Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Trafficserver has additional timeout not tied to configuration #441

Closed
science695 opened this issue Feb 19, 2019 · 3 comments
Closed

Trafficserver has additional timeout not tied to configuration #441

science695 opened this issue Feb 19, 2019 · 3 comments
Milestone

Comments

@science695
Copy link

science695 commented Feb 19, 2019

I have been trying to get an api long-running process (more than 15 minutes) to run through umbrella.

I have set all 3 nginx timeouts to 3600 seconds, but it keeps getting a gateway timeout at 900 (15minutes).

I found a setting in trafficserver/records.config:

CONFIG proxy.config.http.transaction_active_timeout_in INT 900

It seems like this setting shold also be tied to proxy_read_timeout like

config["trafficserver"]["_transaction_no_activity_timeout_out"] = config["nginx"]["proxy_read_timeout"]

Is there any danger of setting this in the api-umbrella.yml config file:

trafficserver:
  embedded_server_config:
    records:
      - "CONFIG proxy.config.http.transaction_active_timeout_in INT 3600"
      - "CONFIG proxy.config.http.connect_attempts_timeout INT 3600"
      - "CONFIG proxy.config.http.post_connect_attempts_timeout INT 3600"

or setting it to 0 so it doesn't introduce another timeout to the process?

@science695
Copy link
Author

science695 commented Mar 4, 2019

I found 2 more timeouts in trafficserver (they default to 1800 seconds or 30minutes).

CONFIG proxy.config.http.connect_attempts_timeout INT 3600
CONFIG proxy.config.http.post_connect_attempts_timeout INT 3600

(See edited main post)

So far I have not had any problems....

@science695
Copy link
Author

I noticed that there have been alot of improvements since the v0.14.4 release, I believe this problem was fixed in June of 2018.

Is there any new release so that we could more easily use your updates?

@GUI GUI added this to the v0.15.0 milestone Apr 29, 2019
@GUI
Copy link
Member

GUI commented May 24, 2019

@science695: Sorry for the delay, but thanks for the heads up! Yes, you're right that this was fixed earlier and is now part of the v0.15.1 release. This should now be configurable via /etc/api-umbrella/api-umbrella.yml with the following settings:

nginx:
  proxy_connect_timeout: 60
  proxy_read_timeout: 60
  proxy_send_timeout: 60

The documentation for these timeout settings can be found for the underlying nginx settings here: http://nginx.org/en/docs/http/ngx_http_proxy_module.html

What's fixed in v0.15.1 is that configuring these nginx settings in API Umbrella should now automatically tune the underlying TrafficServer settings to match.

@GUI GUI closed this as completed May 24, 2019
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants