Skip to content

Epic: Migrate to New Server

Tom Elliott edited this page Oct 9, 2020 · 8 revisions

Tummy.com is provisioning a new server as our current one is running Ubuntu 16.04 LTS, which reaches Maintenance Update EOL in April 2021. Since they normally like to move to new hardware when making an EOL-induced OS upgrade, we have contracted with Tummy to switch to a new dedicated server with the specs listed below. This Epic is intended to group all related stories so that we can migrate to the new server sometime after it becomes available in December 2020.

Stories and tasks:

Label is epic: migrate

For planning purposes, here is the task list we used the last time we did a server migration, with a few additional items added. These will need to be turned into separate stories as appropriate:

  • review any ZMI-based customizations/overrides and move to code; test, release and deploy to legacy server
  • review code on production server to find any cases where a bugfix or feature enhancement branch has been checked out; inspect corresponding repos and perform merge/PR/release and deploy to legacy server as necessary
  • #435: Write and test code to expire daily backups
  • verify developer logins and sudo on the new box; get tummy help to fix as necessary
  • install and verify prerequisites for running our ansible setup; update pleiades.ansible documentation to reflect these steps
  • identify components, products, and other aspects that need upgrade or change in order to run under Ubuntu 20.04 LTS; make and test changes accordingly but do not push to production
  • run and verify pleiades.ansible to deploy the new full server (iterate as necessary); update pleiades.ansible documentation to reflect these steps
  • copy over the production database and blob content
  • get key members of the isaw community to test and verify
  • get another developer not so far involved to do a configuration and security audit
  • decide on a switchover window (date and time)
  • alert tummy and pleiades community to switchover date
  • reboot new server to verify that all processes restart correctly
  • disable changes on legacy production and post notice about same; also email community list
  • do final content copy-over and archive a copy
  • fire up pleiades on the new box and double-check everything from both a development and user point of view
  • update DNS settings to make the new box the production machine
  • verify all is well
  • alert the community
  • shut down pleiades on the machine being retired

New Server Specs

  • Managed Dedicated server
  • Xeon Quad Core E3-1270v6 with 32GB RAM
  • Dual 450GB SSDs, software RAID
  • Redundant Power Supplies
  • Ubuntu 20.04 LTS
Clone this wiki locally