HI Vikas, thank you for your reply.
I totally understand your point, but saying that factors to not recomend any newer realease are several as "Bugs reported, Stability, Customer acceptance/deployability, feature parity etc" also means that during the following 13 (!!!!) releases, Juniper engineers haven't been able to guarrantee anything better in term of "Bugs reported, Stability, Customer acceptance/deployability, feature parity etc"....
And that's quite worriying.
More than that, if they are not able to find a better software, why don't they:
A) Stop releasing new versions, totally pointless as no one deserves to install them, as they are not realiable according to the experts and Juniper itself
B) Increase de end of support date of the latest stable and recommended version, 15.1?
I can't understand Juniper at this point. The thing is that I have these nodes on my desk (e.g. not in production yet) and now it would be really easy and smooth to upgrade them to any other Junos version, while in less than a year it will be much more painful and risky to upgrade that cluster.
This is not showing up good sensations from the reliability of Junos, and the logic between the releases and the end of life - end of support of those releases, and gives us, the network admins, a really narrow capacity to plan and set up the things reliable for the mid-long term.
ThankS!