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

Snapshot of entire state as a WAL recovery optimization #682

Closed
LefterisJP opened this Issue Jun 28, 2017 · 0 comments

Comments

Projects
None yet
3 participants
@LefterisJP
Collaborator

LefterisJP commented Jun 28, 2017

Problem Definition

When recovering from the Write Ahead Log we would need to replay all the state changes that reside in it. This turns out to be too many state changes and also as result of those state changes too many messages are resent and we end up spamming our peers in Matrix.

Solution

  • Implement a queue of pending chain transactions -- Issue Link

  • Choose a snapshot point, as a point in time where we have no pending on-chain transactions. This will allow us to have a clear restarting point that will not put us in danger of ending up in a broken state when restarting.

  • At recovery/restart load up the state at the snapshot and start replaying WAL changes only after the snapshot has been taken.

@hackaugusto hackaugusto referenced this issue Jun 28, 2017

Closed

Make Raiden recoverable #205

10 of 13 tasks complete

@hackaugusto hackaugusto added this to the Red Eyes milestone Apr 19, 2018

@hackaugusto hackaugusto removed this from the Red Eyes milestone Jun 14, 2018

@LefterisJP LefterisJP added this to the Red Eyes milestone Jul 25, 2018

@LefterisJP LefterisJP added the P1 label Aug 3, 2018

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment