<div dir="ltr">I have 2 nodes in an rabbit cluster, one disc, one ram node. I'm seeing messages get lostwhen testing failover scenarios and want to see if I'm misunderstanding behavior or seeing a bug. I'm using Rabbit 3.0.4 on Erlang R15B03. I've got an HA policy on my queue.<div>
<br></div><div><div>Publish a persistent message to HA queue - Message 1 - the queues are in sync.</div><div>Stop disk node</div><div>Publish a persistent message to HA queue - Message 2 - the queues are in no longer in sync.</div>
<div>Start disk node</div><div>Publish a persistent message to HA queue - Message 3 - the queues are in no longer in sync.</div><div>Stop ram node</div><div style>Checking at this point, I only see message 3 in the queue?? Messages 1 and 2 have disappeared.</div>
<div style>I then start ram node, and am still only seeing message 3, and that the queues are out of sync. Where did message 1 and 2 go?</div><div><br></div><div style>Is there a bug or something I'm missing here? At no point has the entire cluster been shut off, only individual nodes. Am I misunderstanding the intended behavior of clusters here? A guess might have been that messages 1 and 3 existed on the disc node and message 2 and 3 should have been on the ram node. Instead I'm ONLY seeing message 3 in the queue.</div>
<div><br></div>-- <br>Jason McIntosh<br><a href="http://mcintosh.poetshome.com/blog/">http://mcintosh.poetshome.com/blog/</a><br>573-424-7612
</div></div>