In regards to ClusterXL Load Sharing, it is true that there are no official recommendations against it but it was "unofficially" discouraged. But based on my experience I went a lot further in the first edition of my book, explicitly stating in writing that ClusterXL Load Sharing introduces a level of complexity that is generally not worth the performance gains in most situations. This pronouncement remained basically the same in the second edition of my book.
Now I'll admit that I was quite nervous about this statement in the first edition (along with some others involving SecureXL) and actively wondered if I should soften it prior to publication, expecting that I might catch some serious flak for it from individuals both inside and outside of Check Point. To date the amount of grief I have received for this statement is...exactly zero. In fact I have received wholehearted support and praise for saying it, which I want to reiterate is only my personal opinion.
Even though I am definitely not a fan of ClusterXL Load Sharing, I was a bit surprised to see that it wasn't supported in R80.20, along with the ability to use IPv6 with the new 3.10 kernel. I assume R80.30 will eliminate these limitations, especially since the R80.30 EA was announced shortly after R80.20 was officially recommended.
"IPS Immersion Training" Self-paced Video Class
Now Available at http://www.maxpowerfirewalls.com
New 2-day Live "Max Power" Series Course Now Available:
"Gateway Performance Optimization R81.20" at maxpowerfirewalls.com