Upstream prematurely closed connection while reading upstream #5706
Comments
|
Hi, I don't think this is a client bug. Most likely a configuration problem on the server side. |
|
Hi Ogoffart, Indeed. It was a server configuration. I just had to set the following on the Ngnix (frontend reverse proxy): Thanks for the help and for the links. |
|
@codesmaker Thanks. I added "proxy_buffering off;" which fixed a similar issue. Without this option, nginx throws the error message "upstream prematurely closed connection while reading upstream" and connection gets terminated. But.. I don't quite understand why adding this option would fix this. According to nginx documentation, buffering should automatically stop when it reaches the maximum buffer size (should be very small by default). Can someone explain what is going on inside Nginx? |
|
Had a similar problem with my reverse proxy. |
Expected behaviour
Owncloud client should successfully sync large files
Actual behaviour
Can not sync large files. Syncing stops as soon as a file reaches 1GB.
Server configuration
Operating system: Debian Jessie 8
Web server: Apache/2.4.10
Database: MySQL 5.5.54
PHP version: PHP 7.0.17-1~dotdeb+8.1
ownCloud version: 9.1.4
Storage backend (external storage):
Client configuration
Client version: 2.3.1
Operating system: Windows 10
OS language: English
Qt version used by client package (Linux only, see also Settings dialog):
Client package (From ownCloud or distro) (Linux only):
Installation path of client:
Logs
Please use Gist (https://gist.github.com/) or a similar code paster for longer
logs.
Web server error log:
2017/04/16 11:00:59 [error] 28461#28461: *2072 upstream timed out (110: Connection timed out) while reading response header from upstream, client: 192.168.1.1, server: owncloud.example.com, request: "GET /remote.php/webdav/NewArrivals/file HTTP/1.1", upstream: "http://192.168.1.204:80/remote.php/webdav/NewArrivals/file", host: "owncloud.example.com"
Server logfile: ownCloud log (data/owncloud.log):
{"reqId":"cYxtvnjVwfi7BJd9ks5q","remoteAddr":"192.168.1.1","app":"PHP","message":"Invalid argument supplied for foreach() at /var/www/owncloud/apps/activity/lib/Formatter/CloudIDFormatter.php#89","level":3,"time":"2017-04-16T12:50:50+02:00","method":"GET","url":"/ocs/v1.php/cloud/activity?page=0&pagesize=100&format=json","user":"user"}
{"reqId":"WcFGzfKmU0dyKXkVKyIY","remoteAddr":"192.168.1.1","app":"PHP","message":"fseek(): stream does not support seeking at /var/www/owncloud/apps/files_external/3rdparty/icewind/streams/src/Wrapper.php#74","level":3,"time":"2017-04-16T12:51:03+02:00","method":"GET","url":"/remote.php/webdav/NewArrivals/file","user":"user"}
Hi,
I have 2 Owncloud instances which are installed in two different countries. Both setups are identical. They are behind a reverse proxy (Nginx) that redirect the connections to the backend server. The owncloud installation is almost standard (Apache+MySQL+PHP7) as mentioned above. When I try to sync a large file (10GB+), the sync stops with an error "connection closed". I get the error message I mentioned above from the Nginx server. It is clear that the connection times out. I tried to increase the timout values in PHP (max execution time) and the reverse proxy but nothing helped. This is what I have in the reverse proxy:
location / { proxy_connect_timeout 159s; proxy_send_timeout 600; proxy_read_timeout 600; proxy_buffer_size 64k; proxy_buffers 16 32k; proxy_busy_buffers_size 64k; proxy_temp_file_write_size 64k; proxy_pass http://192.168.1.204; }It's also worth noting that I'm using federated shares. How can I track what exactly is timing out?
The text was updated successfully, but these errors were encountered: