#712 [spike] Duffy staging and prod environment automation
Closed: Fixed with Explanation 2 years ago by arrfab. Opened 2 years ago by siddharthvipul1.

Duffy development is complete. We have tested all open nebula functionality and need the env to deploy it. Please prepare a production deployment node that’s

  • 8 stream (or above)
  • Able to talk to seamicro chassis
  • Able to talk to open nebula cluster
  • Has networking, images, initrds, pxe configs etc in place to be able to install baremetal nodes using the existing CentOS Infra playbook

Metadata Update from @zlopez:
- Issue tagged with: blocked, centos-ci-infra

2 years ago

hey folks, do we have any updates on this?
we wanted to test final deployment, can this be done in a day or 2 (by Tuesday end)?
Otherwise, we will go with "knowledge transfer" to an infra and releng member so that can be taken on your own time?

@siddharthvipul1 Has this been deployed in staging

Metadata Update from @arrfab:
- Issue tagged with: doc, high-gain, medium-trouble

2 years ago

Metadata Update from @arrfab:
- Issue assigned to arrfab

2 years ago

Per meeting with @amoloney, and thenkick-off meeting around this, it was agreed that :

  • it's still blocked for production deployment
  • we can work on needed documentation for deployment steps
  • work on ansible-role-duffy to a point where we can safely run it to deploy a fresh instance
  • do some validation tests :
    • bare-metal deployment
    • opennebula VM deployment
    • migration script ran to export/import from old to new duffy
    • compatibily for existing python-cicoclient and tenants is working

Some meetings with @phsmoura , @nphilipp are now scheduled to work on this

Metadata Update from @zlopez:
- Issue priority set to: High Priority (was: Needs Review)

2 years ago

Metadata Update from @zlopez:
- Issue untagged with: blocked

2 years ago

Status update :

  • ansible role is more or less complete (in staging branch)
  • bare-metal deployment works (but just needs a little bit of tuning for parallelism but WIP)
  • opennebula VM deployment fully works
  • minimum compatibility test was done for legacy API endpoint

Metadata Update from @zlopez:
- Issue priority set to: Medium Priority (was: High Priority)

2 years ago

Quick status update :

  • role should be now ready
  • we're exploring other options for EC2 to slowly replace bare-metal/on premises infra
  • haproxy role tested to protect api call behind https (using duffy api as backend server)
  • quotas and session-lifetime are now also "per tenants"

Last update on this ticket:

  • we can now provision ec2 instances, including metal options (for x86_64 and aarch64)
  • a tunnel was created between AWS (dedicated ci VPC) and is operational
  • new pool.ci.centos.org dns zone created (on route53) and updated by ansible-playbook provisioner
  • we have a roadmap about communication plan for ci tenants

based on discussion in the [backlog refinement] call, it was decided that this part is now done and can be closed as it's now part of a wider cpe initiative but all technical work is (almost)done and so just waiting to be announced and be worked on

Metadata Update from @arrfab:
- Issue close_status updated to: Fixed with Explanation
- Issue status updated to: Closed (was: Open)

2 years ago

Login to comment on this ticket.

Metadata
Boards 1
CentOS CI Infra Status: Backlog