cloudstack-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "ASF subversion and git services (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (CLOUDSTACK-9738) Optimize vm expunge process for instances with vm snapshots
Date Wed, 08 Feb 2017 06:59:41 GMT

    [ https://issues.apache.org/jira/browse/CLOUDSTACK-9738?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15857526#comment-15857526
] 

ASF subversion and git services commented on CLOUDSTACK-9738:
-------------------------------------------------------------

Commit e303baa1bf10e464d55b078d39cf3f42de1dc2c0 in cloudstack's branch refs/heads/master from
[~rajanik]
[ https://git-wip-us.apache.org/repos/asf?p=cloudstack.git;h=e303baa ]

Merge pull request #1905 from nvazquez/expungeVmRefactor

CLOUDSTACK-9738: [Vmware] Optimize vm expunge process for instances with vm snapshots## Description
It was noticed that expunging instances with many vm snapshots took a look of time, as hypervisor
received as many tasks as vm snapshots instance had, apart from the delete vm task. We propose
a way to optimize this process for instances with vm snapshots by sending only one delete
task to hypervisor, which will delete vm and its snapshots

## Use cases

1. deleteVMsnapohsot-> no changes to current behavior
2. destroyVM with expunge=false ->  no actions to VMsnaphsot is performed at the moment.
When VM cleanup thread is executed it will perform the same sequence as (3). If instance is
recovered before expunged by the cleanup thread it will remain intact with VMSnapshot chain
present
3. destroyVM with expunge=true:
   * Vmsnaphsot is marked with removed timestamp and state = Expunging in DB
   * VM is deleted in HW

* pr/1905:
  CLOUDSTACK-9738: [Vmware] Optimize vm expunge process for instances with vm snapshots

Signed-off-by: Rajani Karuturi <rajani.karuturi@accelerite.com>


> Optimize vm expunge process for instances with vm snapshots
> -----------------------------------------------------------
>
>                 Key: CLOUDSTACK-9738
>                 URL: https://issues.apache.org/jira/browse/CLOUDSTACK-9738
>             Project: CloudStack
>          Issue Type: Improvement
>      Security Level: Public(Anyone can view this level - this is the default.) 
>    Affects Versions: 4.10.0.0
>            Reporter: Nicolas Vazquez
>            Assignee: Nicolas Vazquez
>             Fix For: 4.10.0.0
>
>
> h2. Description
> It was noticed that expunging instances with many vm snapshots took a look of time, as
hypervisor received as many tasks as vm snapshots instance had, apart from the delete vm task.
We propose a way to optimize this process for instances with vm snapshots by sending only
one delete task to hypervisor, which will delete vm and its snapshots
> h2. Use cases
> # deleteVMsnapohsot-> no changes to current behavior
> # destroyVM with expunge=false ->  no actions to VMsnaphsot is performed at the moment.
When VM cleanup thread is executed it will perform the same sequence as #3. If instance is
recovered before expunged by the cleanup thread it will remain intact with VMSnapshot chain
present
> # destroyVM with expunge=true:
> #*       Vmsnaphsot is  marked with removed timestamp and state = Expunging in DB
> #*       VM is deleted in HW



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

Mime
View raw message