I've set up a server on my local network, but I'm stymied troubleshooting HTTPS connections. When I use curl -v to make a request, I'm told, "requested domain name does not match the server's certificate."
But the output itself appears to indicate that the domain does match: "reg.qa"
What's the real issue here?
$ curl -v https://reg.qa/
* Added reg.qa:443:172.18.0.4 to DNS cache
* About to connect() to reg.qa port 443 (#0)
* Trying 172.18.0.4...
* Connected to reg.qa (172.18.0.4) port 443 (#0)
* Initializing NSS with certpath: sql:/etc/pki/nssdb
* CAfile: /etc/pki/tls/certs/ca-bundle.crt
CApath: none
* Server certificate:
* subject: CN=reg.qa,OU=Web Test,O=MYORG,ST=CA,C=US
* start date: Mar 11 20:53:04 2020 GMT
* expire date: Mar 10 20:53:04 2025 GMT
* common name: reg.qa
* issuer: CN=reg.qa,OU=Web Test,O=MYORG,ST=CA,C=US
* NSS error -12276 (SSL_ERROR_BAD_CERT_DOMAIN)
* Unable to communicate securely with peer: requested domain name does not match the server's certificate.
* Closing connection 0
curl: (51) Unable to communicate securely with peer: requested domain name does not match the server's certificate.
NSS error -12276 (SSL_ERROR_BAD_CERT_DOMAIN)
As far as I know newer versions of NSS require that the certificate uses subject alternative names (SAN) to describe the valid domains. Using the common name (CN) is obsolete for many years and several browsers and TLS stacks started to enforce this a while ago.
I am trying to connect to server B from server A using curl (https). I have already tried with -k and it doesn't work.
I have looked into several posts and I spotted blog on this link but still issue exists.
When I do a curl from server A, I am getting following error:
* Rebuilt URL to: https://x.x.x.x:8443/
* Hostname was NOT found in DNS cache
* Trying x.x.x.x...
* Connected to x.x.x.x (x.x.x.x) port 8443 (#0)
* successfully set certificate verify locations:
* CAfile: /tmp/cert_test/certRepo
CApath: /etc/ssl/certs/
* SSLv3, TLS handshake, Client hello (1):
* error:140770FC:SSL routines:SSL23_GET_SERVER_HELLO:unknown protocol
* Closing connection 0
curl: (35) error:140770FC:**SSL routines:SSL23_GET_SERVER_HELLO:**unknown protocol
I went on the server B (https://x.x.x.x:8443/) from the browser and downloaded the root, intermediate and the client certificate. As suggested in the blog, I have created a new folder and combined all the public certs into one directory and tried to execute the curl command
curl -v --cacert /tmp/cert_test/certRepo https://x.x.x.x:8443
I am getting GET_SERVER_HELLO:unknown protocol
any thoughts?
Curl version from the Client machine:
curl 7.37.0 (x86_64-suse-linux-gnu)
libcurl/7.37.0 OpenSSL/0.9.8j
zlib/1.2.7
libidn/1.10
Protocols: dict file ftp ftps gopher http https imap imaps ldap ldaps pop3 pop3s rtsp smtp smtps telnet
tftp
Features: GSS-Negotiate IDN IPv6 Largefile NTLM NTLM_WB SSL libz
I am very sure the server is using TLSv1.2.
you did not post your curl/libssl version, but my best guess is that you're using an ancient build of a ssl/tls library, and/or an ancient version of curl which does not support whatever version of ssl/tls that server us ysubg. update your libssl and curl and try again. also post the output of curl --version.
PS, if you're on linux, you can get rough curl+openssl compile instructions here.
I'm trying to set up a frontend redirect using Traefik, but the docs don't explain how to do so.
I have an existing VirtualHost configuration in Apache that does this via:
ServerName jira.mycompany.com
RedirectMatch /.* https://mycompany.atlassian.net/
I thought that I should be able to do the same thing via my routes.toml file, like so:
[frontends]
[frontends.jira]
backend = "jira"
[frontends.jira.redirect]
regex = "^https?://jira.mycompany.com/(.*)"
replacement = "https://mycompany.atlassian.net/$1"
[backends]
[backend.jira]
[backend.jira.servers.primary]
url = "https://mycompany.atlassian.net"
However, trying to connect to https://jira.mycompany.com gives me an SSL protocol error.
curl -v tells me the following:
Toms-iMac:~ tgolden$ curl -v https://jira.mycompany.com
* Rebuilt URL to: https://jira.mycompany.com/
* Trying <external IP>...
