Custom Query (2296 matches)

Filters
 
Or
 
  
 
Columns

Show under each result:


Results (1 - 3 of 2296)

1 2 3 4 5 6 7 8 9 10 11
Ticket Resolution Summary Owner Reporter
#785 SPDY not supporting domain sharding steve.exley.mayden.co.uk@…
Description

When using multiple domains on the same ip & ssl certificate. Nginx does not support the client switching domains on the same SPDY session. Tested in 1.8.0

Server block looks like:

server {
    listen       443 ssl spdy;
    server_name  *.domain.com;
    ssl_certificate wildcarddomain.crt;
    ...

If the client visits a.domain.com and then b.domain.com on the same IP using spdy, they get a HTTP 400 response.

The log file records:

"client attempted to request the server name different from that one was negotiated while processing SPDY"

When using ssl instead of spdy. There is no issue.

The error message comes from ngx_http_request.c, the code in 1.9.4 appears the same so this issue may persist with the switch to HTTP2.

#808 Defining http2 without ssl leads to HTTP/1.1 client failure LPardue@…
Description

This issue probably relates to support of HTTP/2 over cleartext (h2c). Note that I have successfully tested HTTP/2 over TLS on this server.

I have a simple server setup using an out of the box config, if I use a HTTP/1.1 client then I see what I would expect e.g:

curl -v http://localhost/index.html
* About to connect() to localhost port 80 (#0)
*   Trying ::1...
* Connection refused
*   Trying 127.0.0.1...
* Connected to localhost (127.0.0.1) port 80 (#0)
> GET /index.html HTTP/1.1
> User-Agent: curl/7.29.0
> Host: localhost
> Accept: */*
>
< HTTP/1.1 200 OK
< Server: nginx/1.9.5
< Date: Fri, 09 Oct 2015 15:53:40 GMT
< Content-Type: text/html
< Content-Length: 612
< Last-Modified: Wed, 07 Oct 2015 10:41:07 GMT
< Connection: keep-alive
< ETag: "5614f6c3-264"
< Accept-Ranges: bytes
<
<!DOCTYPE html>
...
* Connection #0 to host localhost left intact

If I add http2 to the config, e.g.

server {
  listen 80 http2;
  server_name localhost;
...

Then I see this

curl -v http://localhost/index.html

* About to connect() to localhost port 80 (#0)
*   Trying ::1...
* Connection refused
*   Trying 127.0.0.1...
* Connected to localhost (127.0.0.1) port 80 (#0)
> GET /index.html HTTP/1.1
> User-Agent: curl/7.29.0
> Host: localhost
> Accept: */*
>
* Connection #0 to host localhost left intact

Nothing is reported in the server logs.

So my questions are:

  • Is h2c supported?
    • If not, why am I allowed to attempt such a configuration without some error/warnings.
    • If so, what am I doing incorrectly? Shouldn't a HTTP/1.1 client continue to work if it offers no upgrade header or "prior knowledge" magic?
#8 worksforme ignoring stale global SSL error somebody mikerogerz.myopenid.com
Description
[alert] 14079#0: *5910589 ignoring stale global SSL error (SSL: error:140940E5:SSL routines:SSL3_READ_BYTES:ssl handshake failure) while closing request, client: 115.184.236.136, server: 0.0.0.0:443
1 2 3 4 5 6 7 8 9 10 11
Batch Modify
Note: See TracBatchModify for help on using batch modify.
Note: See TracQuery for help on using queries.