VMware GemFire Error: Failed to update cluster configuration for cluster Region/RegionNameXXXX not found

Products

• VMware Tanzu Gemfire

Issue/Introduction

When you define regions under "groups", "gfsh destroy" or "gfsh create index" may fail to update cluster configuration with the error "Failed to update cluster configuration for cluster. Region /RegionNameXXXX not found," even  though it did destroy or create the index eventually.

For example:

gfsh>start locator --port=10334
gfsh>configure pdx --read-serialized=true --auto-serializable-classes=".*"
gfsh>import cluster-configuration --zip-file-name=./cluster.zip
gfsh>start server --server-port=7901 --locators=localhost[10334] --include-system-classpath=true --dir=cacheserver1 --groups=regionGroup1

gfsh>destroy index --name=indexname1 --region=/exampleRegion
                    Member                      | Status | Message
----------------------------------------------- | ------ | ----------------------------------------------
192.168.10.106(zoom-calm-sigma:31903)<v1>:41001 | OK     | Destroyed index indexname1 on region exampleRegion

failed to update cluster config for cluster. Reason: Region exampleRegion not found

gfsh>create index --name=indexname1 --expression="exampleExpression" --region="/exampleRegion" --type=range
                    Member                      | Status | Message
----------------------------------------------- | ------ | --------------------------
192.168.10.106(zoom-calm-sigma:31903)<v1>:41001 | OK     | Index successfully created

failed to update cluster config for cluster. Reason: Region /exampleRegion not found.

In the locator log, there is some exception stack as in the log below:

[error 2020/02/24 18:20:22.640 JST jump-alive-hall <RMI TCP Connection(5)-192.168.10.106> tid=0x63] failed to update cluster config for cluster
org.apache.geode.management.internal.cli.exceptions.EntityNotFoundException: Region UnitTelemetry not found
	at org.apache.geode.management.internal.cli.commands.DestroyIndexCommand.updateClusterConfig(DestroyIndexCommand.java:106)
	at org.apache.geode.management.internal.cli.remote.CommandExecutor.lambda$invokeCommand$0(CommandExecutor.java:141)
	at org.apache.geode.distributed.internal.InternalConfigurationPersistenceService.updateCacheConfig(InternalConfigurationPersistenceService.java:975)
	at org.apache.geode.management.internal.cli.remote.CommandExecutor.invokeCommand(CommandExecutor.java:139)
	at org.apache.geode.management.internal.cli.remote.CommandExecutor.execute(CommandExecutor.java:54)
	at org.apache.geode.management.internal.cli.remote.CommandExecutor.execute(CommandExecutor.java:48)
	at org.apache.geode.management.internal.cli.remote.OnlineCommandProcessor.executeCommand(OnlineCommandProcessor.java:123)
	at org.apache.geode.management.internal.beans.MemberMBeanBridge.processCommand(MemberMBeanBridge.java:1505)
	at org.apache.geode.management.internal.beans.MemberMBean.processCommand(MemberMBean.java:417)
......

Environment

Product Version: 9.7

Resolution

Since you define this cluster's regions under "groups", "gfsh destroy" or "gfsh create index," commands hit a known issue in VMware Gemfire's cluster configuration update which is not fixed yet.

The quick workaround for this issue is to destroy/create the index by adding "--groups=regionGroup1" option to your “create index” and “destroy index” commands.

For example:

destroy index --name=indexname1 --region=/exampleRegion --groups=regionGroup1

create index --name=indexname1 --expression="exampleExpression" --region="/exampleRegion" --type=range --groups=regionGroup1