Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

release-20.2: kvserver: allow healing ranges whose replicas are all lost #57941

Closed

Conversation

tbg
Copy link
Member

@tbg tbg commented Dec 15, 2020

Backport:

Please see individual PRs for details.

/cc @cockroachdb/release

@tbg tbg requested a review from irfansharif December 15, 2020 09:57
@tbg tbg requested a review from a team as a code owner December 15, 2020 09:57
@cockroach-teamcity
Copy link
Member

This change is Reviewable

@tbg tbg changed the title release-20.2: TODO release-20.2: kvserver: allow healing ranges whose replicas are all lost Dec 15, 2020
Copy link
Contributor

@irfansharif irfansharif left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Sorry about the delayed review, was asleep.

TheSamHuang and others added 4 commits January 4, 2021 17:39
Introduces and implements an `ResetQuorumRequest` RPC. `ResetQuorumRequest`
takes in the range id of a range that that is unavailable due to lost quorum
and makes it available again, at the cost of losing all of the data in that
range. Any existing replica, even one residing on the target node, will
irrevocably be removed. ResetQuorumRequest first uses meta2 to identify the
range descriptor. Then, it removes all replicas from the range descriptor and
adds a store from the target node as the one designated survivor replica. This
change is then written to meta2 and sent as a snapshot to a store local to the
target node in order to use crdb internal upreplication and rebalancing
mechanisms to create further replicas from this fresh snapshot.

This RPC is meant to be called by the user directly. It will not work for
ranges that have not lost quorum or for a meta range.

Release note: None
…PC to

restore quorum to a given range ID. Note that data on any surviving replicas
will not be used to restore quorum. Instead, these replicas will be removed
irrevocably.

Release note: None.
...to the running cluster.

Release note: None
@tbg tbg force-pushed the backport20.2-56333-57034-57712-57825 branch from 5f174a9 to 99efa3f Compare January 4, 2021 16:41
@tbg tbg closed this Apr 20, 2021
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants