You are viewing a plain text version of this content. The canonical link for it is here.
Posted to notifications@ignite.apache.org by "alievmirza (via GitHub)" <gi...@apache.org> on 2024/02/18 20:47:18 UTC

[PR] IGNITE-21253 Implement a counter for number of rebalancing tables inside the zone [ignite-3]

alievmirza opened a new pull request, #3225:
URL: https://github.com/apache/ignite-3/pull/3225

   
   Thank you for submitting the pull request.
   
   To streamline the review process of the patch and ensure better code quality
   we ask both an author and a reviewer to verify the following:
   
   ### The Review Checklist
   - [ ] **Formal criteria:** TC status, codestyle, mandatory documentation. Also make sure to complete the following:  
   \- There is a single JIRA ticket related to the pull request.  
   \- The web-link to the pull request is attached to the JIRA ticket.  
   \- The JIRA ticket has the Patch Available state.  
   \- The description of the JIRA ticket explains WHAT was made, WHY and HOW.  
   \- The pull request title is treated as the final commit message. The following pattern must be used: IGNITE-XXXX Change summary where XXXX - number of JIRA issue.
   - [ ] **Design:** new code conforms with the design principles of the components it is added to.
   - [ ] **Patch quality:** patch cannot be split into smaller pieces, its size must be reasonable.
   - [ ] **Code quality:** code is clean and readable, necessary developer documentation is added if needed.
   - [ ] **Tests code quality:** test set covers positive/negative scenarios, happy/edge cases. Tests are effective in terms of execution time and resources.
   
   ### Notes
   - [Apache Ignite Coding Guidelines](https://cwiki.apache.org/confluence/display/IGNITE/Java+Code+Style+Guide)


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: notifications-unsubscribe@ignite.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


Re: [PR] IGNITE-21253 Implement a counter for number of rebalancing tables inside the zone [ignite-3]

Posted by "kgusakov (via GitHub)" <gi...@apache.org>.
kgusakov commented on code in PR #3225:
URL: https://github.com/apache/ignite-3/pull/3225#discussion_r1506390478


##########
modules/distribution-zones/src/main/java/org/apache/ignite/internal/distributionzones/rebalance/RebalanceRaftGroupEventsListener.java:
##########
@@ -413,6 +510,14 @@ private void doOnNewPeersConfigurationApplied(PeersAndLearners configuration) {
                 retryPreconditions = and(retryPreconditions, con5);
             }
 
+            retryPreconditions = and(
+                    or(
+                            notExists(stableChangeTriggerKey(tablePartitionId)),
+                            value(stableChangeTriggerKey(tablePartitionId)).lt(ByteUtils.longToBytes(revision))

Review Comment:
   So, if this condition failed, we will produce the failCase result and schedule new attempt? I think that's not right - we should produce the new result about the stale update and do nothing then. The same trick we already use here https://github.com/apache/ignite-3/blob/c9b1ab9cebed8fc04db90abf18b30f7f0bafb090/modules/distribution-zones/src/main/java/org/apache/ignite/internal/distributionzones/rebalance/RebalanceUtil.java#L191



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: notifications-unsubscribe@ignite.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


Re: [PR] IGNITE-21253 Implement a counter for number of rebalancing tables inside the zone [ignite-3]

Posted by "alievmirza (via GitHub)" <gi...@apache.org>.
alievmirza commented on code in PR #3225:
URL: https://github.com/apache/ignite-3/pull/3225#discussion_r1507402444


##########
modules/distribution-zones/src/main/java/org/apache/ignite/internal/distributionzones/rebalance/RebalanceRaftGroupEventsListener.java:
##########
@@ -413,6 +510,14 @@ private void doOnNewPeersConfigurationApplied(PeersAndLearners configuration) {
                 retryPreconditions = and(retryPreconditions, con5);
             }
 
+            retryPreconditions = and(
+                    or(
+                            notExists(stableChangeTriggerKey(tablePartitionId)),
+                            value(stableChangeTriggerKey(tablePartitionId)).lt(ByteUtils.longToBytes(revision))

