nginx client_body_buffer_size increase

[Jörg Wiesmann]

Hi @ all,

i was wondering if somebody could help me out. I am running a node server on an app engine (flex) and i see the following warning in the log files:

textPayload:  "[warn] 32#32: *2096 a client request body is buffered to a temporary file 

I tried to understand what that means and i figured out, that google puts a nginx server as a proxy to all flex app engine applications. Since i never used nginx so far and i could not find any info @ the google documentation i am asking you guys, how do i change the value for

client_body_buffer_size

for the app engine? (or how can i improve the current situation and get rid of the warning message?)

i also found this: 
https://github.com/google/appengine-jenkins/tree/master/image/master_images with an example nginx.conf. Since i only want to change one single value (maybe 2) i am trying to find out what the best way could be. Can i just provide a nginx.conf with 2 values and it will only override this two values provided by default? Or do i have to create an own nginx.conf with all possible settings? 

Thank you for your help!

[Olu]

I found this Serverfault link that confirms that this warning is actually an Nginx message which is sent to indicate that the size of your uploaded file was larger than the in-memory buffer reserved for uploads.

About using the Nginx.config file provided on the link you shared, the link shared is for "a package that provides a Docker image running Jenkins behind NGINX". It is unclear how it directly applies to your setup. Can you provide more information about your application in the AppEngine Flex environment?

[Jörg Wiesmann]

Hi Olu,

thank you for your reply! My current setup is actually quite simple. It is "just" a node application on the flex environment. I did not setup any nginx config whatsoever. I am using a pretty straight forward app.yaml and i deploy my application with the command 

gcloud app deploy

app.yaml

runtime: nodejs
env: flex

automatic_scaling:
  min_num_instances: 2
  max_num_instances: 5
resources:
  cpu: 2
  memory_gb: 2.3
  disk_size_gb: 10
  volumes:
  - name: ramdisk1
    volume_type: tmpfs
    size_gb: 0.5

in my package.json i additionally had to add

"engines": {
  "node": "8.9.4"
}

to tell google to use this node version.

That is basically all i did :)

[George (Cloud Platform Support)]

Hello Jörg, 

This is a warning, not an actual error. It is known to Developers, but there are no definite plans to address it in the near future. If you feel this is of importance to you, and a solution will address an actual problem you are affected by, you are encouraged to open a new issue in the Public Issue Tracker. A advantage of opening an issue there is traceability and certitude that Developers will have a look at it, evaluate and fix it in accordance with merit.