[Judy]

In the ss-config we had the default value 32 32k and couldn’t remember why… well, after uncommenting the FastCGI buffers settings this week to allow users to properly customize the settings, we found out why.

It’s because if you are using a fastcgi_busy_buffers_size value of 256k which is recommended then it’s not going to be compatible with 16 16k

[Cheryl]

https://github.com/littlebizzy/slickstack/issues/179

[William]

There are lots and lots of threads and tutorials recommending 16 16k to avoid upstream errors in Nginx but these fail to take into account busy buffers eg.

https://stackoverflow.com/questions/23844761/upstream-sent-too-big-header-while-reading-response-header-from-upstream

https://stackoverflow.com/questions/19539501/nginx-502-bad-gateway-solved-by-increasing-buffer-why

[Susan]

one of those cases where people are copy and pasting the same (bad) settings in 1000 different tutorials for several years lol

[Jacqueline]

I’ve noticed that some threads suggest using smaller buffer sizes to avoid upstream errors. However, in busy situations, larger buffers like 32k might be necessary to prevent issues. It seems this is one of those cases where outdated settings are widely copied and pasted.

[Author]

Posts