mcs: register tso allocator #7891
Conversation
|
[REVIEW NOTIFICATION] This pull request has not been approved. To complete the pull request process, please ask the reviewers in the list to review by filling The full list of commands accepted by this bot can be found here. Reviewer can indicate their review by submitting an approval review. |
ti-chi-bot
bot
added
do-not-merge/needs-linked-issue
release-note-none
ti-chi-bot
bot
added
the
size/L
rleungx
force-pushed
the
register-allocator
branch
from
5a3803a to
acf9c7d
Compare
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## master #7891 +/- ##
==========================================
+ Coverage 75.61% 75.63% +0.02%
==========================================
Files 460 461 +1
Lines 72135 72381 +246
==========================================
+ Hits 54543 54745 +202
- Misses 14104 14129 +25
- Partials 3488 3507 +19
Flags with carried forward coverage won't be shown. Click here to find out more. |
pkg/tso/global_allocator.go
Outdated
| func (gta *GlobalTSOAllocator) txnWithTTL(key, value string) (clientv3.LeaseID, error) { | ||
| ctx, cancel := context.WithTimeout(gta.ctx, etcdutil.DefaultRequestTimeout) | ||
| defer cancel() | ||
| grantResp, err := gta.am.etcdClient.Grant(ctx, 3) |
There was a problem hiding this comment.
What about setting a const for 3?
|
|
||
| func (gta *GlobalTSOAllocator) register() (<-chan *clientv3.LeaseKeepAliveResponse, bool) { | ||
| ctx, cancel := context.WithTimeout(gta.ctx, time.Duration(3)*time.Second) | ||
| resp, err := gta.am.etcdClient.Get(ctx, gta.am.allocatorKeyPrefix, clientv3.WithPrefix()) |
There was a problem hiding this comment.
The allocatorKeyPrefix varies in different modes, and it cannot retrieve keys from other allocators as expected.
ti-chi-bot
bot
added
the
needs-rebase
rleungx
force-pushed
the
register-allocator
branch
from
056d12a to
7f23d25
Compare
ti-chi-bot
bot
removed
the
needs-rebase
|
PR needs rebase. Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. |
ti-chi-bot
bot
added
the
needs-rebase
rleungx
force-pushed
the
register-allocator
branch
from
7f23d25 to
9a75c5e
Compare
ti-chi-bot
bot
added
the
dco-signoff: yes
|
[APPROVALNOTIFIER] This PR is NOT APPROVED This pull-request has been approved by: The full list of commands accepted by this bot can be found here.
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
rleungx
force-pushed
the
register-allocator
branch
from
cddbb93 to
c8a6dd3
Compare
rleungx
removed
the
needs-rebase
rleungx
force-pushed
the
register-allocator
branch
from
c8a6dd3 to
de79e49
Compare
rleungx
force-pushed
the
register-allocator
branch
from
d2519ae to
40913a5
Compare
Signed-off-by: Ryan Leung <[email protected]>
Signed-off-by: Ryan Leung <[email protected]>
Signed-off-by: Ryan Leung <[email protected]>
Signed-off-by: Ryan Leung <[email protected]>
Signed-off-by: Ryan Leung <[email protected]>
Signed-off-by: Ryan Leung <[email protected]>
Signed-off-by: Ryan Leung <[email protected]>
Signed-off-by: Ryan Leung <[email protected]>
Signed-off-by: Ryan Leung <[email protected]>
Signed-off-by: Ryan Leung <[email protected]>
Signed-off-by: Ryan Leung <[email protected]>
rleungx
force-pushed
the
register-allocator
branch
from
40913a5 to
c314f61
Compare
What problem does this PR solve?
Issue Number: ref #8477.
What is changed and how does it work?
This PR tries to avoid there are two allocators at the same time. If the TSO service has an allocator that might finish the sync timestamp. Then PD still allocates TSO, after enabling the dynamic switch, it might have a TSO fallback issue.
TSO service sync t1, PD is allocating from t2, t2 is less than t1 at the moment. But if PD syncs again and starts from t2 which is larger than t1, meanwhile TSO is not discovered immediately. The TSO might fall back.
Check List
Tests
Release note