On Fri, Aug 19, 2011 at 12:28:47PM -0400, Simo Sorce wrote:
Perhaps we should resolve all servers at the same time with parallel
searches so at least for the DNS step we have one single timeout and no
risk of leaving stuff out.
I was thinking about this as well, this would be an elegant solution to
the service vs. server timeout issue.
My only worry was that especially with SRV lookups we could potentially
issue many DNS queries..but perhaps we could limit the number of concurrently
running queries in similar way how we do it with RFC2307bis LDAP
searches w/o nesting.