Increase default heat stack configurations to accommodate systems with lower specs

Bug #1521323 reported by Sripriya
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
tacker
Fix Released
Low
Sripriya

Bug Description

Currently, tacker has a default wait time out of 50 seconds for heat stack creation to be completed. Sometimes, the stack creation can take a longer time to spawn the resources on a VM or a system with lower specs (CPU and RAM).

Actual behavior:

Tacker sets the VNF to ERROR state after trying for 50 seconds even thought stack creation would be eventually completed after 50 seconds

Expected behavior:
Tacker should not set the VNF to ERROR state as this is a valid use case.

To handle this case, tacker default configurations for stack retry and stack retry wait will be bumped up to a max of 300 seconds (default max timeout) instead of the current default of 50 seconds. With this setting, Tacker will be waiting for 300 seconds for stack creation to be completed before eventually giving up.

There is no rational for choosing configuration value '300' but as a fair number considering systems with varied specs running Tacker/OpenStack.

Sripriya (sseetha)
Changed in tacker:
assignee: nobody → Sripriya (sseetha)
Changed in tacker:
status: New → In Progress
Changed in tacker:
importance: Undecided → Low
Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Fix merged to tacker (master)

Reviewed: https://review.openstack.org/251565
Committed: https://git.openstack.org/cgit/openstack/tacker/commit/?id=e876c5b06bf1bc939afd93add01e57e57f26c7e7
Submitter: Jenkins
Branch: master

commit e876c5b06bf1bc939afd93add01e57e57f26c7e7
Author: Sripriya Seetharam <email address hidden>
Date: Mon Nov 30 14:22:57 2015 -0800

    Increase the heat stack wait time out

    This fix handles the usecase where Tacker falsely sets the VNF to
    ERROR state after waiting for 50 seconds even though the actual
    stack is succesfully completed whose creation time is greater than
    50 seconds.

    The current default wait time out of 50 seconds is bumped up to
    300 seconds as a fair timeout, keeping the memory/cpu constrained
    vms/systems in consideration.

    Closes-Bug: #1521323
    Change-Id: Ied1d904cc0dac2df91dee856100baeb988d65676

Changed in tacker:
status: In Progress → Fix Released
Revision history for this message
Thierry Carrez (ttx) wrote : Fix included in openstack/tacker 0.3.0

This issue was fixed in the openstack/tacker 0.3.0 release.

Revision history for this message
Amit Bhardwaj (bhardwajamit23) wrote :

Hey, What if I want to increase it further. My setup is a bit slow.. Is there any config/command parameter available using which I can override the 300 seconds time.

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.