Dear Adi, jeihii, that solved my problem. Sorry guys, could have solved it by myself reading the docs.
Thanks a lot! Carsten -----Ursprüngliche Nachricht----- Von: [email protected] <[email protected]> Im Auftrag von Adi Kriegisch Gesendet: Freitag, 6. Mai 2022 14:22 An: [email protected] Betreff: Re: [SOGo] problems with 5.6.0 Dear Carsten, > I thought it is a good idea to update to 5.6.0 to get rid of it, but > anyhow, no login at all is possible any longer. > > After login of any user I get Unbehandelte Fehlerantwort unhandled > error response. I assume, you're using nginx. As I've already written[1], the error message is a consequence of bigger header data. You'll need to update the buffer values in your nginx config: proxy_buffer_size 8k; proxy_buffers 4 64k; To find the error message, just check your nginx error log. In my case, the following got logged to sogo.log: [ERROR] <0xXXXXX[WOHttpTransaction]> client disconnected during delivery of response for <WORequest[0xXXXXX]: method=POST uri=/SOGo/connect app=SOGo rqKey=connect rqPath=(null)> (len=50): the socket was shutdown and nginx' error log: *123456 upstream sent too big header while reading response header from upstream, client: ip.add.re.ss, server: sogo.example.com, request: "POST /SOGo/connect HTTP/1.1", upstream: "http://127.0.0.1:20000/SOGo/connect";, host: "sogo.example.com", referrer: https://sogo.example.com/SOGo/"; After increasing the buffer size and reloading nginx it worked again. (btw. do not forget to upgrade your database scheme as this is required for 5.5.1 -> 5.6.0 upgrade). -- Adi [1] https://www.mail-archive.com/users%40sogo.nu/msg31359.html -- [email protected] https://inverse.ca/sogo/lists -- [email protected] https://inverse.ca/sogo/lists
