ambari-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Yusaku Sako (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (AMBARI-10331) BE issue: install repo version is stuck in INSTALLING after sudden server restart
Date Wed, 15 Apr 2015 20:17:58 GMT

     [ https://issues.apache.org/jira/browse/AMBARI-10331?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Yusaku Sako updated AMBARI-10331:
---------------------------------
    Description: 
Installed the GA bits of HDP 2.2 on 3node cluster on local vagrant.
after the install is completed registered new version - 
http://s3.amazonaws.com/dev.hortonworks.com/HDP/centos6/2.x/BUILDS/2.2.2.0-2538
while the install was in progress machine went down (including ambari-server) and once it
was back up, noticed that install on host1 was completed and other 2 hosts were pending. Progress
bar shows 'in progress' where as there were no background operations running. please see the
screenshot attached. 
the status of progress bar contradicts the bgops status.this could be intermittent as it might
have happened because the machine went down when install was in progress. Please take a look.
Since ambari-server went down, the server was unable to transition to cluster_version from
INSTALLING->INSTALLED. This transition must happen while the version is running. The other
transitions (e.g., UPGRADING->UPGRADED) happen because the agents advertise them.
One way to fix this manually is to modify the DB by updating the cluster_version record from
INSTALLING->OUT_OF_SYNC in order to retry.

  was:
Installed the GA bits of champlain on 3node cluster on local vagrant.
after the install is completed registered new version - 
http://s3.amazonaws.com/dev.hortonworks.com/HDP/centos6/2.x/BUILDS/2.2.2.0-2538
while the install was in progress machine went down (including ambari-server) and once it
was back up, noticed that install on host1 was completed and other 2 hosts were pending. Progress
bar shows 'in progress' where as there were no background operations running. please see the
screenshot attached. 
the status of progress bar contradicts the bgops status.this could be intermittent as it might
have happened because the machine went down when install was in progress. Please take a look.
Since ambari-server went down, the server was unable to transition to cluster_version from
INSTALLING->INSTALLED. This transition must happen while the version is running. The other
transitions (e.g., UPGRADING->UPGRADED) happen because the agents advertise them.
One way to fix this manually is to modify the DB by updating the cluster_version record from
INSTALLING->OUT_OF_SYNC in order to retry.


> BE issue: install repo version is stuck in INSTALLING after sudden server restart
> ---------------------------------------------------------------------------------
>
>                 Key: AMBARI-10331
>                 URL: https://issues.apache.org/jira/browse/AMBARI-10331
>             Project: Ambari
>          Issue Type: Bug
>          Components: ambari-server
>            Reporter: Dmitry Lysnichenko
>            Assignee: Dmitry Lysnichenko
>             Fix For: 2.1.0
>
>         Attachments: AMBARI-10331.patch
>
>
> Installed the GA bits of HDP 2.2 on 3node cluster on local vagrant.
> after the install is completed registered new version - 
> http://s3.amazonaws.com/dev.hortonworks.com/HDP/centos6/2.x/BUILDS/2.2.2.0-2538
> while the install was in progress machine went down (including ambari-server) and once
it was back up, noticed that install on host1 was completed and other 2 hosts were pending.
Progress bar shows 'in progress' where as there were no background operations running. please
see the screenshot attached. 
> the status of progress bar contradicts the bgops status.this could be intermittent as
it might have happened because the machine went down when install was in progress. Please
take a look.
> Since ambari-server went down, the server was unable to transition to cluster_version
from INSTALLING->INSTALLED. This transition must happen while the version is running. The
other transitions (e.g., UPGRADING->UPGRADED) happen because the agents advertise them.
> One way to fix this manually is to modify the DB by updating the cluster_version record
from INSTALLING->OUT_OF_SYNC in order to retry.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message