Review Comment:
   Added



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: notifications-unsubscribe@ignite.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


Re: [PR] IGNITE-21253 Implement a counter for number of rebalancing tables inside the zone [ignite-3]

Posted by "alievmirza (via GitHub)" <gi...@apache.org>.
alievmirza commented on code in PR #3225:
URL: https://github.com/apache/ignite-3/pull/3225#discussion_r1507403088


##########
modules/distribution-zones/src/main/java/org/apache/ignite/internal/distributionzones/rebalance/DistributionZoneRebalanceEngine.java:
##########
@@ -214,6 +242,85 @@ public void onError(Throwable e) {
         };
     }
 
+    /**
+     * Creates watch listener for tables from a zone. This counter is needed for tracking rebalances for a specified partition along all
+     * tables. Once all rebalances for specified partition number are finished (meaning that counter is equal to 0), we can perform stable
+     * switch for all stable keys from meta storage.
+     *
+     * @return Corresponding watch listener.
+     */
+    private WatchListener createPartitionsCounterListener() {
+        return new WatchListener() {
+            @Override
+            public CompletableFuture<Void> onUpdate(WatchEvent event) {
+                return IgniteUtils.inBusyLockAsync(busyLock, () -> {
+
+                    int counter = ((Set<Integer>) fromBytes(event.entryEvent().newEntry().value())).size();
+
+                    assert counter >= 0 : "Tables counter for rabalances cannot be negative.";
+
+                    if (counter > 0) {
+                        return nullCompletedFuture();
+                    }
+
+                    int zoneId = RebalanceUtil.extractZoneIdFromTablesCounter(event.entryEvent().newEntry().key());
+
+                    List<CatalogTableDescriptor> tables = findTablesByZoneId(zoneId, catalogService.latestCatalogVersion(), catalogService);
+
+                    rebalanceScheduler.schedule(() -> {
+                        if (!busyLock.enterBusy()) {
+                            return;
+                        }
+
+                        LOG.info("Started to update stable keys for tables from the zone [zoneId = {}, tables = [{}]]",
+                                zoneId,
+                                tables.stream().map(CatalogObjectDescriptor::name).collect(Collectors.toSet())
+                        );
+
+                        try {
+                            Map<ByteArray, TablePartitionId> partitionTablesKeys = new HashMap<>();
+
+                            int partId = extractPartitionNumber(event.entryEvent().newEntry().key());
+
+                            for (CatalogTableDescriptor table : tables) {
+                                TablePartitionId replicaGrpId = new TablePartitionId(table.id(), partId);
+                                partitionTablesKeys.put(raftConfigurationAppliedKey(replicaGrpId), replicaGrpId);
+                            }
+
+                            Map<ByteArray, Entry> entriesMap = metaStorageManager.getAll(partitionTablesKeys.keySet()).get();
+
+                            entriesMap.forEach((key, stable) -> {
+                                doOnNewPeersConfigurationApplied(

Review Comment:
   I've added check for revision and we skip outdated events



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: notifications-unsubscribe@ignite.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


Re: [PR] IGNITE-21253 Implement a counter for number of rebalancing tables inside the zone [ignite-3]

Posted by "kgusakov (via GitHub)" <gi...@apache.org>.
kgusakov commented on code in PR #3225:
URL: https://github.com/apache/ignite-3/pull/3225#discussion_r1502949601


##########
modules/distribution-zones/src/main/java/org/apache/ignite/internal/distributionzones/rebalance/DistributionZoneRebalanceEngine.java:
##########
@@ -214,6 +242,85 @@ public void onError(Throwable e) {
         };
     }
 
