[
https://issues.apache.org/jira/browse/IGNITE-26792?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Mirza Aliev updated IGNITE-26792:
---------------------------------
Description:
h3. Motivation
Current Disaster Recovery design (e.g. resetPartition, restartPartition, etc.)
was developed with persistent storage in mind and does not account for
in-memory partitions where partitions reside in volatile stores.
We need to analyze how DR should behave in such cases and decide whether the
current approach can be extended or if a separate design is required.
h3. Definition of done
* Identify gaps in the current DR implementation regarding in-memory partitions.
* Define possible options for handling volatile stores (reuse existing logic vs
new concept).
* Prepare a proposal or follow-up design task based on findings.
was:
h3. Motivation
Current Disaster Recovery design (e.g. resetPartition, restartPartition, etc.)
was developed with persistent storage in mind and does not account for
in-memory zones where partitions reside in volatile stores.
We need to analyze how DR should behave in such cases and decide whether the
current approach can be extended or if a separate design is required.
h3. Definition of done
* Identify gaps in the current DR implementation regarding in-memory partitions.
* Define possible options for handling volatile stores (reuse existing logic vs
new concept).
* Prepare a proposal or follow-up design task based on findings.
> Analyze Disaster Recovery semantics for in-memory partitions
> ------------------------------------------------------------
>
> Key: IGNITE-26792
> URL: https://issues.apache.org/jira/browse/IGNITE-26792
> Project: Ignite
> Issue Type: Task
> Reporter: Mirza Aliev
> Priority: Major
> Labels: ignite-3
>
> h3. Motivation
> Current Disaster Recovery design (e.g. resetPartition, restartPartition,
> etc.) was developed with persistent storage in mind and does not account for
> in-memory partitions where partitions reside in volatile stores.
> We need to analyze how DR should behave in such cases and decide whether the
> current approach can be extended or if a separate design is required.
> h3. Definition of done
> * Identify gaps in the current DR implementation regarding in-memory
> partitions.
> * Define possible options for handling volatile stores (reuse existing logic
> vs new concept).
> * Prepare a proposal or follow-up design task based on findings.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
