group with trunk monitoring/weight, and am able to get it to failover automatically, but based on the weight, it fails over the group if just one interface is shut down, which is not what I'm looking for. I'd only want to fail-over if less than 2 interfaces are active. I thought the active bonus would help here, but....
It appears that the weight is relevant at the device level, and not the traffic-group level, preventing me from manually failing over a traffic group to a unit with no active traffic-groups. For example, if all of the traffic groups are active on lb1, then it receives the active bonus, whereas lb2 will not. If I attempt to manually failover [prod], it sends it to lb2, but then lb2 sends it right back, indicating weight as the reason. In this state, I'm unable to move traffic groups, say once an lb has recovered from a reboot or something.
However, if each lb has active traffic groups, both receive the bonus weight, and I am able to move traffic groups around, UNTIL one unit has no active groups and does not recieve the weight bonus.
Basically, I'm looking for how to best handle failover as it relates to traffic-groups.
Thanks,
-Ken
2 Replies
- Rise_77519
Nimbostratus
Hi Ken,
What about say using the vlan failsafe instead of the HA group.I am using vlan failsafe and satisfy. You can easly configure one or more vlan for vlan failsafe and give a timeout( at leases 10 second) and if any trunk interface goes down then f5 automatically failover the traffic group to another after complated timeot.You can get detail info the below link.
http://support.f5.com/kb/en-us/solu...13297.html
- yammy1688_99834
Nimbostratus
Posted By Rise on 07/17/2013 01:29 AMHi Ken,
What about say using the vlan failsafe instead of the HA group.I am using vlan failsafe and satisfy. You can easly configure one or more vlan for vlan failsafe and give a timeout( at leases 10 second) and if any trunk interface goes down then f5 automatically failover the traffic group to another after complated timeot.You can get detail info the below link.
http://support.f5.com/kb/en-us/solu...13297.html
This seems to work. Thanks!
Help guide the future of your DevCentral Community!
What tools do you use to collaborate? (1min - anonymous)Recent Discussions
Related Content
* Getting Started on DevCentral
* Community Guidelines
* Community Terms of Use / EULA
* Community Ranking Explained
* Community Resources
* Contact the DevCentral Team
* Update MFA on account.f5.com