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.

Sign up for GitHub

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

Jump to bottom

kv: add support for Lease.MinExpiration #125235

Closed
nvanbenschoten opened this issue Jun 6, 2024 · 1 comment · Fixed by #126575
Closed

kv: add support for Lease.MinExpiration #125235

nvanbenschoten opened this issue Jun 6, 2024 · 1 comment · Fixed by #126575
Assignees
@nvanbenschoten
Labels
A-kv-replication Relating to Raft, consensus, and coordination. A-leader-leases Related to the introduction of leader leases C-enhancement Solution expected to add code/behavior + preserve backward-compat (pg compat issues are exception) T-kv KV Team

Comments

@nvanbenschoten
Copy link
Member

nvanbenschoten commented Jun 6, 2024 •
edited by exalate-issue-sync bot
Loading

Part of #123847.

We extend the Lease proto with additional information for leader leases:

  • leader_term: ...
  • min_expiration: This field places a lower-bound on the expiration of the leader-lease. It avoids the need for complex interactions during the transition from an expiration-based lease to a leader lease to ensure that the effective expiration of the lease does not regress. This also avoids the need for a leader lease acquisition to wait for fortification.
    • We recently fixed a bug with the expiration-based to epoch-based transition in #123442 which would have been much simpler to fix had such a minimum expiration field existed for epoch-based leases.

This new field can be used to epoch-based leases as well, as long as it is put behind a cluster version gate.

Jira issue: CRDB-39328

Epic CRDB-37522
@nvanbenschoten nvanbenschoten added C-enhancement Solution expected to add code/behavior + preserve backward-compat (pg compat issues are exception) A-kv-replication Relating to Raft, consensus, and coordination. T-kv KV Team A-leader-leases Related to the introduction of leader leases labels Jun 6, 2024
@blathers-crl blathers[crl]
Copy link

blathers-crl bot commented Jun 6, 2024

cc @cockroachdb/replication

@nvanbenschoten nvanbenschoten self-assigned this Jun 6, 2024
nvanbenschoten added a commit to nvanbenschoten/cockroach that referenced this issue Jun 10, 2024
@nvanbenschoten
kv: retry liveness heartbeat on race with insufficient expiration
b2ac52a 
Fixes cockroachdb#124693.
Fixes cockroachdb#125287.

This commit adds logic to retry the synchronous liveness heartbeat which
ensures that the liveness record has a later expiration than the prior
lease by the time the lease promotion goes into effect, which was added
in cockroachdb#123442. This heartbeat may need to be retried because it may have
raced with another liveness heartbeat which did not extend the liveness
expiration far enough.

We opt to allow this race and retry across it instead of detecting it
and returning an error from `NodeLiveness.Heartbeat` because:
1. returning an error would have a larger blast radius and could cause
   other issues.
2. returning an error wouldn't actually fix the tests that are failing,
   because they would still get an error, just a different one.

Before this commit, `TestLeaseQueueProactiveEnqueueOnPreferences` would
hit this case fail under deadlock and stress every ~150 iterations.
After this commit, the test passes continuously under deadlock stress
for over 2000 runs.

This makes cockroachdb#123442 even uglier. The nicer solution is cockroachdb#125235, but that
is not backportable. Still, we're planning to address that issue as part
of the leader leases work, so this is a temporary fix.

This also removes a TODO added in 1dc18df. As mentioned above, we don't
address it, but instead document the behavior.

Release note (bug fix): resolves a concerning log message that says
"expiration of liveness record ... is not greater than expiration of the
previous lease ... after liveness heartbeat". This message is no longer
possible.
@nvanbenschoten nvanbenschoten mentioned this issue Jun 10, 2024
Merged
craig bot pushed a commit that referenced this issue Jun 10, 2024
@nvanbenschoten @stevendanna @dt
Merge #124885 #125309 #125369 #125411
d554c0e 
124885: kv: retry liveness heartbeat on race with insufficient expiration r=nvanbenschoten a=nvanbenschoten

Fixes #124693.
Fixes #125287.

This commit adds logic to retry the synchronous liveness heartbeat which ensures that the liveness record has a later expiration than the prior lease by the time the lease promotion goes into effect, which was added in #123442. This heartbeat may need to be retried because it may have raced with another liveness heartbeat which did not extend the liveness expiration far enough.

We opt to allow this race and retry across it instead of detecting it and returning an error from `NodeLiveness.Heartbeat` because:
1. returning an error would have a larger blast radius and could cause other issues.
2. returning an error wouldn't actually fix the tests that are failing, because they would still get an error, just a different one.

Before this commit, `TestLeaseQueueProactiveEnqueueOnPreferences` would hit this case fail under deadlock and stress every ~150 iterations. After this commit, the test passes continuously under deadlock stress for over 2000 runs.

