server: bump timeout to 3600s

[ngxson]

Overview

IMPORTANT: server's --timeout works fine. For users who reported the problem related to timeout, check your client code first, some HTTP framework and browsers may have a default client-side timeout.

Ref discussion from #22907

Fix #23832

Fix #22997

Bump timeout to one hour. This "ought to be enough for anybody"

Also print a message to remind about client's timeout.

How I tested this change

Here is how I test it:

1. add this to either before or after llama_decode() in server-context.cpp: std::this_thread::sleep_for(std::chrono::seconds(1000000));
   that will simulate a long blocking task
2. remember to configure timeout of the client. in my case, postman: https://stackoverflow.com/questions/36355732/how-to-increase-postman-client-request-timeout
3. send a request and wait
4. 17 minutes later, I stop the request:

0.36.243.889 I srv          load:  - looking for better prompt, base f_keep = -1.000, sim = 0.000
0.36.243.895 I srv        update:  - cache state: 0 prompts, 0.000 MiB (limits: 8192.000 MiB, 131072 tokens, 8589934592 est)
0.36.243.899 I srv  get_availabl: prompt cache update took 0.04 ms
0.36.244.264 I slot launch_slot_: id  3 | task 0 | processing task, is_child = 0
17.11.781.940 W srv          next: request cancelled after 30s, likely a client-side timeout; please check your client's code
17.11.781.948 W srv          stop: cancel task, id_task = 0

Requirements

• I have read and agree with the contributing guidelines
• AI usage disclosure: nope

[ngxson]

note: it would be better to detect if time_elapsed_ms > server's --timeout here, then log another message. but due to the way things are structured, this proved to be quite complicated

[sasa7812]

Would the client-set timeout for request be easier to log? It would have avoided the confusion

[ngxson]

but how? AFAIK client never communicate such info to server

[bonswouar]

I might misunderstand the PR but isn't the log request cancelled after 30s a bit confusing?
It could show this message with a time_elapsed_ms anywhere between 30s & 3600s, but for an user reading this log it looks like it's exactly 30s.
Why not use the actual time_elapsed_ms value?

[sasa7812]

but how? AFAIK client never communicate such info to server

You are right, I should have checked before asking. The whole log messaging was just implying that request is cancelled on server initiative.