I have an API server (Debian Apache2) with OpenSSL installed and working. I also have a staging and production web server (also Debian Apache2, exactly the same spec - they are VM clones). All servers are on the same subnet. I can browse to the wsdl from my local machine on 443 successfully, and I can wget
the wsdl successfully from my staging server on 443, but a wget
from my production web server will not connect:
--2015-04-16 10:26:18-- https://www.example.com/index.php/api?wsdl
Resolving https://www.example.com (https://www.example.com)... XX.XX.XX.XX
Connecting to https://www.example.com (https://www.example.com)|XX.XX.XX.XX|:443... failed: Connection refused.
I can connect over https from a PHP nusoap client on staging no problem, but the same code on my production server returns:
wsdl error: HTTP ERROR: cURL ERROR: 7: couldn't connect to host
url: https://www.example.com/index.php/api?wsdl
content_type:
http_code: 0
header_size: 0
request_size: 0
filetime: -1
ssl_verify_result: 0
redirect_count: 0
total_time: 5.272228
namelookup_time: 5.271805
connect_time: 0
pretransfer_time: 0
size_upload: 0
size_download: 0
speed_download: 0
speed_upload: 0
download_content_length: -1
upload_content_length: -1
starttransfer_time: 0
redirect_time: 0
certinfo: Array
primary_ip: XX.XX.XX.XX
primary_port: 443
local_ip:
local_port: 0
redirect_url:
An openssl s_client -connect
from both web servers produces the same output.
After my production server returns connection refused, there are no new entries in the API server's error.log, therefore this must be a client issue.
Is there a Debian-specific/internal firewall config I may have inadvertently changed that would prevent the one client from connecting to a secure web server over HTTPS and not another?