+    /**
+     * Creates watch listener for tables from a zone. This counter is needed for tracking rebalances for a specified partition along all
+     * tables. Once all rebalances for specified partition number are finished (meaning that counter is equal to 0), we can perform stable
+     * switch for all stable keys from meta storage.
+     *
+     * @return Corresponding watch listener.
+     */
+    private WatchListener createPartitionsCounterListener() {
+        return new WatchListener() {
+            @Override
+            public CompletableFuture<Void> onUpdate(WatchEvent event) {
+                return IgniteUtils.inBusyLockAsync(busyLock, () -> {
+
+                    int counter = ((Set<Integer>) fromBytes(event.entryEvent().newEntry().value())).size();
+
+                    assert counter >= 0 : "Tables counter for rabalances cannot be negative.";
+
+                    if (counter > 0) {
+                        return nullCompletedFuture();
+                    }
+
+                    int zoneId = RebalanceUtil.extractZoneIdFromTablesCounter(event.entryEvent().newEntry().key());
+
+                    List<CatalogTableDescriptor> tables = findTablesByZoneId(zoneId, catalogService.latestCatalogVersion(), catalogService);

Review Comment:
   Why the `catalogService.latestCatalogVersion()`? What about the stale counter updates? Will it work as expected?



##########
modules/distribution-zones/src/main/java/org/apache/ignite/internal/distributionzones/rebalance/DistributionZoneRebalanceEngine.java:
##########
@@ -214,6 +242,85 @@ public void onError(Throwable e) {
         };
     }
 
+    /**
+     * Creates watch listener for tables from a zone. This counter is needed for tracking rebalances for a specified partition along all
+     * tables. Once all rebalances for specified partition number are finished (meaning that counter is equal to 0), we can perform stable
+     * switch for all stable keys from meta storage.
+     *
+     * @return Corresponding watch listener.
+     */
+    private WatchListener createPartitionsCounterListener() {
+        return new WatchListener() {
+            @Override
+            public CompletableFuture<Void> onUpdate(WatchEvent event) {
+                return IgniteUtils.inBusyLockAsync(busyLock, () -> {
+
+                    int counter = ((Set<Integer>) fromBytes(event.entryEvent().newEntry().value())).size();
+
+                    assert counter >= 0 : "Tables counter for rabalances cannot be negative.";
+
+                    if (counter > 0) {
+                        return nullCompletedFuture();
+                    }
+
+                    int zoneId = RebalanceUtil.extractZoneIdFromTablesCounter(event.entryEvent().newEntry().key());
+
+                    List<CatalogTableDescriptor> tables = findTablesByZoneId(zoneId, catalogService.latestCatalogVersion(), catalogService);
+
+                    rebalanceScheduler.schedule(() -> {
+                        if (!busyLock.enterBusy()) {
+                            return;
+                        }
+
+                        LOG.info("Started to update stable keys for tables from the zone [zoneId = {}, tables = [{}]]",

Review Comment:
   The list of tables can be huge in production and in some kind of benchmark tests - it will produce huge log record. Maybe move the tables list to debug level at least?



##########
modules/distribution-zones/src/main/java/org/apache/ignite/internal/distributionzones/rebalance/DistributionZoneRebalanceEngine.java:
##########
@@ -214,6 +242,85 @@ public void onError(Throwable e) {
         };
     }
 
