Closed Bug 1310349 Opened 8 years ago Closed 8 years ago

Deploy the new CEP with the Hindsight Administration UI

Categories

(Cloud Services Graveyard :: Metrics: Pipeline, defect, P1)

defect

Tracking

(Not tracked)

RESOLVED FIXED

People

(Reporter: trink, Assigned: whd)

Details

      No description provided.
Proxy the Hindsight admin UI behind an Nginx instance (example config from hsadmin)

server {
        listen 80;
        server_name hsadmin.trink.com;
        return 301 https://$server_name$request_uri;
    }

server {
        listen 443 ssl;
        server_name hsadmin.trink.com;

    keepalive_requests    10;
    keepalive_timeout     60 60;
    ssl_certificate      /etc/letsencrypt/live/lpeg.trink.com/fullchain.pem;
    ssl_certificate_key  /etc/letsencrypt/live/lpeg.trink.com/privkey.pem;

    location /dashboard_output/ {
        alias  /mnt/work/hsadmin_dashboard/;
        autoindex on;
    }

    location / {
        proxy_pass http://localhost:2020;
        proxy_redirect http://localhost:2020/ /;
        proxy_buffering off;
        proxy_set_header X-Forwarded-For \$proxy_add_x_forwarded_for;
    }
}
Assignee: nobody → whd
The hindsight_admin_server.xml will need the following fields set

<property name="appRoot">/var/tmp</property>  # user database will lives here
<property name="hs_cfg">/work/hindsight.cfg</property> # location of the Hindsight cfg to connect this UI to

# Google creds
<property name="google-oauth2-client-id">*CONFIGURE_ME*.apps.googleusercontent.com</property>
<property name="google-oauth2-client-secret">*CONFIGURE_ME*</property>

Hindsight itself will require
- kafka input - to load the production data
- prune_input - to clean up the data queues (input and analysis)
- heka_inject_payload output - to populate the dashboard. The output_dir should match the /dashboard_output/ location in the Nginx config.
Points: --- → 2
Priority: -- → P1
This is implemented in https://github.com/mozilla-services/puppet-config/compare/df07b800a64e1ad120aedf5565d338282c97e36b...new_telemetry and https://github.com/mozilla-services/svcops/compare/7beda09f34fa5bcb50d45dc3518e4ee5e2260a3c...new_telemetry

All that's left I believe is to move the dev subset and test pilot jobs to the new CEP, which I will file a separate bug for.
Status: NEW → RESOLVED
Closed: 8 years ago
Resolution: --- → FIXED
Product: Cloud Services → Cloud Services Graveyard
You need to log in before you can comment on or make changes to this bug.