tacker

Controlling the duration of "PENDING" scale tasks

Asked by Gianpietro Lavado

Whenever a scale_in/out task is called, Heat proceeds according to the policy.
Sometimes, for example when the maximum/minimum number of instances has been reached, the VNF remains in the PENDING_SCALE_xxx state for a long time. I thought this time would be related to the 'cooldown' timer in the policy but it's not, is there a way to control it?
Thanks!

Question information

Language:
English Edit question
Status:
Expired
For:
tacker Edit question
Assignee:
No assignee Edit question
Last query:
Last reply:
Revision history for this message
Gianpietro Lavado (gianpietro.lavado) said : 		#1

This is relevant because while the VNF is in PENDING_SCALE_xxx state, it does not attend any other request.
So, when we use scale_in policies, the VNF starts with an alarm activated, and Heat responds that it will not perform the scaling as we have the minimum amount of instances. The problem is that Tacker doesn't seem to acknowledge those kind of messages.

Revision history for this message
Gianpietro Lavado (gianpietro.lavado) said : 		#2

Just realized how it works: https://ask.openstack.org/en/question/102095/tacker-vnfs-stuck-in-pending_scale_xxx/

So through the retries*retry_wait timer, Tacker is just monitoring Heat until it finishes any creation/deletion/scaling task. If we configure the timer too short to avoid the VNF being locked in PENDING tasks, it may not wait enough time for the VNF actual creation and it would throw an error.

Wouldn't it be ideal for Heat to signal Tacker whenever it couldn't perform the task?

Revision history for this message
Launchpad Janitor (janitor) said : 		#3

This question was expired because it remained in the 'Open' state without activity for the last 15 days.