This makes #123442 even uglier. The nicer solution is #125235, but that is not backportable. Still, we're planning to address that issue as part of the leader leases work, so this is a temporary fix.

This also removes a TODO added in 1dc18df. As mentioned above, we don't address it, but instead document the behavior.

Release note (bug fix): resolves a concerning log message that says "expiration of liveness record ... is not greater than expiration of the previous lease ... after liveness heartbeat". This message is no longer possible.

125309: streamingccl: fix metric name r=msbutler a=stevendanna

This metric was re-using the name of another metric.

Epic: none
Release note: None

125369: streamproducer: add emit_wait and produce_wait to vtable r=rickystewart a=dt

Release note: none.
Epic: none.

125411: streamingccl: cleanup unit test terminology r=stevendanna a=stevendanna

Mixing source/target with serverA/serverB was a bit confusing.

Epic: none
Release note: None

Co-authored-by: Nathan VanBenschoten <nvanbenschoten@gmail.com>
Co-authored-by: Steven Danna <danna@cockroachlabs.com>
Co-authored-by: David Taylor <tinystatemachine@gmail.com>
blathers-crl bot pushed a commit that referenced this issue Jun 10, 2024
@nvanbenschoten
kv: retry liveness heartbeat on race with insufficient expiration
2268baf 
Fixes #124693.
Fixes #125287.

This commit adds logic to retry the synchronous liveness heartbeat which
ensures that the liveness record has a later expiration than the prior
lease by the time the lease promotion goes into effect, which was added
in #123442. This heartbeat may need to be retried because it may have
raced with another liveness heartbeat which did not extend the liveness
expiration far enough.

We opt to allow this race and retry across it instead of detecting it
and returning an error from `NodeLiveness.Heartbeat` because:
1. returning an error would have a larger blast radius and could cause
   other issues.
2. returning an error wouldn't actually fix the tests that are failing,
   because they would still get an error, just a different one.

Before this commit, `TestLeaseQueueProactiveEnqueueOnPreferences` would
hit this case fail under deadlock and stress every ~150 iterations.
After this commit, the test passes continuously under deadlock stress
for over 2000 runs.

This makes #123442 even uglier. The nicer solution is #125235, but that
is not backportable. Still, we're planning to address that issue as part
of the leader leases work, so this is a temporary fix.

This also removes a TODO added in 1dc18df. As mentioned above, we don't
address it, but instead document the behavior.

Release note (bug fix): resolves a concerning log message that says
"expiration of liveness record ... is not greater than expiration of the
previous lease ... after liveness heartbeat". This message is no longer
possible.
@blathers-crl blathers-crl bot mentioned this issue Jun 10, 2024
Merged
@nvanbenschoten nvanbenschoten mentioned this issue Jun 14, 2024
Merged
@nvanbenschoten nvanbenschoten mentioned this issue Jun 25, 2024
Merged
nvanbenschoten added a commit to nvanbenschoten/cockroach that referenced this issue Jun 26, 2024
@nvanbenschoten
roachpb: add MinExpiration field to Lease
be02654 
Informs cockroachdb#125235.

This commit adds a new MinExpiration field to the Lease struct. This
field defines the minimum expiration at which the lease expires,
independent of any other expiry condition. This field can be used to
place a floor on the expiration for epoch-based leases and leader leases
(not yet implemented) to prevent expiration regressions when upgrading
from an expiration-based lease.

The field is not yet used.

Release note: None
nvanbenschoten added a commit to nvanbenschoten/cockroach that referenced this issue Jun 26, 2024
@nvanbenschoten
ui: add MinExpiration field to range debug page
4428079 
Informs cockroachdb#125235.

This commit adds the new MinExpiration field to the range debug page.

Release note: None
nvanbenschoten added a commit to nvanbenschoten/cockroach that referenced this issue Jul 1, 2024
@nvanbenschoten
roachpb: add MinExpiration field to Lease
7781cfa 
Informs cockroachdb#125235.

This commit adds a new MinExpiration field to the Lease struct. This
field defines the minimum expiration at which the lease expires,
independent of any other expiry condition. This field can be used to
place a floor on the expiration for epoch-based leases and leader leases
(not yet implemented) to prevent expiration regressions when upgrading
from an expiration-based lease.

The field is not yet used.

Release note: None
nvanbenschoten added a commit to nvanbenschoten/cockroach that referenced this issue Jul 1, 2024
@nvanbenschoten
ui: add MinExpiration field to range debug page
96cf596 
Informs cockroachdb#125235.

This commit adds the new MinExpiration field to the range debug page.

Release note: None
craig bot pushed a commit that referenced this issue Jul 1, 2024
@nvanbenschoten
Merge #126202
ccb6f96 
126202: roachpb: add MinExpiration field to Lease r=arulajmani a=nvanbenschoten

First half of #125235.

