You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@ambari.apache.org by "Sumit Mohanty (JIRA)" <ji...@apache.org> on 2013/02/19 21:53:13 UTC
[jira] [Updated] (AMBARI-1425) Upgrading Stack from 1.2.0/1.2.1 to
1.3.0.
[ https://issues.apache.org/jira/browse/AMBARI-1425?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Sumit Mohanty updated AMBARI-1425:
----------------------------------
Description:
h3. Getting Ready Upgrade
* API support for current version of stack and available versions to upgrade
** Store minimum version to upgrade from
** Test for all queries to be performed by FE
* Report current stack version for all host components
** Agent to return the current stack version of all host-components
** In the store, current_stack_version column for a host-component reflects the value returned by the agent
** Agent records the current version using a file that is updated after each successful upgrade
** Support default value to be HDP
** Ability to query all host-components that are not upgraded
h3. Cluster upgrade - BE support
* Upgrade action on cluster
** POST an action to cluster to start upgrade
** Return request id - similar to smoke test actions
** Modify all desired_stack_version in the store
* Validate an upgrade request
** Fresh upgrade
** Some service are not completely upgraded
** Service is upgraded but not started
** Store and compare current stack version for a cluster
* Stage/task plan generation - fresh upgrade
** Schedule upgrade on all host-components
** Schedule start of the service
** Update the cluster version
* Stage/task plan generation - incomplete upgrade
** Schedule upgrade on host-components that are not upgraded
** Schedule start of the service
** Schedule smoke test on started service
* Validating task results
** Check against upgrade of MASTERs only
** Report warning on SLAVE/CLIENT failures
** Check for smoke test failure (evaluate against ignore_failure flag)
* Upgrade action support in server and agent
** Server side definition of the task and its results
*** Version number as parameter
** Agent side support for upgrading a host-component
*** Puppet integration for downloading the new repo and upgrading
*** Calling pre-install script (python) to perform backup
*** Calling post-install script (python) for restore
h3. Upgrade of individual services
This cost is based primarily on efforts needed to develop service specific python scripts for pre-/post-installation and a common effort required for puppet changes to perform upgrade.
* Upgrade HDFS and MapReduce
* Upgrade ZooKeeper and HBase
* Upgrade Hive and HCatalog
* Upgrade Oozie
* Upgrade WebHCat
* Upgrade Pig
* Upgrade Sqoop
was:
h3. Getting Ready Upgrade
* API support for current version of HDP and available versions to upgrade
** Store minimum version to upgrade from
** Test for all queries to be performed by FE
* Report current HDP version for all host components
** Agent to return the current HDP version of all host-components
** In the store, current_stack_version column for a host-component reflects the value returned by the agent
** Agent records the current version using a file that is updated after each successful upgrade
** Support default value to be HDP
** Ability to query all host-components that are not upgraded
h3. Cluster upgrade - BE support
* Upgrade action on cluster
** POST an action to cluster to start upgrade
** Return request id - similar to smoke test actions
** Modify all desired_stack_version in the store
* Validate an upgrade request
** Fresh upgrade
** Some service are not completely upgraded
** Service is upgraded but not started
** Store and compare current stack version for a cluster
* Stage/task plan generation - fresh upgrade
** Schedule upgrade on all host-components
** Schedule start of the service
** Update the cluster version
* Stage/task plan generation - incomplete upgrade
** Schedule upgrade on host-components that are not upgraded
** Schedule start of the service
** Schedule smoke test on started service
* Validating task results
** Check against upgrade of MASTERs only
** Report warning on SLAVE/CLIENT failures
** Check for smoke test failure (evaluate against ignore_failure flag)
* Upgrade action support in server and agent
** Server side definition of the task and its results
*** Version number as parameter
** Agent side support for upgrading a host-component
*** Puppet integration for downloading the new repo and upgrading
*** Calling pre-install script (python) to perform backup
*** Calling post-install script (python) for restore
h3. Upgrade of individual services
This cost is based primarily on efforts needed to develop service specific python scripts for pre-/post-installation and a common effort required for puppet changes to perform upgrade.
* Upgrade HDFS and MapReduce
* Upgrade ZooKeeper and HBase
* Upgrade Hive and HCatalog
* Upgrade Oozie
* Upgrade WebHCat
* Upgrade Pig
* Upgrade Sqoop
Summary: Upgrading Stack from 1.2.0/1.2.1 to 1.3.0. (was: Upgrading HDP Stack from 1.2.0/1.2.1 to 1.3.0.)
> Upgrading Stack from 1.2.0/1.2.1 to 1.3.0.
> ------------------------------------------
>
> Key: AMBARI-1425
> URL: https://issues.apache.org/jira/browse/AMBARI-1425
> Project: Ambari
> Issue Type: Bug
> Affects Versions: 1.3.0
> Reporter: Sumit Mohanty
> Assignee: Sumit Mohanty
> Fix For: 1.3.0
>
> Attachments: HDP_Upgrade_Using_Ambari-Design.docx
>
>
> h3. Getting Ready Upgrade
> * API support for current version of stack and available versions to upgrade
> ** Store minimum version to upgrade from
> ** Test for all queries to be performed by FE
> * Report current stack version for all host components
> ** Agent to return the current stack version of all host-components
> ** In the store, current_stack_version column for a host-component reflects the value returned by the agent
> ** Agent records the current version using a file that is updated after each successful upgrade
> ** Support default value to be HDP
> ** Ability to query all host-components that are not upgraded
> h3. Cluster upgrade - BE support
> * Upgrade action on cluster
> ** POST an action to cluster to start upgrade
> ** Return request id - similar to smoke test actions
> ** Modify all desired_stack_version in the store
> * Validate an upgrade request
> ** Fresh upgrade
> ** Some service are not completely upgraded
> ** Service is upgraded but not started
> ** Store and compare current stack version for a cluster
> * Stage/task plan generation - fresh upgrade
> ** Schedule upgrade on all host-components
> ** Schedule start of the service
> ** Update the cluster version
> * Stage/task plan generation - incomplete upgrade
> ** Schedule upgrade on host-components that are not upgraded
> ** Schedule start of the service
> ** Schedule smoke test on started service
> * Validating task results
> ** Check against upgrade of MASTERs only
> ** Report warning on SLAVE/CLIENT failures
> ** Check for smoke test failure (evaluate against ignore_failure flag)
> * Upgrade action support in server and agent
> ** Server side definition of the task and its results
> *** Version number as parameter
> ** Agent side support for upgrading a host-component
> *** Puppet integration for downloading the new repo and upgrading
> *** Calling pre-install script (python) to perform backup
> *** Calling post-install script (python) for restore
> h3. Upgrade of individual services
> This cost is based primarily on efforts needed to develop service specific python scripts for pre-/post-installation and a common effort required for puppet changes to perform upgrade.
> * Upgrade HDFS and MapReduce
> * Upgrade ZooKeeper and HBase
> * Upgrade Hive and HCatalog
> * Upgrade Oozie
> * Upgrade WebHCat
> * Upgrade Pig
> * Upgrade Sqoop
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira