Skip to content

Disaster recovery in case of data lost #3583

@truong-hua

Description

@truong-hua

FEATURE REQUEST

Sometimes, we lost all data of a bookie (it's ledgers, cookies and others). The problem here is that if a ledger of that bookie can not be recover due to any reason. The shell decommission command will stuck in waiting for these ledger to be replicated but of course it will never happen.

In reality we have some data that is defined not important and we are willing to lost it to optimize for performance, so we configure the quorum to only 1.

Currently, there are no way for us to remove the dead bookie as the shell decommission will be stuck in waiting for dead ledger to be replicated. The only workaround solution for us is to separate the non-important cluster and re-init the whole cluster every time the disaster happen.

EXPECTATION
A shell decommission -force option which will force remove the dead bookie and fill empty (or some dummy data) to dead ledgers.

Metadata

Metadata

Assignees

No one assigned

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions