[IGNITE-14744] Restore snapshot taken on different topologies

• Type: Improvement
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 2.10
• Fix Version/s: 2.12
• Labels: iep-43 important ise
• Release Note: Added restore snapshot taken on different topologies with index rebuild
• Ignite Flags: Docs Required, Release Notes Required
• Assignee: Maxim Muzafarov
• Reporter: Maxim Muzafarov
• Reviewer: Aleksey Plekhanov
• Created: 19/May/21 18:42
• Updated: 09/Sep/22 17:17
• Resolved: 22/Nov/21 07:40

Description

Currently, the snapshot restore procedure can be performed in the same cluster topology. This means that it will not work in topologies with a different number of cluster nodes (for example, fewer nodes) since it requires a complete rebuild of the index for the cache groups.
We must support the restore procedure on different cluster topologies.

Issue Links

• depends upon: IGNITE-15208 Remove unnecessary rebalance order classes for exchange thread
• depends upon: IGNITE-15283 Remove duplicated managing of CacheDataStore in offheap manager
• depends upon: IGNITE-15413 Extract cancellation index rebuild token on upper level
• depends upon: IGNITE-15802 Remove duplicated code for affinity assignment manager
• relates to: IGNITE-15967 Copying the remote index from the snapshot if the node has the same partition distribution
• relates to: IGNITE-15975 Snapshot restore should wait for the index rebuild prior to releasing future
• links to: GitHub Pull Request #9321
• links to: GitHub Pull Request #9539
• links to: GitHub Pull Request #9575