enhancement/bump-resources-on-dev

[bwalsh]

Based on my search through the helm/revproxy folder, here are the configurations that could produce a 502 Bad Gateway error:

Timeout Configurations

Both nginx.conf and nginxPrivate/nginx.conf have identical timeout settings that could lead to 502 errors:

proxy_buffer_size          16k;
proxy_buffers              8 16k;
proxy_busy_buffers_size    32k;
client_body_buffer_size    16k;
proxy_read_timeout         400;
proxy_send_timeout         400;
proxy_connect_timeout      400;

Key timeout values:

• proxy_read_timeout: 400 seconds - If backend doesn't respond within 400s, returns 502
• proxy_send_timeout: 400 seconds - If backend doesn't accept data within 400s, returns 502
• proxy_connect_timeout: 400 seconds - If connection to backend fails within 400s, returns 502
• keepalive_timeout: 65 seconds (line 36)

Service-Specific Extended Timeouts

Some services override the default 400s with longer timeouts (600s):

Guppy Service

proxy_connect_timeout 600s;
proxy_send_timeout 600s;
proxy_read_timeout 600s;
send_timeout 600s;

Sheepdog Service (API)

proxy_connect_timeout 300;
proxy_send_timeout 300;
proxy_read_timeout 300;
send_timeout 300;

Peregrine Service (GraphQL)

proxy_connect_timeout 300;
proxy_send_timeout 300;
proxy_read_timeout 300;
send_timeout 300;

Workspace Services (Ambassador/Hatchery)

# for fixing noVNC connection timeout issue
proxy_read_timeout 36000s;

Rate Limiting

I didn't find explicit rate limiting directives (like limit_req_zone or limit_conn_zone) in the revproxy configuration, which could also cause 502-like errors when backends are overwhelmed.

Other 502 Triggers

1. Buffer size limits - If headers/responses exceed buffer sizes:

   • proxy_buffer_size: 16k
   • large_client_header_buffers: 4 64k
   • client_header_buffer_size: 4k

2. Backend service unavailability - When upstream services are down or unreachable

3. Error page handling:

error_page 500 501 502 503 504 @5xx;

location @5xx {
    internal;
    return 500 "{ \"error\": \"service failure - try again later\"}";
}

---

Note: These results may be incomplete. For a comprehensive search, you can view more results in the GitHub code search UI.