[
https://issues.apache.org/jira/browse/IGNITE-18022?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Kirill Tkalenko updated IGNITE-18022:
-------------------------------------
Description:
Suggested API for a full rebalance for
*org.apache.ignite.internal.tx.storage.state.TxStateStorage*:
{code:java}
public interface TxStateStorage extends ManuallyCloseable {
....
long FULL_REBALANCE_IN_PROGRESS = -1;
/**
* Prepares the transaction state storage for a full rebalance: clears the
storage, sets the {@link #lastAppliedIndex()} to
* {@link #FULL_REBALANCE_IN_PROGRESS}, and closes all cursors.
*
* <p>After the complete of future, only write methods will be available,
and read methods will throw {@link IgniteInternalException}.
*
* <p>This method must be called before every full rebalance of transaction
state storage and ends with a call to one of the methods:
* <ul>
* <li>{@link #abortFullRebalance()} - in case of errors or
cancellation of a full rebalance;</li>
* <li>{@link #finishFullRebalance(long, long)} - in case of successful
completion of a full rebalance.</li>
* </ul>
*
* <p>If the {@link #lastAppliedIndex()} is {@link
#FULL_REBALANCE_IN_PROGRESS} after a node restart, then the storage needs to be
* cleared before it can be used.
*
* @return Future of the start a full rebalance for transaction state
storage.
*/
CompletableFuture<Void> startFullRebalance();
/**
* Aborts a full rebalance for transaction state storage: clears the
storage, sets the {@link #lastAppliedIndex()} to {@code 0}.
*
* <p>After the complete of future, methods for writing and reading will be
available.
*
* <p>If a full rebalance has not started, then nothing will happen.
*
* @return Future of the abort a full rebalance for transaction state
storage.
*/
CompletableFuture<Void> abortFullRebalance();
/**
* Completes a full rebalance for transaction state storage: updates {@link
#lastAppliedIndex()} and {@link #lastAppliedTerm()}.
*
* <p>After the complete of future, methods for writing and reading will be
available.
*
* <p>If a full rebalance has not started, then nothing will happen.
*
* @param lastAppliedIndex Last applied index.
* @param lastAppliedTerm Last applied term.
* @return Future of the finish a full rebalance for transaction state
storage.
*/
CompletableFuture<Void> finishFullRebalance(long lastAppliedIndex, long
lastAppliedTerm);
}
{code}
We will also need to write tests and make a test implementation.
The main idea is described in IGNITE-17990.
was:
We need to create an API for a complete rebalance without losing user data in
class *TxStateStorage* and *TxStateTableStorage* on receiver.
And also to make a test implementation and basic tests.
The main idea is described in IGNITE-17990.
> Creating an API for full rebalance without losing user data in TxStateStorage
> on receiver
> -----------------------------------------------------------------------------------------
>
> Key: IGNITE-18022
> URL: https://issues.apache.org/jira/browse/IGNITE-18022
> Project: Ignite
> Issue Type: Task
> Reporter: Kirill Tkalenko
> Assignee: Kirill Tkalenko
> Priority: Major
> Labels: ignite-3
> Fix For: 3.0.0-beta2
>
> Time Spent: 1.5h
> Remaining Estimate: 0h
>
> Suggested API for a full rebalance for
> *org.apache.ignite.internal.tx.storage.state.TxStateStorage*:
> {code:java}
> public interface TxStateStorage extends ManuallyCloseable {
> ....
> long FULL_REBALANCE_IN_PROGRESS = -1;
> /**
> * Prepares the transaction state storage for a full rebalance: clears
> the storage, sets the {@link #lastAppliedIndex()} to
> * {@link #FULL_REBALANCE_IN_PROGRESS}, and closes all cursors.
> *
> * <p>After the complete of future, only write methods will be available,
> and read methods will throw {@link IgniteInternalException}.
> *
> * <p>This method must be called before every full rebalance of
> transaction state storage and ends with a call to one of the methods:
> * <ul>
> * <li>{@link #abortFullRebalance()} - in case of errors or
> cancellation of a full rebalance;</li>
> * <li>{@link #finishFullRebalance(long, long)} - in case of
> successful completion of a full rebalance.</li>
> * </ul>
> *
> * <p>If the {@link #lastAppliedIndex()} is {@link
> #FULL_REBALANCE_IN_PROGRESS} after a node restart, then the storage needs to
> be
> * cleared before it can be used.
> *
> * @return Future of the start a full rebalance for transaction state
> storage.
> */
> CompletableFuture<Void> startFullRebalance();
> /**
> * Aborts a full rebalance for transaction state storage: clears the
> storage, sets the {@link #lastAppliedIndex()} to {@code 0}.
> *
> * <p>After the complete of future, methods for writing and reading will
> be available.
> *
> * <p>If a full rebalance has not started, then nothing will happen.
> *
> * @return Future of the abort a full rebalance for transaction state
> storage.
> */
> CompletableFuture<Void> abortFullRebalance();
> /**
> * Completes a full rebalance for transaction state storage: updates
> {@link #lastAppliedIndex()} and {@link #lastAppliedTerm()}.
> *
> * <p>After the complete of future, methods for writing and reading will
> be available.
> *
> * <p>If a full rebalance has not started, then nothing will happen.
> *
> * @param lastAppliedIndex Last applied index.
> * @param lastAppliedTerm Last applied term.
> * @return Future of the finish a full rebalance for transaction state
> storage.
> */
> CompletableFuture<Void> finishFullRebalance(long lastAppliedIndex, long
> lastAppliedTerm);
> }
> {code}
> We will also need to write tests and make a test implementation.
> The main idea is described in IGNITE-17990.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
