CODEWITHSUNDEEP
erlangrabbitmqcluster-computing
D.Zhu
D.Zhu

Reputation: 71

Rabbitmq log flood with discarding message when restart

I have a rabbitmq cluster using mirror queue with two nodes,rabbitmq1 and rabbitmq2.When I restart rabbitmq1(It has master queue.),it log so many error messages as follows:

=ERROR REPORT==== 25-Jan-2017::11:55:20 === Discarding message {'$gen_call',{<0.824.0>,#Ref<0.0.0.98906>},stat} from <0.824.0> to <0.270.0> in an old incarnation (1) of this node (2)

I restart node by using [service rabbitmq-server stop] and [service rabbitmq-server start] and when error happened,start command had no response.

It seems that rabbitmq send message to the old version of node,but why version did not update?

According to log,the slave queue on rabbitmq2 did not detect master queue's death.Does it have something to do with the error?

The situation happens occasionally.How can I make it again?

Upvotes: 7

Views: 12220

Answers (1)

Gabriele Santomaggio
Gabriele Santomaggio

Reputation: 22750

most likely you got this issue: https://github.com/rabbitmq/rabbitmq-server/issues/802

you are using an old version of RabbitMQ, there are lot of fixes and improvement for mirroring.

if you have a chance you should update it

Upvotes: 2

Related Questions