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

lease not found on one member of cluster after restart #12651

Closed
AALOON opened this issue Jan 29, 2021 · 6 comments
Closed

lease not found on one member of cluster after restart #12651

AALOON opened this issue Jan 29, 2021 · 6 comments
Labels

Comments

@AALOON
Copy link

AALOON commented Jan 29, 2021

Hi.
We have a problem. After unexpected restart service/reboot one member of the cluster, this member returns to the cluster and missing a lot of leases and as a result a lot of keys (lease not found on add/update and not found on get keys). On other hand other two members work properly and have such leases => keys.
Our configuration is 3 members cluster, etcd version 3.4.13
From clent side is Include="dotnet-etcd" Version="4.1.1"
image

Usage flow:

  1. creating leases
  2. creating and updating some keys with such lease
  3. reding some keys

Logs from members of cluster.

logs.zip

@ptabor
Copy link
Contributor

ptabor commented Jan 30, 2021

@tangcong Might be the same issue as the one addressed in #12531 ?

@AALOON AALOON changed the title leease not found on one memeber of cluster after restart lease not found on one memeber of cluster after restart Feb 5, 2021
@AALOON AALOON changed the title lease not found on one memeber of cluster after restart lease not found on one member of cluster after restart Feb 5, 2021
@jiapeish
Copy link

Could you try this PR #12531 and see if this issue still occurs?

@AALOON
Copy link
Author

AALOON commented Mar 11, 2021

@jiapeish It didn't work.

@tangcong
Copy link
Contributor

@AALOON can you provide more detail info to reproduce it ? it does not seem to be the same issue as #12531 .

@AALOON
Copy link
Author

AALOON commented May 6, 2021

@tangcong Hi.
I couldn't reproduce it by synthetic test, but we had such flow:

  1. We create two leases for each session with it is timeout (from few minutes to few days). id = 2, lease1 = 2, lease2 = -2 (we do not keep alive leases, leases revoked (remove) in the end of session or by expiration)
  2. During session We Create keys with specified leases depending on business logic.
  3. During session Using this keys , search, lookup, update, delete
  4. but after restart one of instance cluster, it returns, but cluster inconsistent
  5. two instances have keys and leases. the one which was restarted does not have some of leases
  6. but interesting thing some of keys (not all) is presented in this instance if search them without specify lease and I think problem is here.

For now we just removed usage of leases from our application. And it works fine. But we have to manage expiration logic and some other logic by application.
Seems to me leases not synced by raft or revoked on start of broken instance.

@stale
Copy link

stale bot commented Aug 15, 2021

This issue has been automatically marked as stale because it has not had recent activity. It will be closed after 21 days if no further activity occurs. Thank you for your contributions.

@stale stale bot added the stale label Aug 15, 2021
@stale stale bot closed this as completed Sep 5, 2021
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Development

No branches or pull requests

4 participants