Support Questions
Find answers, ask questions, and share your expertise

Namenode High Availability : BootstrapStandby failed with QJM

Highlighted

Namenode High Availability : BootstrapStandby failed with QJM

New Contributor

Hi everyone,

 

I ran into a problem when deploying CDH4.4 High Availability using the linux packages. Following the steps described into the cloudera documentation http://www.cloudera.com/content/cloudera-content/cloudera-docs/CDH4/latest/CDH4-High-Availability-Gu... once all wonfigured, I tried to synchronised the standby Namenode by running the command sudo -u hdfs hdfs namenode -bootstrapStandby. This never worked, even after running hdfs namenode -initializeSharedEdits or even sudo -u hdfs hdfs dfsadmin -saveNamespace.

The issue I get seems really closed to the Apache issue https://issues.apache.org/jira/browse/HDFS-5080. A very similar issue was said to be corrected in CDH4.2 : https://issues.apache.org/jira/browse/HDFS-3752.

 

Here is the end of the output from the bootstrapStandby command :

13/10/16 14:33:19 INFO namenode.NNStorage: Storage directory /HDFS/hdfs-name has been successfully formatted.
13/10/16 14:33:19 WARN common.Util: Path /HDFS/hdfs-name should be specified as a URI in configuration files. Please update hdfs configuration.
13/10/16 14:33:19 WARN common.Util: Path /HDFS/hdfs-name should be specified as a URI in configuration files. Please update hdfs configuration.
13/10/16 14:33:19 FATAL ha.BootstrapStandby: Unable to read transaction ids 1169-1169 from the configured shared edits storage qjournal://node-admin.test.saga:8485;master0.test.saga:8485;master1.test.saga:8485/mycluster. Please copy these logs into the shared edits storage or call saveNamespace on the active node.
Error: Gap in transactions. Expected to be able to read up until at least txid 1169 but unable to find any edit logs containing txid 1169
13/10/16 14:33:19 INFO util.ExitUtil: Exiting with status 6

 

The only solution I found is to rsync the active namenode metadata directory.

 

Thank you for any input you could give.

3 REPLIES 3
Highlighted

Re: Namenode High Availability : BootstrapStandby failed with QJM

New Contributor

same problem?  anyone help?

Re: Namenode High Availability : BootstrapStandby failed with QJM

Cloudera Employee

This sounds like HDFS-5080, which is going to be fixed in CDH4.5.  On earlier versions, you can either use rsync, or shut down the active NN and use bootstrapStandby to work around the issue.

Highlighted

Re: Namenode High Availability : BootstrapStandby failed with QJM

Explorer

we had the same issue with CDH 4.3.

 

Now we upgraded to 4.5 and i can confirm that it seems to work now.

As workaround you can simply use scp or rsync as suggested to copy the data. Worked for us.

 

 

 

Don't have an account?