<div>The problem of redistributing 32-bit metric space to 16-bit metric space and vice versa still remains. The policy tools (add, sub) are fairly crude and error prone for this purpose. As the network scales and goes through other redistributions, then this simple case becomes more complex.</div>
<div><br></div><div>Suggestion #1 - Wouldn't it make more sense to have a native 'redistribute' command built into xorp that takes a number of policies with well-chosen defaults and wraps behind a single command? Quagga has a simple redistribute that can be applied to rip and ospf. </div>
<div><br></div><div>Suggestion #2 - Further, for redistributing 16/32-bit metrics a "bit shift" function would be helpful for the policy tools to use.</div><br>Phil<div><div>On Mon, Jul 23, 2012 at 5:51 AM, Igor Maravić <span dir="ltr"><<a href="mailto:igorm@etf.rs" target="_blank">igorm@etf.rs</a>></span> wrote:</div>
</div><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Good catch :)<div>Well if the redistribution was bad, TAGS would be bad. Because they aren't redistribution is good.</div>
<div><br></div><div>You're problem is that you've hit RIP_INFINITY. RIP_INFINITY is set to 16.</div>
<div>So, routes that have metric larger then 16, won't even enter the RIP code. I have to say that this isn't a redistribution BUG.</div><div><br></div></blockquote></div>