allow_overlapping_ips=False fails for Concurrent requests

Bug #1194206 reported by Ritesh Shetty
16
This bug affects 2 people
Affects Status Importance Assigned to Milestone
neutron
Expired
Medium
Unassigned

Bug Description

In my use case if i set allow_overlapping_ips=False i expect no duplicate CIDR's . This works fine. But under concurrent requests this fails. It creates duplicate CIDR's

Changed in quantum:
assignee: nobody → yong sheng gong (gongysh)
importance: Undecided → High
milestone: none → havana-2
status: New → Confirmed
Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Fix proposed to quantum (master)

Fix proposed to branch: master
Review: https://review.openstack.org/34300

Changed in quantum:
status: Confirmed → In Progress
Gary Kotton (garyk)
tags: added: grizzly-backport-potential
Revision history for this message
Salvatore Orlando (salvatore-orlando) wrote :

I have tried an approach based on db consistency (using a unique constraint between cidr and another 'dummy' attribute).
This won't work anyway, as it won't verify overlapping cidrs.

See here for the diff: http://paste.openstack.org/show/39432/

Changed in neutron:
milestone: havana-2 → none
tags: added: db
Alan Pevec (apevec)
tags: removed: grizzly-backport-potential
Changed in neutron:
status: In Progress → Triaged
Revision history for this message
Eugene Nikanorov (enikanorov) wrote :

I think we need more input on this from submitter.
The code might have been changed during last year.
Marking as Incomplete, will reopen if new information appears.

Changed in neutron:
importance: High → Medium
status: Triaged → Incomplete
Revision history for this message
Ritesh Shetty (ritesh-s-shetty) wrote :

What more information is needed?

Revision history for this message
Cedric Brandily (cbrandily) wrote :

This bug is > 365 days without activity. We are unsetting assignee and milestone and setting status to Incomplete in order to allow its expiry in 60 days.

If the bug is still valid, then update the bug status.

Changed in neutron:
assignee: yong sheng gong (gongysh) → nobody
Revision history for this message
Ihar Hrachyshka (ihar-hrachyshka) wrote :

We validate the constraint in code (_validate_subnet_cidr) but if two transactions are started in parallel, and both try to allocate the same CIDR, then validation will succeed for both, and then they will be able to create those resources. Which would violate the expectation.

I presume the bug is real, so moving to another state.

Changed in neutron:
status: Incomplete → Confirmed
tags: added: l3-ipam-dhcp
Changed in neutron:
assignee: nobody → Cedric Brandily (cbrandily)
Revision history for this message
Nam (namnh) wrote :

@Ihar Hrachyshka

Hi Ihar. I also report a bug [1] that it is similar to this bug. Anh I purposed a solution [2]. Could you review my solution and leave some comments, please.
BTW, Kevin suggest that I should create new table for this. However, the ideal is similer.

[1] https://launchpad.net/bugs/1532695
[2] https://review.openstack.org/#/c/267470/

Changed in neutron:
status: Confirmed → Incomplete
assignee: Cedric Brandily (cbrandily) → nobody
Revision history for this message
Launchpad Janitor (janitor) wrote :

[Expired for neutron because there has been no activity for 60 days.]

Changed in neutron:
status: Incomplete → Expired
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.