Removing from load balancer after marathon killing the tasks

[vamsu]

I am wondering how bamboo handles this use case. Let’s say we created a marathon app with minimumHealthCapacity=0.9 and maximumOverCapacity=0 and this does a 10% rolling restart. Based on our understanding of Bamboo, it would listen to marathon events and update HAProxy. Does that mean Bamboo updates HAProxy after marathon kills the 10% tasks? If that is the case, wouldn't HAProxy send the requests to the old nodes during brief time period between marathon sending the event and Bamboo reloading HAProxy configuration?

[lclarkmichalek]

Bamboo adds and removes tasks as they appear in Marathon's list of tasks. There may be a delay. Use HAProxy health checks to handle this. On 15 Apr 2016 6:26 am, "vamsu" notifications@github.com wrote:

I am wondering how bamboo handles this use case. Let’s say we created a marathon app with minimumHealthCapacity=0.9 and maximumOverCapacity=0 and this does a 10% rolling restart. Based on our understanding of Bamboo, it would listen to marathon events and update HAProxy. Does that mean Bamboo updates HAProxy after marathon kills the 10% tasks? If that is the case, wouldn't HAProxy send the requests to the old nodes during brief time period between marathon sending the event and Bamboo reloading HAProxy configuration?

— You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub https://github.com/QubitProducts/bamboo/issues/202