This commit adds a new `MinExpiration` field to the `Lease` struct. This field defines the minimum expiration at which the lease expires, independent of any other expiry condition. This field can be used to place a floor on the expiration for epoch-based leases and leader leases (not yet implemented) to prevent expiration regressions when upgrading from an expiration-based lease.

The field is not yet used.

Release note: None

Co-authored-by: Nathan VanBenschoten <nvanbenschoten@gmail.com>
@nvanbenschoten nvanbenschoten mentioned this issue Jul 2, 2024
Merged
nvanbenschoten added a commit to nvanbenschoten/cockroach that referenced this issue Jul 2, 2024
@nvanbenschoten
kv: add support for Lease.MinExpiration
d10fbb1 
Closes cockroachdb#125235.

This commit begins using the new `Lease.MinExpiration` field, which was
introduced in 7781cfa. `Lease.MinExpiration` is assigned to epoch-based
leases (and soon, leader leases) during upgrades from expiration-based
lease. It is a simple, direct way to ensure that the expiration of the
lease does not regress during such promotions.

We previously solved this issue in 6dd54b4 by manually heartbeating the
node liveness record when we detected an expiration regression. This was
backwards compatible, so it was a better backport target, but it was
also complex, as demonstrated by the need for b2ac52a. In the future,
we will be able to remove the manual heartbeat logic, as it is no longer
necessary once MinTimestamp is supported by all clusters.

Epic: None
craig bot pushed a commit that referenced this issue Jul 8, 2024
@annrpom @xinhaoz @nvanbenschoten
Merge #126336 #126419 #126575
b5ab804 
126336: sql: override StmtTimeout to 0 for background jobs r=annrpom a=annrpom

For background jobs and schema changes, our Internal
Executor inherits from cluster settings. If a statement
timeout is set cluster-wide, then background jobs will
respect this -- which does not make much sense; thus,
this patch overrides the StmtTimeout to 0s for background
jobs.

Fixes: #126261
Release note (bug fix): Background jobs no longer respect
a statement timeout.

126419: ui: fix query fetching store ids for db/tables r=xinhaoz a=xinhaoz

These queries were updated in #118904 to make them more performant. In that change we mistakenly removed the null check before reading the `rows` field, causing pages using this request to crash if this query fails.

Epic: none
Fixes: #126348

Release note (bug fix): Database overview and db details pages should not crash if the range information is not available.

126575: kv: add support for Lease.MinExpiration r=nvanbenschoten a=nvanbenschoten

Closes #125235.

This commit begins using the new `Lease.MinExpiration` field, which was introduced in 7781cfa. `Lease.MinExpiration` is assigned to epoch-based leases (and soon, leader leases) during upgrades from expiration-based lease. It is a simple, direct way to ensure that the expiration of the lease does not regress during such promotions.

We previously solved this issue in 6dd54b4 by manually heartbeating the node liveness record when we detected an expiration regression. This was backwards compatible, so it was a better backport target, but it was also complex, as demonstrated by the need for b2ac52a. In the future, we will be able to remove the manual heartbeat logic, as it is no longer necessary once MinTimestamp is supported by all clusters.

Epic: None

Co-authored-by: Annie Pompa <annie@cockroachlabs.com>
Co-authored-by: Xin Hao Zhang <xzhang@cockroachlabs.com>
Co-authored-by: Nathan VanBenschoten <nvanbenschoten@gmail.com>
@craig craig bot closed this as completed in ba758c0 Jul 8, 2024
asg0451 pushed a commit to asg0451/cockroach that referenced this issue Jul 10, 2024
@nvanbenschoten @asg0451
kv: add support for Lease.MinExpiration
95915bb 
Closes cockroachdb#125235.

This commit begins using the new `Lease.MinExpiration` field, which was
introduced in 7781cfa. `Lease.MinExpiration` is assigned to epoch-based
leases (and soon, leader leases) during upgrades from expiration-based
lease. It is a simple, direct way to ensure that the expiration of the
lease does not regress during such promotions.

We previously solved this issue in 6dd54b4 by manually heartbeating the
node liveness record when we detected an expiration regression. This was
backwards compatible, so it was a better backport target, but it was
also complex, as demonstrated by the need for b2ac52a. In the future,
we will be able to remove the manual heartbeat logic, as it is no longer
necessary once MinTimestamp is supported by all clusters.

Epic: None
nvanbenschoten added a commit that referenced this issue Aug 29, 2024
@nvanbenschoten
kv: retry liveness heartbeat on race with insufficient expiration
34ae87f 
Fixes #124693.
Fixes #125287.

