Increase nginx buffer size to avoid too big header errors

[thespad]

---

• I have read the contributing guideline and understand that I have made the correct modifications

---

Description:

Benefits of this PR and context:

How Has This Been Tested?

Source / References:

[LinuxServer-CI]

I am a bot, here are the test results for this PR:
https://ci-tests.linuxserver.io/lspipepr/speedtest-tracker/v1.7.2-pkg-0945974d-dev-197d76b64ff8ee9b724b711257a79981d0707eb1-pr-58/index.html
https://ci-tests.linuxserver.io/lspipepr/speedtest-tracker/v1.7.2-pkg-0945974d-dev-197d76b64ff8ee9b724b711257a79981d0707eb1-pr-58/shellcheck-result.xml

Tag	Passed
amd64-v1.7.2-pkg-0945974d-dev-197d76b64ff8ee9b724b711257a79981d0707eb1-pr-58	✅
arm64v8-v1.7.2-pkg-0945974d-dev-197d76b64ff8ee9b724b711257a79981d0707eb1-pr-58	✅

[nemchik]

I'm approving this with a side note that it gives me an itch to do some more nginx conf revamps so that smaller changes like this don't need to replace the whole default.conf (and maybe I'll scratch that itch at some point).

[jdevrie]

Wel this didn't work on my synology DSM 7.3.1. Now it does not give a 502 but I do not get a response at all from the docker container.

[jdevrie]

I tried this but it didn't fix the problem. The problem changed. I am no longer getting an error 502 but the whole thing would not react to the reverse proxy anymore. This works when I access the container directly on its mapped external port but fails when I try to access it through my reverse proxy. I can access the main page and Data Integration, Notifications and Thresholds pages through the proxy but as soon as I try to access any other page I get a result that the reverse proxy cannot handle.
I guess the header of the page produced is just to big for the reverse proxy on a Synology DSM to handle.

[thespad]

That is quite possible, and it might be an option to modify the proxy config, not sure what Synology exposes in that regard.

[jdevrie]

@thespad this is not good.The headers are huge. I cannot change the buffer size on the reverse proxy of a Synology DSM since that will be overwritten on boot time. The headers should be reduced to a normal size that a standard nginx configuration can handle otherwise thousands of users can not run this container on their Synology NAS with a reverse proxy.

I just activated the error log on my reverse proxy for the speedtest container and see indeed the same error as I first had on the nginx inside the container so my previous assumption was right. Hereby the message:

2025/11/11 14:49:43 [error] 25546#25546: *3167 upstream sent too big header while reading response header from upstream, client: 192.168.1.1, server: speedtest.jdvl.nl, request: "GET /admin/results HTTP/2.0", upstream: "http://127.0.0.1:54080/admin/results", host: "speedtest.jdvl.nl", referrer: "https://speedtest.jdvl.nl/admin/results"

[thespad]

It's not something we have the power to change, it needs to be discussed with the upstream project here alexjustesen/speedtest-tracker#2400

[drizuid]

It is also worth noting that systems with kernels and docker versions that have been unsupported for years are not generally going to be the target that things are designed for. Synology, unfortunately, is extremely irresponsible in terms of non-nas things. They are great nas units and bad for anything else because synology refuses to properly support them.

with that said, I will go ahead and lock this, but jdevrie, i would expect you will see this more and more often and technology moves forward and synology stays behind.