* TCP_NODELAY set
* Connected to jira.mycompany.com (<external IP>) port 443 (#0)
* ALPN, offering h2
* ALPN, offering http/1.1
* Cipher selection: ALL:!EXPORT:!EXPORT40:!EXPORT56:!aNULL:!LOW:!RC4:#STRENGTH
* successfully set certificate verify locations:
* CAfile: /etc/ssl/cert.pem
CApath: none
* TLSv1.2 (OUT), TLS handshake, Client hello (1):
* TLSv1.2 (IN), TLS alert, Server hello (2):
* error:14077438:SSL routines:SSL23_GET_SERVER_HELLO:tlsv1 alert internal error
* stopped the pause stream!
* Closing connection 0
curl: (35) error:14077438:SSL routines:SSL23_GET_SERVER_HELLO:tlsv1 alert internal error
What am I missing here?
I tried adding a "Host:jira.mycompany.com" rule as well as SSLRedirect and SSLTemporaryRedirect headers as well -- no dice.
The reason for the SSL error was due to my Traefik configuration using the deprecated acme.ondemand parameter. For an unrelated reason, the ACME challenge failed when trying to get a certificate for jira.mycompany.com, which caused the SSL handshake to fail.
I'm surprised the failed challenge for ondemand wouldn't fall back to the self-signed Traefik cert, but since the parameter was deprecated it shouldn't be as surprising.
Once I resolved this issue (ignoring the cert issue for the moment) the redirect worked as expected.
When I try to connect to any server (e.g. google.com) using curl (or libcurl) I get the error message:
curl: (35) error:1408F10B:SSL routines:ssl3_get_record:wrong version number
Verbose output:
$ curl www.google.com --verbose
* Rebuilt URL to: www.google.com/
* Uses proxy env variable no_proxy == 'localhost,127.0.0.1,localaddress,.localdomain.com'
* Uses proxy env variable http_proxy == 'https://proxy.in.tum.de:8080'
* Trying 131.159.0.2...
* TCP_NODELAY set
* Connected to proxy.in.tum.de (131.159.0.2) port 8080 (#0)
* successfully set certificate verify locations:
* CAfile: /etc/ssl/certs/ca-certificates.crt
CApath: none
* TLSv1.3 (OUT), TLS handshake, Client hello (1):
* error:1408F10B:SSL routines:ssl3_get_record:wrong version number
* Closing connection 0
curl: (35) error:1408F10B:SSL routines:ssl3_get_record:wrong version number'
For some reason curl seems to use TLSv1.3 even if I force it to use TLSv1.2 with the command --tlsv1.2 (it will still print TLSv1.3 (OUT), ..."
I am using the newest version of both Curl and OpenSSL :
$ curl -V
curl 7.61.0-DEV (x86_64-pc-linux-gnu) libcurl/7.61.0-DEV OpenSSL/1.1.1 zlib/1.2.8
Release-Date: [unreleased]
Protocols: dict file ftp ftps gopher http https imap imaps pop3 pop3s rtsp smb smbs smtp smtps telnet tftp
Features: AsynchDNS IPv6 Largefile NTLM NTLM_WB SSL libz TLS-SRP UnixSockets HTTPS-proxy
I think this is a problem related to my installation of the programms.
Can somebody explain to me what this error message means?
* Uses proxy env variable http_proxy == 'https://proxy.in.tum.de:8080'
^^^^^
The https:// is wrong, it should be http://. The proxy itself should be accessed by HTTP and not HTTPS even though the target URL is HTTPS. The proxy will nevertheless properly handle HTTPS connection and keep the end-to-end encryption. See HTTP CONNECT method for details how this is done.
If anyone is getting this error using Nginx, try adding the following to your server config:
server {
listen 443 ssl;
...
}
The issue stems from Nginx serving an HTTP server to a client expecting HTTPS on whatever port you're listening on. When you specify ssl in the listen directive, you clear this up on the server side.
This is a telltale error that you are serving HTTP from the HTTPS port.
You can easily test with telnet
telnet FQDN 443
GET / HTTP/1.0
[hit return twice]
and if you see regular HTTP document here [not some kind of error], you know that your configuration is incorrect and the responding server is not SSL encrypting the response.
Simple answer
If you are behind a proxy server, please set the proxy for curl. The curl is not able to connect to server so it shows wrong version number.