This commit adds logic to retry the synchronous liveness heartbeat which
ensures that the liveness record has a later expiration than the prior
lease by the time the lease promotion goes into effect, which was added
in #123442. This heartbeat may need to be retried because it may have
raced with another liveness heartbeat which did not extend the liveness
expiration far enough.

We opt to allow this race and retry across it instead of detecting it
and returning an error from `NodeLiveness.Heartbeat` because:
1. returning an error would have a larger blast radius and could cause
   other issues.
2. returning an error wouldn't actually fix the tests that are failing,
   because they would still get an error, just a different one.

Before this commit, `TestLeaseQueueProactiveEnqueueOnPreferences` would
hit this case fail under deadlock and stress every ~150 iterations.
After this commit, the test passes continuously under deadlock stress
for over 2000 runs.

This makes #123442 even uglier. The nicer solution is #125235, but that
is not backportable. Still, we're planning to address that issue as part
of the leader leases work, so this is a temporary fix.

This also removes a TODO added in 1dc18df. As mentioned above, we don't
address it, but instead document the behavior.

Release note (bug fix): resolves a concerning log message that says
"expiration of liveness record ... is not greater than expiration of the
previous lease ... after liveness heartbeat". This message is no longer
possible.
celiala pushed a commit that referenced this issue Sep 12, 2024
@nvanbenschoten @celiala
kv: retry liveness heartbeat on race with insufficient expiration
e7ed406 
Fixes #124693.
Fixes #125287.

This commit adds logic to retry the synchronous liveness heartbeat which
ensures that the liveness record has a later expiration than the prior
lease by the time the lease promotion goes into effect, which was added
in #123442. This heartbeat may need to be retried because it may have
raced with another liveness heartbeat which did not extend the liveness
expiration far enough.

We opt to allow this race and retry across it instead of detecting it
and returning an error from `NodeLiveness.Heartbeat` because:
1. returning an error would have a larger blast radius and could cause
   other issues.
2. returning an error wouldn't actually fix the tests that are failing,
   because they would still get an error, just a different one.

Before this commit, `TestLeaseQueueProactiveEnqueueOnPreferences` would
hit this case fail under deadlock and stress every ~150 iterations.
After this commit, the test passes continuously under deadlock stress
for over 2000 runs.

This makes #123442 even uglier. The nicer solution is #125235, but that
is not backportable. Still, we're planning to address that issue as part
of the leader leases work, so this is a temporary fix.

This also removes a TODO added in 1dc18df. As mentioned above, we don't
address it, but instead document the behavior.

Release note (bug fix): resolves a concerning log message that says
"expiration of liveness record ... is not greater than expiration of the
previous lease ... after liveness heartbeat". This message is no longer
possible.
nvanbenschoten added a commit to nvanbenschoten/cockroach that referenced this issue Sep 12, 2024
@nvanbenschoten
kv: retry liveness heartbeat on race with insufficient expiration
09ffbd8 
Fixes cockroachdb#124693.
Fixes cockroachdb#125287.

This commit adds logic to retry the synchronous liveness heartbeat which
ensures that the liveness record has a later expiration than the prior
lease by the time the lease promotion goes into effect, which was added
in cockroachdb#123442. This heartbeat may need to be retried because it may have
raced with another liveness heartbeat which did not extend the liveness
expiration far enough.

We opt to allow this race and retry across it instead of detecting it
and returning an error from `NodeLiveness.Heartbeat` because:
1. returning an error would have a larger blast radius and could cause
   other issues.
2. returning an error wouldn't actually fix the tests that are failing,
   because they would still get an error, just a different one.

Before this commit, `TestLeaseQueueProactiveEnqueueOnPreferences` would
hit this case fail under deadlock and stress every ~150 iterations.
After this commit, the test passes continuously under deadlock stress
for over 2000 runs.

This makes cockroachdb#123442 even uglier. The nicer solution is cockroachdb#125235, but that
is not backportable. Still, we're planning to address that issue as part
of the leader leases work, so this is a temporary fix.

This also removes a TODO added in 1dc18df. As mentioned above, we don't
address it, but instead document the behavior.

Release note (bug fix): resolves a concerning log message that says
"expiration of liveness record ... is not greater than expiration of the
previous lease ... after liveness heartbeat". This message is no longer
possible.
@nvanbenschoten nvanbenschoten mentioned this issue Sep 12, 2024
Merged
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@nvanbenschoten nvanbenschoten

Labels
A-kv-replication Relating to Raft, consensus, and coordination. A-leader-leases Related to the introduction of leader leases C-enhancement Solution expected to add code/behavior + preserve backward-compat (pg compat issues are exception) T-kv KV Team
Projects
No open projects
KV
Status: Incoming
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

kv: add support for Lease.MinExpiration nvanbenschoten/cockroach
1 participant
@nvanbenschoten