+    /**
+     * Creates watch listener for tables from a zone. This counter is needed for tracking rebalances for a specified partition along all
+     * tables. Once all rebalances for specified partition number are finished (meaning that counter is equal to 0), we can perform stable
+     * switch for all stable keys from meta storage.
+     *
+     * @return Corresponding watch listener.
+     */
+    private WatchListener createPartitionsCounterListener() {
+        return new WatchListener() {
+            @Override
+            public CompletableFuture<Void> onUpdate(WatchEvent event) {
+                return IgniteUtils.inBusyLockAsync(busyLock, () -> {
+
+                    int counter = ((Set<Integer>) fromBytes(event.entryEvent().newEntry().value())).size();
+
+                    assert counter >= 0 : "Tables counter for rabalances cannot be negative.";
+
+                    if (counter > 0) {
+                        return nullCompletedFuture();
+                    }
+
+                    int zoneId = RebalanceUtil.extractZoneIdFromTablesCounter(event.entryEvent().newEntry().key());
+
+                    List<CatalogTableDescriptor> tables = findTablesByZoneId(zoneId, catalogService.latestCatalogVersion(), catalogService);
+
+                    rebalanceScheduler.schedule(() -> {
+                        if (!busyLock.enterBusy()) {
+                            return;
+                        }
+
+                        LOG.info("Started to update stable keys for tables from the zone [zoneId = {}, tables = [{}]]",
+                                zoneId,
+                                tables.stream().map(CatalogObjectDescriptor::name).collect(Collectors.toSet())
+                        );
+
+                        try {
+                            Map<ByteArray, TablePartitionId> partitionTablesKeys = new HashMap<>();
+
+                            int partId = extractPartitionNumber(event.entryEvent().newEntry().key());
+
+                            for (CatalogTableDescriptor table : tables) {
+                                TablePartitionId replicaGrpId = new TablePartitionId(table.id(), partId);
+                                partitionTablesKeys.put(raftConfigurationAppliedKey(replicaGrpId), replicaGrpId);
+                            }
+
+                            Map<ByteArray, Entry> entriesMap = metaStorageManager.getAll(partitionTablesKeys.keySet()).get();
+
+                            entriesMap.forEach((key, stable) -> {
+                                doOnNewPeersConfigurationApplied(

Review Comment:
   Could you check the following case: we have N nodes, which handled some rebalances successfully, but 1 node didn't process any counter updates yet.
   
   At the moment, this node started to process the counter update events - so, it can call the keys update (potentially with the same stable keys) and finish the current rebalance (move planned->pending->stable and etc.) because of the 0-value counter of old stale rebalance. 



##########
modules/table/src/main/java/org/apache/ignite/internal/table/distributed/TableManager.java:
##########
@@ -432,7 +440,8 @@ public TableManager(
             HybridTimestampTracker observableTimestampTracker,
             PlacementDriver placementDriver,
             Supplier<IgniteSql> sql,
-            FailureProcessor failureProcessor
+            FailureProcessor failureProcessor,
+            ScheduledExecutorService rebalanceScheduler

Review Comment:
   javadoc must be updated



##########
modules/table/src/main/java/org/apache/ignite/internal/table/distributed/TableManager.java:
##########
@@ -1778,9 +1765,7 @@ private CompletableFuture<Void> handleChangePendingAssignmentEvent(
                         // Table can be null only recovery, because we use a revision from the future. See comment inside
                         // performRebalanceOnRecovery.
                         if (table == null) {
-                            if (LOG.isInfoEnabled()) {
-                                LOG.info("Skipping Pending Assignments update, because table {} does not exist", tblId);
-                            }
+                            assert false : "Table cannot be null.";

Review Comment:
   Why? We have a possible race between the table delete and new rebalance scheduling.



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: notifications-unsubscribe@ignite.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


Re: [PR] IGNITE-21253 Implement a counter for number of rebalancing tables inside the zone [ignite-3]

Posted by "alievmirza (via GitHub)" <gi...@apache.org>.
alievmirza commented on code in PR #3225:
URL: https://github.com/apache/ignite-3/pull/3225#discussion_r1507403088


##########
modules/distribution-zones/src/main/java/org/apache/ignite/internal/distributionzones/rebalance/DistributionZoneRebalanceEngine.java:
##########
@@ -214,6 +242,85 @@ public void onError(Throwable e) {
         };
     }
 
