Performance degradation archiving DB with large numbers of FK related records

[melanie witt]

[Impact]
Originally, Nova archives deleted rows in batches consisting of a maximum number of parent rows (max_rows) plus their child rows, all within a single database transaction.
This approach limits the maximum value of max_rows that can be specified by the caller due to the potential size of the database transaction it could generate.
Additionally, this behavior can cause the cleanup process to frequently encounter the following error:
oslo_db.exception.DBError: (pymysql.err.InternalError) (3100, "Error on observer while running replication hook 'before_commit'.")

The error arises when the transaction exceeds the group replication transaction size limit, a safeguard implemented to prevent potential MySQL crashes [1].
The default value for this limit is approximately 143MB.

[Fix]
An upstream commit has changed the logic to archive one parent row and its related child rows in a single database transaction.
This change allows operators to choose more predictable values for max_rows and achieve more progress with each invocation of archive_deleted_rows.
Additionally, this commit reduces the chances of encountering the issue where the transaction size exceeds the group replication transaction size limit.

commit 697fa3c000696da559e52b664c04cbd8d261c037
Author: melanie witt <email address hidden>
CommitDate: Tue Jun 20 20:04:46 2023 +0000

    database: Archive parent and child rows "trees" one at a time

[Test Plan]
1. Create an instance and delete it in OpenStack.
2. Log in to the Nova database and confirm that there is an entry with a deleted_at value that is not NULL.
select display_name, deleted_at from instances where deleted_at <> 0;
3. Execute the following command, ensuring that the timestamp specified in --before is later than the deleted_at value:
nova-manage db archive_deleted_rows --before "XXX-XX-XX XX:XX:XX" --verbose --until-complete
4. Log in to the Nova database again and confirm that the entry has been archived and removed.
select display_name, deleted_at from instances where deleted_at <> 0;

[Where problems could occur]
The commit changes the logic for archiving deleted entries to reduce the size of transactions generated during the operation.
If the patch contains errors, it will only impact the archiving of deleted entries and will not affect other functionalities.

[1] https://bugs.mysql.com/bug.php?id=84785

[Original Bug Description]

Observed downstream in a large scale cluster with constant create/delete
server activity and hundreds of thousands of deleted instances rows.

Currently, we archive deleted rows in batches of max_rows parents +
their child rows in a single database transaction. Doing it that way
limits how high a value of max_rows can be specified by the caller
because of the size of the database transaction it could generate.

For example, in a large scale deployment with hundreds of thousands of
deleted rows and constant server creation and deletion activity, a
value of max_rows=1000 might exceed the database's configured maximum
packet size or timeout due to a database deadlock, forcing the operator
to use a much lower max_rows value like 100 or 50.

And when the operator has e.g. 500,000 deleted instances rows (and
millions of deleted rows total) they are trying to archive, being
forced to use a max_rows value several orders of magnitude lower than
the number of rows they need to archive is a poor user experience and
makes it unclear if archive progress is actually being made.

[OpenStack Infra (hudson-openstack)]

This issue was fixed in the openstack/nova 28.0.0.0rc1 release candidate.

[melanie witt (melwitt)]

https://review.opendev.org/c/openstack/nova/+/877056 merged to master

[melanie witt (melwitt)]

https://review.opendev.org/c/openstack/nova/+/887979 proposed to stable/2023.1

[melanie witt (melwitt)]

https://review.opendev.org/c/openstack/nova/+/887981 proposed to stable/zed

[melanie witt (melwitt)]

https://review.opendev.org/c/openstack/nova/+/887983 proposed to stable/yoga

[melanie witt (melwitt)]

https://review.opendev.org/c/openstack/nova/+/887985 proposed to stable/xena

[melanie witt (melwitt)]

https://review.opendev.org/c/openstack/nova/+/887988 proposed to stable/wallaby

[OpenStack Infra (hudson-openstack)]

Change abandoned by "Elod Illes <email address hidden>" on branch: stable/yoga
Review: https://review.opendev.org/c/openstack/nova/+/887983
Reason: stable/yoga branch of openstack/nova is about to be deleted. To be able to do that, all open patches need to be abandoned. Please cherry pick the patch to unmaintained/yoga if you want to further work on this patch.

[OpenStack Infra (hudson-openstack)]

Change abandoned by "Elod Illes <email address hidden>" on branch: stable/ussuri
Review: https://review.opendev.org/c/openstack/nova/+/888002
Reason: stable/ussuri branch of openstack/nova transitioned to End of Life and is about to be deleted. To be able to do that, all open patches need to be abandoned.

[OpenStack Infra (hudson-openstack)]

Change abandoned by "Elod Illes <email address hidden>" on branch: stable/victoria
Review: https://review.opendev.org/c/openstack/nova/+/887998
Reason: stable/victoria branch of openstack/nova is about to be deleted. To be able to do that, all open patches need to be abandoned. Please cherry pick the patch to unmaintained/victoria if you want to further work on this patch.

[OpenStack Infra (hudson-openstack)]

Change abandoned by "Elod Illes <email address hidden>" on branch: stable/wallaby
Review: https://review.opendev.org/c/openstack/nova/+/887988
Reason: stable/wallaby branch of openstack/nova is about to be deleted. To be able to do that, all open patches need to be abandoned. Please cherry pick the patch to unmaintained/wallaby if you want to further work on this patch.

[OpenStack Infra (hudson-openstack)]

Change abandoned by "Elod Illes <email address hidden>" on branch: stable/xena
Review: https://review.opendev.org/c/openstack/nova/+/887985
Reason: stable/xena branch of openstack/nova is about to be deleted. To be able to do that, all open patches need to be abandoned. Please cherry pick the patch to unmaintained/xena if you want to further work on this patch.

[OpenStack Infra (hudson-openstack)]

Change abandoned by "Elod Illes <email address hidden>" on branch: stable/zed
Review: https://review.opendev.org/c/openstack/nova/+/887981
Reason: stable/zed branch of openstack/nova is about to be deleted. To be able to do that, all open patches need to be abandoned. Please cherry pick the patch to unmaintained/zed if you want to further work on this patch.

[Chengen Du (chengendu)]

Debdiff for Focal

[Chengen Du (chengendu)]

Debdiff for Jammy

[Ubuntu Foundations Team Bug Bot (crichton)]

The attachment "lp2024258-nova-focal.debdiff" seems to be a debdiff. The ubuntu-sponsors team has been subscribed to the bug report so that they can review and hopefully sponsor the debdiff. If the attachment isn't a patch, please remove the "patch" flag from the attachment, remove the "patch" tag, and if you are member of the ~ubuntu-sponsors, unsubscribe the team.

[This is an automated message performed by a Launchpad user owned by ~brian-murray, for any issue please contact him.]