Set proxy by opening subl ~/.curlrc or use any other text editor. Then add the following line to file:
proxy= proxyserver:proxyport
For e.g. proxy = 10.8.0.1:8080
If you are not behind a proxy, make sure that the curlrc file does not contain the proxy settings.
Also check your /etc/hosts file. Wasted 2 hours on this. If you have an url rerouted to 127.0.0.1 or any other loopback, this will fail the ssl handshake.
In my case the cause of this error was that my web server was not configured to listen to IPv6 on SSL port 443. After enabling it the error disappeared.
Here's how you do it for Apache:
<VirtualHost ip.v4.address:443 ip:v::6:address:443>
...
</VirtualHost>
And for nginx:
listen 443 ssl http2;
listen [::]:443 ssl http2;
Thanks to #bret-weinraub,
I found that something is weird about the server's reply. After a bit of investigation, it turned out that I have a static IP in /etc/hosts file for the target domain and as they have changed their IP address I'm not getting to the correct server.
More simply in one line:
proxy=192.168.2.1:8080;curl -v example.com
eg. $proxy=192.168.2.1:8080;curl -v example.com
xxxxxxxxx-ASUS:~$ proxy=192.168.2.1:8080;curl -v https://google.com|head -c 15 % Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0
* Trying 172.217.163.46:443...
* TCP_NODELAY set
* Connected to google.com (172.217.163.46) port 443 (#0)
* ALPN, offering h2
* ALPN, offering http/1.1
* successfully set certificate verify locations:
* CAfile: /etc/ssl/certs/ca-certificates.crt
CApath: /etc/ssl/certs
} [5 bytes data]
* TLSv1.3 (OUT), TLS handshake, Client hello (1):
} [512 bytes data]
Another possible cause of this problem is if you have not enabled the virtual host's configuration file in Apache (or if you don't have that virtual host at all) and the default virtual host in Apache is only configured for non-SSL connections -- ie there's no default virtual host which can talk SSL. In this case because Apache is listening on port 443 the request for the virtual host that doesn't exist will arrive at the default virtual host -- but that virtual host doesn't speak SSL.
In the case of using MySQL CLI to connect to an external MySQL DB, depending on the version of MySQL, you can pass the --ssl-mode=disabled like:
$ mysql --ssl-mode=disabled -h yourhost.tld -p
Or simply in your client config, for example in /etc/my.cnf.d/client.cnf:
[client]
ssl-mode=DISABLED
This is for dev and sometimes security and these things can be forfeited in certain situations in a closed, private dev environment.
I have a UCC SSL certificate which holds up to 10 domain names and when I browse to a page using the certificate I get an ERR_CONNECTION_REFUSED error.
Here's what I've done:
Purchased a UCC certificate
Created a Heroku SSL endpoint
Uploaded the certificate: heroku certs:add server.crt server.key
Checked the certs: heroku certs:info
Info returns:
Fetching SSL Endpoint blah-1234.herokussl.com info for app-name... done
Certificate details:
Common Name(s):domain.com
www.domain.com
others......
Expires At: 2015-05-25 23:48 UTC
Issuer: /OU=Domain Control Validated/CN=www.domain.com
Starts At: 2014-05-25 23:48 UTC
Subject: /OU=Domain Control Validated/CN=www.domain.com
SSL certificate is verified by a root authority.
heroku certs
Gives:
Endpoint Common Name(s) Expires Trusted
------------------------ ------------------------------------------------------------------------------------------- -------------------- -------
blah-1234.herokussl.com www.domain.com, domain.com, .......... 2015-05-25 23:48 UTC True
- Updated the domain to point to blah-1234.herokussl.com
CName:
www blah-1234.herokussl.com TTL1hr
Waited TTL
Run a curl test: curl -kvI https://www.domain.com
Response:
* About to connect() to www.domain.com port 443 (#0)
* Trying 50.19.XXX.XXX...
* Connection refused
* Trying 54.204.XXX.XXX...
* Connection refused
* Trying 23.21.XXX.XXX...
* Connection refused
* couldn't connect to host
* Closing connection #0
curl: (7) couldn't connect to host
The http urls work fine.
Any idea why this is happening? Is this because I'm using a UCC certificate?
The tech support guys at Heroku suggested removing and recreating the SSL end-point add-on and that solved the problem.
* Connection refused
These are no certificate errors, either your web server does not listen on port 443 or there is a firewall in between which is blocking connections.