+    /**
+     * Creates watch listener for tables from a zone. This counter is needed for tracking rebalances for a specified partition along all
+     * tables. Once all rebalances for specified partition number are finished (meaning that counter is equal to 0), we can perform stable
+     * switch for all stable keys from meta storage.
+     *
+     * @return Corresponding watch listener.
+     */
+    private WatchListener createPartitionsCounterListener() {
+        return new WatchListener() {
+            @Override
+            public CompletableFuture<Void> onUpdate(WatchEvent event) {
+                return IgniteUtils.inBusyLockAsync(busyLock, () -> {
+
+                    int counter = ((Set<Integer>) fromBytes(event.entryEvent().newEntry().value())).size();
+
+                    assert counter >= 0 : "Tables counter for rabalances cannot be negative.";
+
+                    if (counter > 0) {
+                        return nullCompletedFuture();
+                    }
+
+                    int zoneId = RebalanceUtil.extractZoneIdFromTablesCounter(event.entryEvent().newEntry().key());
+
+                    List<CatalogTableDescriptor> tables = findTablesByZoneId(zoneId, catalogService.latestCatalogVersion(), catalogService);
+
+                    rebalanceScheduler.schedule(() -> {
+                        if (!busyLock.enterBusy()) {
+                            return;
+                        }
+
+                        LOG.info("Started to update stable keys for tables from the zone [zoneId = {}, tables = [{}]]",
+                                zoneId,
+                                tables.stream().map(CatalogObjectDescriptor::name).collect(Collectors.toSet())
+                        );
+
+                        try {
+                            Map<ByteArray, TablePartitionId> partitionTablesKeys = new HashMap<>();
+
+                            int partId = extractPartitionNumber(event.entryEvent().newEntry().key());
+
+                            for (CatalogTableDescriptor table : tables) {
+                                TablePartitionId replicaGrpId = new TablePartitionId(table.id(), partId);
+                                partitionTablesKeys.put(raftConfigurationAppliedKey(replicaGrpId), replicaGrpId);
+                            }
+
+                            Map<ByteArray, Entry> entriesMap = metaStorageManager.getAll(partitionTablesKeys.keySet()).get();
+
+                            entriesMap.forEach((key, stable) -> {
+                                doOnNewPeersConfigurationApplied(

Review Comment:
   I've added check for revision and we skipp outdared events



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: notifications-unsubscribe@ignite.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


Re: [PR] IGNITE-21253 Implement a counter for number of rebalancing tables inside the zone [ignite-3]

Posted by "alievmirza (via GitHub)" <gi...@apache.org>.
alievmirza commented on code in PR #3225:
URL: https://github.com/apache/ignite-3/pull/3225#discussion_r1507402569


##########
modules/table/src/main/java/org/apache/ignite/internal/table/distributed/TableManager.java:
##########
@@ -432,7 +440,8 @@ public TableManager(
             HybridTimestampTracker observableTimestampTracker,
             PlacementDriver placementDriver,
             Supplier<IgniteSql> sql,
-            FailureProcessor failureProcessor
+            FailureProcessor failureProcessor,
+            ScheduledExecutorService rebalanceScheduler

Review Comment:
   done



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: notifications-unsubscribe@ignite.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


Re: [PR] IGNITE-21253 Implement a counter for number of rebalancing tables inside the zone [ignite-3]

Posted by "vldpyatkov (via GitHub)" <gi...@apache.org>.
vldpyatkov merged PR #3225:
URL: https://github.com/apache/ignite-3/pull/3225


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: notifications-unsubscribe@ignite.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


Re: [PR] IGNITE-21253 Implement a counter for number of rebalancing tables inside the zone [ignite-3]

Posted by "alievmirza (via GitHub)" <gi...@apache.org>.
alievmirza commented on code in PR #3225:
URL: https://github.com/apache/ignite-3/pull/3225#discussion_r1507403328


##########
modules/distribution-zones/src/main/java/org/apache/ignite/internal/distributionzones/rebalance/DistributionZoneRebalanceEngine.java:
##########
@@ -214,6 +242,85 @@ public void onError(Throwable e) {
         };
     }
 
