Closed
Bug 1399162
Opened 8 years ago
Closed 7 years ago
Move aggregates service to ops purview (deploy prod)
Categories
(Data Platform and Tools :: General, enhancement, P1)
Data Platform and Tools
General
Tracking
(Not tracked)
RESOLVED
FIXED
People
(Reporter: frank, Assigned: hwoo)
Details
(Whiteboard: [DataOps])
No description provided.
Reporter | ||
Updated•8 years ago
|
Priority: -- → P2
Assignee | ||
Updated•8 years ago
|
Assignee: nobody → hwoo
Updated•7 years ago
|
Whiteboard: [SvcOps] → [DataOps]
Assignee | ||
Updated•7 years ago
|
Priority: P2 → P3
Assignee | ||
Comment 1•7 years ago
|
||
Just some notes for self:
- rerun deployment to stage with master branch of python_mozaggregator to verify everything is still working properly
- verify deployment of prod service wont take CNAME of current running service (app_dns_name: "{{ stack }}.{{ dns_zone }}" in projects/mozaggregator/ansible/envs/default.yml?)
- push deployment to prod
- change rds password of temp db in prod manually to whatever is in hiera
- test service functionality (may need to take db snapshot and use stale data. try importing snapshot from staging)
- schedule time to do cutover and stop ingestion to old prod service's db
- make backup of prod RDS
- make changes to cloudops-deployment to reference prod-rds (env vars set in hiera sops?). This may require a redeploy of prod to pull new secrets
- test elb endpoint and service again
- move cname of current service to new deployment
- test
- start ingestion of data to new prod db (iam perms required?)
old bug -
https://bugzilla.mozilla.org/show_bug.cgi?id=1342947
old prs
https://github.com/mozilla/python_mozaggregator/pull/63
https://github.com/mozilla-services/cloudops-deployment/pull/1564
https://github.com/mozilla/python_mozaggregator/pull/50
open issues
- add newrelic-admin entrypoint?
Assignee | ||
Updated•7 years ago
|
Priority: P3 → P2
Summary: Move aggregates service to ops purview → Move aggregates service to ops purview (deploy prod)
Assignee | ||
Comment 2•7 years ago
|
||
- front service with openresty/nginx
- investigate whether or not we can use the same prod database as currently in use (this will simplify the migration greatly)
Assignee | ||
Updated•7 years ago
|
Priority: P2 → P1
Assignee | ||
Comment 3•7 years ago
|
||
x - rerun mozaggregator-env jenkins job for stage|prod
x - ensure elb listener ports change (for nginx change)
x - manually repoint rds dns cnames to telemetry-aggregates rds endpoints (for prod point this to the existing dev dbs)
x - modify hiera-sops so that mozaggregator points to the production cnames (dbs in dev) with correct password
x - log on and ensure that nginx/openresty is installed
x - continue mozaggregator jenkins job for prod and redo steps
- frank to test functionality of prod endpiont
- ask ckolos or it to modify cname aggregates.telemetry.mozilla.org -> mozaggregator.prod.mozaws.net...or the ELB endpoint directly
---> ask ckolos what his opinion is on which is better. If we use ELB directly we need to modify the SSL cert on the elb to use tmo (da1ea4a6-bcac-43aa-b3b9-84b4912f63de)
projects/mozaggregator/ansible/envs/prod.yml:
elb_cert_arn: arn:aws:acm:us-west-2:361527076523:certificate/da1ea4a6-bcac-43aa-b3b9-84b4912f63de
Assignee | ||
Updated•7 years ago
|
Status: NEW → RESOLVED
Closed: 7 years ago
Resolution: --- → FIXED
Updated•3 years ago
|
Component: Telemetry Aggregation Service → General
You need to log in
before you can comment on or make changes to this bug.
Description
•