Comentários (2)

1 Joseph.Gramig comentou às Link permanente

You left out this important tid bit at this link: <br /> http://pic.dhe.ibm.com/infocenter/wmqv7/v7r5/topic/com.ibm.mq.con.doc/q017270_.htm <div>&nbsp;</div> "The full repositories republish the publications they receive through the manually defined CLUSSDR channels, which must point to other full repositories in the cluster."

2 Anthony Beardsmore comentou às Link permanente

Hi Joseph, thanks for the comment. That's a very good point, it's important to ensure Full Repositories are always directly connected by sender channels to all other FRs for this reason and to 'bootstrap' the cluster first time or following a REFRESH, and that partial repositories define a CLUSSDR pointing to an FR as an initial bootstrap mechanism. One of the most common reasons for unexpected behaviour in a cluster is that some of these CLUSSDR definitions are missing. (In general, the content in this blog should not replace a good understanding of steps detailed in the Infocenter for configuring your system, but it is rather intended to help give some detailed background on the 'why' as well as the 'what and how' - I hope this is useful.)