[Bug 1882657] Re: Batch update member to pool failed due to new_member is not exist in DB

[Hemanth Nakkina]

** Patch added: "Debdiff for UCA stein"
   https://bugs.launchpad.net/cloud-archive/ussuri/+bug/1882657/+attachment/5390714/+files/lp1882657_stein.debdiff

-- 
You received this bug notification because you are a member of Ubuntu
OpenStack, which is subscribed to Ubuntu Cloud Archive.
https://bugs.launchpad.net/bugs/1882657

Title:
  Batch update member to pool failed due to new_member is not exist in
  DB

Status in Ubuntu Cloud Archive:
  Fix Released
Status in Ubuntu Cloud Archive stein series:
  New
Status in Ubuntu Cloud Archive train series:
  New
Status in Ubuntu Cloud Archive ussuri series:
  New
Status in Ubuntu Cloud Archive victoria series:
  Fix Released
Status in octavia package in Ubuntu:
  Fix Released
Status in octavia source package in Focal:
  New
Status in octavia source package in Groovy:
  Fix Released

Bug description:
  Original Upstream bug: https://storyboard.openstack.org/#!/story/2007581
  (The problem is observed in OpenStack Stein)

  Code review in progress (Upstream) -
  https://review.opendev.org/#/c/733670/

  
  ==========================================================================

  [Impact]
  Adding members to loadbalancer in batch mode fails and loadbalancers are
  stuck in pending_create state.
  This patch retries the read from DB 15 times to check if DB is populated properly or not.

  [Test Case]

  1. Reproducing the issue

  1a. Create 2 or more instances to add as members to loadbalancer
      (The issue can be reproduced if there is load on the machine, on my test environment able to reproduce with 2 members)
  1b. Create a loadbalancer, listener, pool and healthmonitor
      openstack loadbalancer create --name lb1 --vip-subnet-id private_subnet
      openstack loadbalancer listener create --name listener1 --protocol HTTP --protocol-port 80 lb1
      openstack loadbalancer pool create --name pool1 --lb-algorithm ROUND_ROBIN --listener listener1 --protocol HTTP
      openstack loadbalancer healthmonitor create --name hm1 --delay 5 --max-retries 4 --timeout 10 --type HTTP --url-path / pool1
  1c. Create a token
      AUTH_TOKEN=$(openstack token issue  -c id -f value)
  1d. Prepare a file members.json based on Request example in the below link
      https://docs.openstack.org/api-ref/load-balancer/v2/?expanded=batch-update-members-detail#batch-update-members
  1e. Add members in batch mode
      curl -X PUT -H "Content-Type: application/json" -H "X-Auth-Token: $AUTH_TOKEN" -d @members.json http://<ip>:<port>/v2.0/lbaas/pools/<pool_id>/members
  1f. Check loadbalancer show. This will be in pending_create state in non-working case
      openstack loadbalancer show lb1

  2. Install the package with fixed code

  3. Confirm bug have been fixed

  3a. Repeat steps 1a-1f. Loadbalancer should be in ACTIVE state

  
  [Regression Potential]

  Upstream CI passed for all the releases including the backported versions.
  Installation of new package will restart apache service and hence there will be downtime of octavia service for few milliseconds.

To manage notifications about this bug go to:
https://bugs.launchpad.net/cloud-archive/+bug/1882657/+subscriptions