Share Replication: File-based locks don't provide concurrency control in multi-node/multi-AZ deployments
Bug #1585241 reported by
Goutham Pacha Ravi
This bug affects 1 person
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
OpenStack Shared File Systems Service (Manila) |
Triaged
|
Low
|
Goutham Pacha Ravi |
Bug Description
Workflows in the share replication feature introduced in Mitaka are coordinated by the share manager with the help of locks deriving from oslo_concurrency. These locks are limited to file based locks. If deployers choose to have the manila-share service run across multiple controller nodes (multi-node deployment), local file locks are of no use.
https:/
Changed in manila: | |
assignee: | nobody → monika (monikaparkar25) |
Changed in manila: | |
assignee: | monika (monikaparkar25) → Goutham Pacha Ravi (gouthamr) |
status: | New → In Progress |
Changed in manila: | |
importance: | Undecided → Medium |
Changed in manila: | |
assignee: | Goutham Pacha Ravi (gouthamr) → Tom Barron (tpb) |
Changed in manila: | |
assignee: | Tom Barron (tpb) → Goutham Pacha Ravi (gouthamr) |
tags: | added: races |
Changed in manila: | |
milestone: | none → train-rc1 |
Changed in manila: | |
milestone: | ussuri-3 → victoria-rc1 |
Changed in manila: | |
milestone: | victoria-rc1 → wallaby-1 |
summary: |
- File-based locks don't provide concurrency over multi-node/multi-AZ - deployments + Share Replication: File-based locks don't provide concurrency control in + multi-node/multi-AZ deployments |
Changed in manila: | |
milestone: | wallaby-1 → wallaby-3 |
Changed in manila: | |
milestone: | wallaby-3 → none |
Changed in manila: | |
importance: | Medium → Low |
milestone: | none → zed-3 |
Changed in manila: | |
milestone: | zed-3 → antelope-1 |
Changed in manila: | |
milestone: | antelope-1 → antelope-2 |
Changed in manila: | |
milestone: | antelope-2 → antelope-rc1 |
Changed in manila: | |
milestone: | antelope-rc1 → bobcat-1 |
Changed in manila: | |
milestone: | bobcat-1 → bobcat-2 |
Changed in manila: | |
milestone: | bobcat-2 → bobcat-3 |
Changed in manila: | |
milestone: | bobcat-3 → caracal-1 |
Changed in manila: | |
milestone: | caracal-1 → caracal-2 |
Changed in manila: | |
milestone: | caracal-2 → caracal-3 |
Changed in manila: | |
milestone: | caracal-3 → dalmation-1 |
Changed in manila: | |
milestone: | dalmatian-1 → dalmatian-3 |
Changed in manila: | |
milestone: | dalmatian-3 → epoxy-1 |
Changed in manila: | |
milestone: | epoxy-1 → epoxy-3 |
To post a comment you must log in.
After much deliberation and multiple design summit discussions, we have the impression that this is not to be treated as a bug. We will introduce a mechanism to provide distributed locking management underneath manila. However, deployers today have an option of using file locks living on a shared file system across distributed services.
The tooz abstraction is proposed here: https:/ /review. openstack. org/#/c/ 318336
It will not be back ported to the releases where this "bug" exists;
Please note, that tooz can also be used with no further configuration, i.e, default to using file locks as is the current behavior for the share replication feature to work as intended.