<html><body><div style="color:#000; background-color:#fff; font-family:times new roman, new york, times, serif;font-size:12pt"><div><span><font size="3">Thanks Matt. It would be nice if the HA documentation is revisited again and </font>explain the basic steps well, it's of great help to the community.</span></div><div><span><font size="3"><br></font></span></div><div><span><font size="3">Best,</font></span></div><div><span><font size="3">Uday. </font></span></div><div><span><font size="3"> </font></span></div><div style="font-family: 'times new roman', 'new york', times, serif; font-size: 12pt; "><br></div> <div style="font-size: 12pt; font-family: 'times new roman', 'new york', times, serif; "> <div style="font-size: 12pt; font-family: 'times new roman', 'new york', times, serif; "> <div dir="ltr"> <font size="2" face="Arial"> <hr size="1"> <b><span style="font-weight:bold;">From:</span></b> Matthew Sackman
<matthew@rabbitmq.com><br> <b><span style="font-weight: bold;">To:</span></b> rabbitmq-discuss@lists.rabbitmq.com <br> <b><span style="font-weight: bold;">Sent:</span></b> Thursday, February 16, 2012 3:14 AM<br> <b><span style="font-weight: bold;">Subject:</span></b> Re: [rabbitmq-discuss] HA Documentation Error?<br> </font> </div> <br>
On Mon, Feb 13, 2012 at 10:15:16AM +0000, Emile Joubert wrote:<br>> On 10/02/12 19:50, Uday Subbarayan wrote:<br>> > Note: There is a risk in the above strategy if the master node goes<br>> > down. If you want to avoid, then publish to all slaves? Here on wards, i<br>> <br>> If the master becomes unavailable then a slave is promoted to take its<br>> place. Message replication between slaves and the master happens<br>> internally and automatically. As long as the queue is declared as<br>> mirrored, the publisher is not required to take any action to republish<br>> messages to slaves.<br><br>Well, that's not necessarily true. If the node that failed was also the<br>node to which the client was connected, then the client needs to detect<br>the death of the connection, establish a new connection to a surviving<br>node and then republish all messages that it never got<br>publisher-confirms back for - assuming the client even
turned<br>publisher-confirms on. That would be the only way to ensure<br>at-least-once delivery.<br><br>> Active-active mirrored queues are recommended instead of the<br>> active-passive approach, because it does not require additional software<br>> and the failover process is quicker.<br><br>However, throughput is lower, and it does not have any chance of<br>integration with other clustering/HA software elsewhere in the stack.<br><br>Matthew<br>_______________________________________________<br>rabbitmq-discuss mailing list<br><a ymailto="mailto:rabbitmq-discuss@lists.rabbitmq.com" href="mailto:rabbitmq-discuss@lists.rabbitmq.com">rabbitmq-discuss@lists.rabbitmq.com</a><br><a href="https://lists.rabbitmq.com/cgi-bin/mailman/listinfo/rabbitmq-discuss" target="_blank">https://lists.rabbitmq.com/cgi-bin/mailman/listinfo/rabbitmq-discuss</a><br><br><br> </div> </div> </div></body></html>