+    /**
+     * Creates watch listener for tables from a zone. This counter is needed for tracking rebalances for a specified partition along all
+     * tables. Once all rebalances for specified partition number are finished (meaning that counter is equal to 0), we can perform stable
+     * switch for all stable keys from meta storage.
+     *
+     * @return Corresponding watch listener.
+     */
+    private WatchListener createPartitionsCounterListener() {
+        return new WatchListener() {
+            @Override
+            public CompletableFuture<Void> onUpdate(WatchEvent event) {
+                return IgniteUtils.inBusyLockAsync(busyLock, () -> {
+
+                    int counter = ((Set<Integer>) fromBytes(event.entryEvent().newEntry().value())).size();
+
+                    assert counter >= 0 : "Tables counter for rabalances cannot be negative.";
+
+                    if (counter > 0) {
+                        return nullCompletedFuture();
+                    }
+
+                    int zoneId = RebalanceUtil.extractZoneIdFromTablesCounter(event.entryEvent().newEntry().key());
+
+                    List<CatalogTableDescriptor> tables = findTablesByZoneId(zoneId, catalogService.latestCatalogVersion(), catalogService);
+
+                    rebalanceScheduler.schedule(() -> {
+                        if (!busyLock.enterBusy()) {
+                            return;
+                        }
+
+                        LOG.info("Started to update stable keys for tables from the zone [zoneId = {}, tables = [{}]]",

Review Comment:
   done



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: notifications-unsubscribe@ignite.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


Re: [PR] IGNITE-21253 Implement a counter for number of rebalancing tables inside the zone [ignite-3]

Posted by "alievmirza (via GitHub)" <gi...@apache.org>.
alievmirza commented on code in PR #3225:
URL: https://github.com/apache/ignite-3/pull/3225#discussion_r1507403896


##########
modules/distribution-zones/src/main/java/org/apache/ignite/internal/distributionzones/rebalance/DistributionZoneRebalanceEngine.java:
##########
@@ -214,6 +242,85 @@ public void onError(Throwable e) {
         };
     }
 
+    /**
+     * Creates watch listener for tables from a zone. This counter is needed for tracking rebalances for a specified partition along all
+     * tables. Once all rebalances for specified partition number are finished (meaning that counter is equal to 0), we can perform stable
+     * switch for all stable keys from meta storage.
+     *
+     * @return Corresponding watch listener.
+     */
+    private WatchListener createPartitionsCounterListener() {
+        return new WatchListener() {
+            @Override
+            public CompletableFuture<Void> onUpdate(WatchEvent event) {
+                return IgniteUtils.inBusyLockAsync(busyLock, () -> {
+
+                    int counter = ((Set<Integer>) fromBytes(event.entryEvent().newEntry().value())).size();
+
+                    assert counter >= 0 : "Tables counter for rabalances cannot be negative.";
+
+                    if (counter > 0) {
+                        return nullCompletedFuture();
+                    }
+
+                    int zoneId = RebalanceUtil.extractZoneIdFromTablesCounter(event.entryEvent().newEntry().key());
+
+                    List<CatalogTableDescriptor> tables = findTablesByZoneId(zoneId, catalogService.latestCatalogVersion(), catalogService);

Review Comment:
   yes, because we are in the ms handler 



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: notifications-unsubscribe@ignite.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


Re: [PR] IGNITE-21253 Implement a counter for number of rebalancing tables inside the zone [ignite-3]

Posted by "alievmirza (via GitHub)" <gi...@apache.org>.
alievmirza commented on code in PR #3225:
URL: https://github.com/apache/ignite-3/pull/3225#discussion_r1507403997


##########
modules/table/src/main/java/org/apache/ignite/internal/table/distributed/TableManager.java:
##########
@@ -1778,9 +1765,7 @@ private CompletableFuture<Void> handleChangePendingAssignmentEvent(
                         // Table can be null only recovery, because we use a revision from the future. See comment inside
                         // performRebalanceOnRecovery.
                         if (table == null) {
-                            if (LOG.isInfoEnabled()) {
-                                LOG.info("Skipping Pending Assignments update, because table {} does not exist", tblId);
-                            }
+                            assert false : "Table cannot be null.";

Review Comment:
   reverted



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: notifications-unsubscribe@ignite.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org