-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/55230/#review160635
-----------------------------------------------------------




ambari-server/src/main/resources/common-services/HDFS/2.1.0.2.0/package/scripts/hdfs_namenode.py
 (line 125)
<https://reviews.apache.org/r/55230/#comment231773>

    Most changes are like this one - just trying to use constants...



ambari-server/src/main/resources/common-services/HDFS/3.0.0.3.0/package/scripts/namenode.py
 (lines 199 - 204)
<https://reviews.apache.org/r/55230/#comment231772>

    This logic was weird, but it caused conf-select to not be called for HOU



ambari-server/src/main/resources/stacks/HDP/2.5/upgrades/host-ordered-upgrade.xml
 
<https://reviews.apache.org/r/55230/#comment231771>

    This, and the one below it, were the biggest causes of the problem. NN 
restarts during HOU should be treated as normal restarts - no need to place NN 
into a "mode" or even create a marker file.


- Jonathan Hurley


On Jan. 5, 2017, 4:47 p.m., Jonathan Hurley wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/55230/
> -----------------------------------------------------------
> 
> (Updated Jan. 5, 2017, 4:47 p.m.)
> 
> 
> Review request for Ambari, Dmytro Grinenko and Nate Cole.
> 
> 
> Bugs: AMBARI-19344
>     https://issues.apache.org/jira/browse/AMBARI-19344
> 
> 
> Repository: ambari
> 
> 
> Description
> -------
> 
> When restarting NameNode during an HOU, NN throws an error because it was 
> improperly put into rollingUpgrade mode.
> 
> *Steps*
> # Deploy HDP-2.5.0.0 with Ambari-2.5.0.0-547 build (non-HA cluster)
> # Start Host Ordered Upgrade to HDP-2.5.3
>  
> *Result:* Error at pre Upgrade HDFS step
> 
> ```
> PREPARE rolling upgrade ...
> 17/01/03 17:07:51 WARN retry.RetryInvocationHandler: Exception while invoking 
> ClientNamenodeProtocolTranslatorPB.rollingUpgrade over null. Not retrying 
> because try once and fail.
> org.apache.hadoop.ipc.RemoteException(java.io.IOException): Safe mode should 
> be turned ON in order to create namespace image.
>       at 
> org.apache.hadoop.hdfs.server.namenode.FSNamesystem.startRollingUpgradeInternalForNonHA(FSNamesystem.java:7839)
> 
> rollingUpgrade: Safe mode should be turned ON in order to create namespace 
> image.
> ```
> 
> 
> Diffs
> -----
> 
>   
> ambari-server/src/main/resources/common-services/HDFS/2.1.0.2.0/package/scripts/hdfs_namenode.py
>  23119f0 
>   
> ambari-server/src/main/resources/common-services/HDFS/2.1.0.2.0/package/scripts/namenode.py
>  86f68e5 
>   
> ambari-server/src/main/resources/common-services/HDFS/3.0.0.3.0/package/scripts/hdfs_namenode.py
>  23119f0 
>   
> ambari-server/src/main/resources/common-services/HDFS/3.0.0.3.0/package/scripts/namenode.py
>  86f68e5 
>   
> ambari-server/src/main/resources/stacks/HDP/2.5/upgrades/host-ordered-upgrade.xml
>  f6480bf 
>   
> ambari-server/src/main/resources/stacks/HDP/2.6/upgrades/host-ordered-upgrade.xml
>  1b29af3 
>   ambari-server/src/test/python/stacks/2.0.6/HDFS/test_namenode.py 2627ddf 
> 
> Diff: https://reviews.apache.org/r/55230/diff/
> 
> 
> Testing
> -------
> 
> ----------------------------------------------------------------------
> Total run:1157
> Total errors:0
> Total failures:0
> OK
> [INFO] 
> ------------------------------------------------------------------------
> [INFO] BUILD SUCCESS
> [INFO] 
> ------------------------------------------------------------------------
> [INFO] Total time: 50.822 s
> [INFO] Finished at: 2017-01-05T16:29:21-05:00
> [INFO] Final Memory: 17M/310M
> [INFO] 
> ------------------------------------------------------------------------
> 
> 
> Thanks,
> 
> Jonathan Hurley
> 
>

Reply via email to