Character encoding issue in Transfer Protocol

[john-d13]

I use sniproxy for name based vitrual redirect for hundreds of hosts and for the most part it works properly. I am having issues with one hostname in particular - www.bbc.co.uk.

Here are the symptoms:

• Initial Query parses successfully
• Subsequent Queries fail with with HTML body error "Not found. Please check your DNS. Internal reference TA-1437"
• Wait 60s (TTL for www.bbc.co.uk) and it works
• Restart SNIPROXY daemon and it works before the 60s TTL expires
• Clear Cache on local web browser

Its worth noting, those workarounds will only work for one or 2 queries to the hostname, errors will repeat the cycle. Also on the sniproxy server, no DNS queries are being initiated to target recursor once entered into this error state.

A good start is to understand logically what data sniproxy holds on to during the DNS Cache period and what triggers a restart to fix the problem.

I have tried a myriad of configuration options without luck, including:

• Using a local resolver instead of public DNS (ie. google/cloudflare)
• Fresh install of SNIProxy (sniproxy 0.6.0+git.8.g3fa47ea)
• Tested on Multiple Virtual Machines
• Tested on FF Dev + Chrome browsers
• Enable Debug logging on sniproxy (no errors logged)

Here is an exceprt of the console error from FF Dev Browser

The character encoding of the HTML document was not declared. The document will render with garbled text in some browser configurations if the document contains characters from outside the US-ASCII range. The character encoding of the page must be declared in the document or in the transfer protocol.

[john-d13]

Creek creek

[bcacbe]

Have you tried the following solution. it fixed all my issues like you're describing. https://github.com/dlundquist/sniproxy/issues/255#issuecomment-521531583