[
https://issues.apache.org/jira/browse/IGNITE-19239?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Ilya Shishkov updated IGNITE-19239:
-----------------------------------
Description: (was: Currently, snapshot check is performed under the
hood during restoring of a snapshot. For big snapshots check operation
execution can take a long time.
In such case, if you call two similar commands, they will return different
results:
Deprecated command _restore --status_ shows, that snapshot operation is in
progress:
{noformat}
Command [SNAPSHOT] started
Arguments: --snapshot restore test_snapshot --status
--------------------------------------------------------------------------------
Command deprecated. Use '--snapshot status' instead.
Snapshot cache group restore operation is running [snapshot=test_snapshot]
Command [SNAPSHOT] finished with code: 0
{noformat}
But actual command _--snapshot status_ does not show that any snapshot
operation is running:
{noformat}
Command [SNAPSHOT] started
Arguments: --snapshot status
--------------------------------------------------------------------------------
There is no create or restore snapshot operation in progress.
Command [SNAPSHOT] finished with code: 0
{noformat}
{{VisorSnapshotStatusTask.java}} rely on snapshot restore metrics[1]. This
metrics are extracted from {{SnapshotRestoreProcess#lastOpCtx}} [2], which is
updated in {{SnapshotRestoreProcess#prepare}} [3], _but this method called
after snapshot check is finished_ [4, 5].
Here is a reproducer: [^SnapshotStatusReproducerTest.patch]
Links:
#
https://github.com/apache/ignite/blob/a8e16899514348ed7b9404e7d122ed430f8e1acd/modules/core/src/main/java/org/apache/ignite/internal/visor/snapshot/VisorSnapshotStatusTask.java#L141
#
https://github.com/apache/ignite/blob/a8e16899514348ed7b9404e7d122ed430f8e1acd/modules/core/src/main/java/org/apache/ignite/internal/processors/cache/persistence/snapshot/SnapshotRestoreProcess.java#L245
#
https://github.com/apache/ignite/blob/a8e16899514348ed7b9404e7d122ed430f8e1acd/modules/core/src/main/java/org/apache/ignite/internal/processors/cache/persistence/snapshot/SnapshotRestoreProcess.java#L670
#
https://github.com/apache/ignite/blob/a8e16899514348ed7b9404e7d122ed430f8e1acd/modules/core/src/main/java/org/apache/ignite/internal/processors/cache/persistence/snapshot/SnapshotRestoreProcess.java#L414
#
https://github.com/apache/ignite/blob/a8e16899514348ed7b9404e7d122ed430f8e1acd/modules/core/src/main/java/org/apache/ignite/internal/processors/cache/persistence/snapshot/SnapshotRestoreProcess.java#L205)
> Checkpoint read lock acquisition timeouts during snapshot restore
> -----------------------------------------------------------------
>
> Key: IGNITE-19239
> URL: https://issues.apache.org/jira/browse/IGNITE-19239
> Project: Ignite
> Issue Type: Bug
> Reporter: Ilya Shishkov
> Priority: Minor
> Labels: iep-43, ise
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
