Re: QUIC with a fcgi backend

2023-07-28 Thread Yaacov Akiba Slama
On 25/07/2023 12:04, Christopher Faulet wrote: With more intensive tests, I'm able to reproduce it time to time. It is timing dependent. It happens when the last 2 DATA frames are decodes in same time. The end of the request is detected to early, on the first DATA frame instead of the last

Re: QUIC with a fcgi backend

2023-07-25 Thread Christopher Faulet
Le 7/24/23 à 19:21, Yaacov Akiba Slama a écrit : On 24/07/2023 19:02, Aleksandar Lazic wrote: Are the HAProxy and the FCGI Server on the same host/network or is there any firewall or anything in between? Both run in the same host. In order to reproduce the problem I had with php-fpm, I am

Re: QUIC with a fcgi backend

2023-07-24 Thread Yaacov Akiba Slama
On 24/07/2023 19:02, Aleksandar Lazic wrote: Are the HAProxy and the FCGI Server on the same host/network or is there any firewall or anything in between? Both run in the same host. In order to reproduce the problem I had with php-fpm, I am using a simple go fcgi server: package main

Re: QUIC with a fcgi backend

2023-07-24 Thread Aleksandar Lazic
Yaacov. On 2023-07-24 (Mo.) 15:08, Christopher Faulet wrote: Le 7/24/23 à 12:24, Yaacov Akiba Slama a écrit : Hi Christopher, Thanks for report. It is not a known issue, but I can confirm it. When H3 HEADERS frames are converted to the internal HTTP representation (HTX), a flag is missing to

Re: QUIC with a fcgi backend

2023-07-24 Thread Christopher Faulet
Le 7/24/23 à 12:24, Yaacov Akiba Slama a écrit : Hi Christopher, Thanks for report. It is not a known issue, but I can confirm it. When H3 HEADERS frames are converted to the internal HTTP representation (HTX), a flag is missing to specify a content-length was found. I pushed a flag, it should

Re: QUIC with a fcgi backend

2023-07-24 Thread Yaacov Akiba Slama
Hi Christopher, Thanks for report. It is not a known issue, but I can confirm it. When H3 HEADERS frames are converted to the internal HTTP representation (HTX), a flag is missing to specify a content-length was found. I pushed a flag, it should be fixed: commit

Re: QUIC with a fcgi backend

2023-07-24 Thread Christopher Faulet
Le 7/22/23 à 21:48, Yaacov Akiba Slama a écrit : Hi, It seems that there is a bug in QUIC when using a fastcgi backend: As soon as the size of the uploaded data is more than bufsize, the server returns 400 Bad request and shows PH-- in the logs. The problem occurs with both haproxy 2.8.1 and

Re: QUIC with a fcgi backend

2023-07-22 Thread Yaacov Akiba Slama
Sending the config file as an attachment. -yas On 23/07/2023 07:20, Yaacov Akiba Slama wrote: On 22/07/2023 23:07, Aleksandar Lazic wrote: Hi. On 2023-07-22 (Sa.) 21:48, Yaacov Akiba Slama wrote: Hi, It seems that there is a bug in QUIC when using a fastcgi backend: As soon as the size of

Re: QUIC with a fcgi backend

2023-07-22 Thread Yaacov Akiba Slama
On 22/07/2023 23:07, Aleksandar Lazic wrote: Hi. On 2023-07-22 (Sa.) 21:48, Yaacov Akiba Slama wrote: Hi, It seems that there is a bug in QUIC when using a fastcgi backend: As soon as the size of the uploaded data is more than bufsize, the server returns 400 Bad request and shows PH-- in

Re: QUIC with a fcgi backend

2023-07-22 Thread Aleksandar Lazic
Hi. On 2023-07-22 (Sa.) 21:48, Yaacov Akiba Slama wrote: Hi, It seems that there is a bug in QUIC when using a fastcgi backend: As soon as the size of the uploaded data is more than bufsize, the server returns 400 Bad request and shows PH-- in the logs. The problem occurs with both haproxy

QUIC with a fcgi backend

2023-07-22 Thread Yaacov Akiba Slama
Hi, It seems that there is a bug in QUIC when using a fastcgi backend: As soon as the size of the uploaded data is more than bufsize, the server returns 400 Bad request and shows PH-- in the logs. The problem occurs with both haproxy 2.8.1 and 2.9-dev2 (both build quictls