Page MenuHomePhabricator

Build replacement for the webservice toolschecker test
Open, MediumPublic

Description

It tested a bunch of different things:

  1. Start a webservice
  2. Verify it is up
  3. Stop the webservice
  4. Verify it is down

on both gridengine and kubernetes. It also used the 'webservice' command that caused racing problems, since that was designed to be used by humans. Need to split that up into multiple smaller checks, and remove the current ones.

Event Timeline

Change 303120 had a related patch set uploaded (by Yuvipanda):
toollabs: Don't have icinga check for webservice

https://gerrit.wikimedia.org/r/303120

Change 303120 merged by Yuvipanda:
toollabs: Don't have icinga check for webservice

https://gerrit.wikimedia.org/r/303120

For kubernetes, we should do the following

  1. Spawn a simple pod with a built in simple web server, no NFS, hit its IP to see if it responds. Kill it after.
  2. Spawn a pod with NFS on with a simple web server, hit its IP to see if it responds. Kill it after.
  3. Spawn a pod with the simple built in web server, no NFS, create a service IP, hit the service IP to see if it responds. Kill it after.
  4. Spawn a pod with the simple built in web server, no NFS, create a service IP, hit proxy to see if it responds properly.
scfc moved this task from Triage to Backlog on the Cloud-Services board.
scfc moved this task from Backlog to Triage on the Cloud-Services board.
scfc moved this task from Triage to Backlog on the Toolforge board.
Bstorm lowered the priority of this task from High to Medium.Feb 25 2020, 5:39 PM