SSL routines:tls_process_client_hello:version too low

[Sergey Kandaurov]

> On 28 Sep 2020, at 23:12, jriker1 <nginx-forum at forum.nginx.org> wrote:
> 
> Hope I can post this as Chrome keeps complaining this site has a data
> breach.

The primary interface is using mailing lists:
http://nginx.org/en/support.html

> 
> I have been using NGINX to route my 443 traffic for two servers for a while
> now.  Now I can't get my RDP side of things working.  Not sure why as it
> used to work.  RDP is thru Essentials Server 2016 and it's Remote Web
> Access.  I can get to the website but when I try to pull up one of my
> servers it fails.  So the website works, the second server I'm using on 443
> works, but not the RDP side of things.  it does work if I remove NGINX and
> set 443 to route directly to the Essentials Server thru my router so know
> it's working still.  What I get in the error logs when this happens is:
> 
> 2020/09/28 05:09:50 [crit] 7556#7556: *1366 SSL_do_handshake() failed (SSL:
> error:1417D18C:SSL routines:tls_process_client_hello:version too low) while
> SSL handshaking, client: 107.6.171.130, server: 0.0.0.0:443
> 2020/09/28 06:01:06 [crit] 7556#7556: *1385 SSL_do_handshake() failed (SSL:
> error:1417D18C:SSL routines:tls_process_client_hello:version too low) while
> SSL handshaking, client: 107.6.171.130, server: 0.0.0.0:443
> 

You didn't provide nginx version, but
the error message suggests you are using OpenSSL 1.1.0.

I don't know what is Essentials Server 2016, but if it is something
from Microsoft, then I heard that it still prefers to use TLSv1.

Then you might want to look at this thread as somewhat related:
http://mailman.nginx.org/pipermail/nginx/2018-November/057154.html

-- 
Sergey Kandaurov