Open ajays20078 opened 8 years ago
My initial thought is that this is more trouble than it's probably worth. Can you describe a use case where this would be a compelling feature? I think it would introduce a lot of complexity, and therefore also introduce a number of new bugs.
The only usecase i could think of, from mlb point of view is
event
mode and in sse
mode) and it goes out of sync from the other mlbs.(And it wont get updated untill next event/notification).But if we think about zdd, the script currently queries all the HAProxy instances to check if the update is propagated, which will have scaling issues once HAProxy instances increase.
Can we have clustering support for mlb, where there is a cluster leader(who pulls config from marathon) and others update their configs from the master/leader.