Re: HTTP2 DATA frames with 0 length

2017-09-01 Thread Muhui Jiang
END_STREAM flag. Do you have any suggestions or could you please propose a configuration that NGINX will only return one DATA frame for one stream. Many thanks Regards Muhui 2017-09-02 4:32 GMT+08:00 Valentin V. Bartenev : > On пятница, 1 сентября 2017 г. 19:59:04 MSK Muhui Jiang wrote: >

HTTP2 DATA frames with 0 length

2017-09-01 Thread Muhui Jiang
Hi I am using nginx 1.9.15. I noticed when I made a HTTP2 request to the nginx. It will send the data frames that carry the object first. But end with a data frame whose length is zero indicating the Data end flag. I am curious why you guys design in this way. I think we don't need this extra dat

Re: different Memory consumption for H1 and H2

2017-05-12 Thread Muhui Jiang
05-11 23:11 GMT+08:00 Maxim Dounin : > Hello! > > On Thu, May 11, 2017 at 10:32:41PM +0800, Muhui Jiang wrote: > > > Recently, I did an experiment to test the memory consumption of nginx. I > > request a large static zip file. I explored the debug information of > nginx. &g

different Memory consumption for H1 and H2

2017-05-11 Thread Muhui Jiang
Hi Recently, I did an experiment to test the memory consumption of nginx. I request a large static zip file. I explored the debug information of nginx. For H2, below is a part of the log, I noticed that every time server will allocate 65536 bytes, I increase the connection number, I noticed that

nginx input

2016-09-07 Thread Muhui Jiang
Hi I am using program analysis to locate the bottleneck of nginx. I know the file nginx under the directory of nginx/sbin is the binary file. My question is that what is the input of the binary. I mean the format. Since a general URL doesn't seem to be a right input. Many Thanks Regards Muhui ___

Re: HTTP2 window update and priority

2016-05-10 Thread Muhui Jiang
request queue order(Here I noticed the request queue is the same as the receiving order) rather than handling priority when generating data frames, it would be too late. So that even the files is small, the response order will also follows the dependency tree. Best Regards Muhui Jiang 2016-05-10

Re: HTTP2 window update and priority

2016-05-10 Thread Muhui Jiang
observed. But I have give enough time to Nginx by block the request. Nginx should handle the priority though the file is really small. If you need more information, please tell me. Many Thanks Best Regards Muhui Jiang 2016-05-10 21:15 GMT+08:00 Valentin V. Bartenev : > On Monday 09 May 2016 21:27

HTTP2 window update and priority

2016-05-09 Thread Muhui Jiang
ent the priority. Could you please give me some suggestions. A failed priority mechanism example debug log: https://github.com/valour01/LOG/blob/master/example.log Best Regards Muhui Jiang ___ nginx mailing list nginx@nginx.org http://mailman.nginx.org/

Re: HTTP2 Multiplexing

2016-05-05 Thread Muhui Jiang
Hi Thanks Valentin. I've know the principle Best Regards Muhui Jiang 2016-05-04 19:18 GMT+08:00 Valentin V. Bartenev : > On Wednesday 04 May 2016 18:50:44 Muhui Jiang wrote: > > Hi > > > > >Nginx allows multiple request and responses in multiple connections > us

Re: HTTP2 Multiplexing

2016-05-04 Thread Muhui Jiang
mprovement of the implementation on the feature of multiplexing compared with Http/1.1 pipeline. How do you solve the problem Head-of-line blocking occurred in Http/1.1. Best Regards Muhui Jiang 2016-05-04 18:19 GMT+08:00 Valentin V. Bartenev : > On Wednesday 04 May 2016 11:25:11 Muhui

HTTP2 Multiplexing

2016-05-03 Thread Muhui Jiang
requests handling. If you can locate the correspond code for me, that would be great Best Regards Muhui Jiang ___ nginx mailing list nginx@nginx.org http://mailman.nginx.org/mailman/listinfo/nginx

Re: HTTP 400 responses

2015-10-21 Thread Muhui Jiang
>>client prematurely closed connection that means your client has closed the socket. You can rebuilt the nginx to enable the detailed logging for debug http://nginx.org/en/docs/debugging_log.html 2015-10-21 15:52 GMT+08:00 Pauls V : > Had tried with reload and that didn't change output. Tried re

Re: Data frame size sent by server

2015-10-19 Thread Muhui Jiang
As I am researching on the flow control strategy of HTTP/2 recently. The internal buffers will show me something about it. Thanks Valentin. Best Regards Muhui Jiang 2015-10-20 0:55 GMT+08:00 Valentin V. Bartenev : > On Sunday 18 October 2015 16:02:04 Muhui Jiang wrote: > > Hi > &g

Data frame size sent by server

2015-10-18 Thread Muhui Jiang
frame sent: 7F942205FBC0 sid:1 bl:0 len:8182 Many Thanks Best Regards Muhui Jiang ___ nginx mailing list nginx@nginx.org http://mailman.nginx.org/mailman/listinfo/nginx

Re: Http2 Priority

2015-10-13 Thread Muhui Jiang
Hi Thank you very much. Valentin Best Regards Muhui Jiang 2015-10-13 23:16 GMT+08:00 Valentin V. Bartenev : > On Tuesday 13 October 2015 22:43:48 Muhui Jiang wrote: > > Hi > > > > Thank you so much. Valentin. > > My understanding: > > Though clinklittle128.pn

Re: Http2 Priority

2015-10-13 Thread Muhui Jiang
Best Regards Muhui Jiang 2015-10-13 22:14 GMT+08:00 Valentin V. Bartenev : > On Tuesday 13 October 2015 16:05:16 Muhui Jiang wrote: > > Hi > > > > move the debug to the main level; And the error.log below: a little bit > long. Thanks > > > [..] > > 2015/10/13 15

Re: Http2 Priority

2015-10-12 Thread Muhui Jiang
this comes. Hope you could understand my statement. I also think this problem is related to flow control or buffer. It would be better if you can locate the code related to flow control and buffer for me. Many thanks. Best Regards Muhui Jiang 2015-10-12 21:03 GMT+08:00 Valentin V. Bartenev :

Http2 Priority

2015-10-09 Thread Muhui Jiang
nother server h2o, it gives me the result I want. A finished first and then B,C,D. Best Regards Muhui Jiang ___ nginx mailing list nginx@nginx.org http://mailman.nginx.org/mailman/listinfo/nginx