Page MenuHomeMiraheze

Deploy new infrastructure
Closed, ResolvedPublic

Description

  • Rackup and install equipment (Dec 29th) T8470
    • Setup ProxMox, with networking and generate all VMs as agreed by resource allocation (Dec 29th onwards, potentially while on site?)
  • Initial install steps - NON TRAFFIC
    • Install and deploy bastion hosts T8471
    • Install and deploy db hosts
    • Install and deploy es cluster T8472
    • Install and deploy gluster cluster
    • Install and deploy jobchron T8473
    • Install and deploy mem T8474
    • Install and deploy mw
    • Install and deploy phab
  • Transition Steps - TRAFFIC - non critical
    • Install and deploy graylog T8475
    • Install and deploy ldap T8476
    • Install and deploy mail T8477
    • Install and deploy mon T8478
    • Install and deploy mwtask T8479
    • Install and deploy test T8480
  • Transition Steps - TRAFFIC - critical
    • Migrate databases over to new infrastructure T8481
    • Migrate gluster over to new infrastructure T8482
    • Migrate mw traffic over to new infrastructure T8483
    • Install and deploy ns T8484
    • Install and deploy puppet T8485

Related Objects

Event Timeline

John triaged this task as Normal priority.Dec 18 2021, 18:43
John created this task.

I will create all subtasks as labelled above and update that.

Tagged this as Notice for @Agent_Isai to pick up as I think a general notice to the community that there will be periods of read-only and reduced services (grafana/static, etc.) potentially coming up in January. More details will be known in the next few weeks (at most I would say 9th of Jan?) and periods of reduce service will receive at least 1 weeks notice to users.

First part is partially done - I'm going to put final testing in tomorrow for networking setup and security harden the more valuable access points into the new cluster.

I also want to keep the servers up and running for a full 24 hour period to see how they bed into their new environment and monitor base level power usage. Further updates on this tomorrow.

Hi @John,

What's the plan to handover MediaWiki (SRE) servers to them?

I'll be doing most of the setup but with no @Reception123 I'll need someone to merge my puppet PRs.

We obviously need puppet (T8485) online first but my understanding is that it must be as gluster is being configured.

Hi @John,

What's the plan to handover MediaWiki (SRE) servers to them?

I'll be doing most of the setup but with no @Reception123 I'll need someone to merge my puppet PRs.

We obviously need puppet (T8485) online first but my understanding is that it must be as gluster is being configured.

I'll link in with Paladox later to understand what they have done and where they are up to. From there, I provision all the MediaWiki (SRE) servers with puppet and then merge the relevant puppet PRs as/when required/possible.

Thank you. Please ping me when doing so. I'll be taking lead from our end.

Status: MediaWiki is running as much as it can be without any databases.

I will be setting up jobs tonight and hopefully connecting the new servers to cache proxies but not in the main backend.

Phabricator has now been migrated.

John updated the task description. (Show Details)

Ok, we are all completed, now, let's go for the "stalled" one. (T8179)