[jira] [Commented] (VCL-1034) Lab.pm does not handle preload reservations correctly

["ASF subversion and git services (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/VCL-1034?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15991023#comment-15991023 ] 

ASF subversion and git services commented on VCL-1034:
------------------------------------------------------

Commit 1793369 from arkurth@apache.org in branch 'vcl/trunk'
[ https://svn.apache.org/r1793369 ]

VCL-1034
Updated Lab.pm::load to not call exit. This was causing the state to get stuck in 'pending' for preload requests. Cleaned up load and node_status subroutines. node_status now returns 'READY' if the computer is responding to SSH, and 'RELOAD' if it is not. The load() subroutine also performs an SSH check. If it is not responding, node_status forces load to be called, which will return false, causing the reservation to fail before the user is presented with the connect button.

> Lab.pm does not handle preload reservations correctly
> -----------------------------------------------------
>
>                 Key: VCL-1034
>                 URL: https://issues.apache.org/jira/browse/VCL-1034
>             Project: VCL
>          Issue Type: Bug
>          Components: vcld (backend)
>    Affects Versions: 2.4.2
>            Reporter: Andy Kurth
>             Fix For: 2.5
>
>
> The reservation gets processed ahead of time:
> {noformat}
> 2017-04-19 19:54:36|14129|3149948|3258996|new|Lab.pm:node_status|287|VCL client daemon is running
> 2017-04-19 19:54:36|14129|3149948|3258996|new|Lab.pm:node_status|312|returning node status hash reference with {status}=READY
> 2017-04-19 19:54:36|14129|3149948|3258996|new|new.pm:reload_image|518|node status is HASH(0x3e752c0), engr-vcl-l-004 will be reloaded
> 2017-04-19 19:54:36|14129|3149948|3258996|new|utils.pm:insertloadlog|2762|inserted 'loadimageblade' computerloadlog entry
> 2017-04-19 19:54:36|14129|3149948|3258996|new|new.pm:reload_image|565|unable to check if image exists, does_image_exist() not implemented by VCL::Module::Provisioning::Lab
> 2017-04-19 19:54:36|14129|3149948|3258996|new|utils.pm:update_computer_state|1462|computer 2869 state updated to: reloading
> 2017-04-19 19:54:36|14129|3149948|3258996|new|new.pm:reload_image|570|computer engr-vcl-l-004 state set to reloading
> 2017-04-19 19:54:36|14129|3149948|3258996|new|utils.pm:insertloadlog|2762|inserted 'info' computerloadlog entry
> 2017-04-19 19:54:36|14129|3149948|3258996|new|new.pm:reload_image|580|calling VCL::Module::Provisioning::Lab->load() subroutine
> 2017-04-19 19:54:36|14129|3149948|3258996|new|utils.pm:insertloadlog|2762|inserted 'info' computerloadlog entry
> 2017-04-19 19:54:36|14129|3149948|3258996|new|DataStructure.pm:is_parent_reservation|1140|this is the parent reservation
> 2017-04-19 19:54:36|14129|3149948|3258996|new|Lab.pm:load|97|computer id: 2869, is parent reservation: 1
> 2017-04-19 19:54:36|14129|3149948|3258996|new|Lab.pm:load|102|check_time result is preload, nothing needs to be done for lab preloads
> 2017-04-19 19:54:36|14129|3149948|3258996|new|utils.pm:insertloadlog|2762|inserted 'info' computerloadlog entry
> 2017-04-19 19:54:36|14129|3149948|3258996|new|utils.pm:update_preload_flag|1321|preload flag updated for request_id 3149948
> 2017-04-19 19:54:36|14129|3149948|3258996|new|Lab.pm:load|110|parent reservation: updated preload flag to 1
> 2017-04-19 19:54:36|14129|3149948|3258996|new|utils.pm:insertloadlog|2762|inserted 'info' computerloadlog entry
> 2017-04-19 19:54:36|14129|3149948|3258996|new|Lab.pm:load|121|preload lab reservation process exiting
> 2017-04-19 19:54:36|14129|3149948|3258996|new|State.pm:DESTROY|1045|VCL::new destructor called, address: 2ebbc08
> 2017-04-19 19:54:36|14129|3149948|3258996|new|State.pm:state_exit|863|beginning state module exit tasks
> |14129|3149948|3258996|new|State.pm:state_exit|863| request state argument: <not specified>
> |14129|3149948|3258996|new|State.pm:state_exit|863| computer state argument: <not specified>
> |14129|3149948|3258996|new|State.pm:state_exit|863| log ending argument: <not specified>
> 2017-04-19 19:54:36|14129|3149948|3258996|new|DataStructure.pm:is_parent_reservation|1140|this is the parent reservation
> 2017-04-19 19:54:36|14129|3149948|3258996|new|utils.pm:delete_computerloadlog_reservation|5848|deleted rows from computerloadlog matching loadstate regex '(begin)' for reservation IDs: 3258996
> 2017-04-19 19:54:36|14129|3149948|3258996|new|utils.pm:insertloadlog|2762|inserted 'exited' computerloadlog entry
> 2017-04-19 19:54:36|14129|3149948|3258996|new|State.pm:state_exit|1014|calling subroutine: VCL::Module::State::DESTROY, skipping call to exit
> 2017-04-19 19:54:36|14129|3149948|3258996|new|Module.pm:DESTROY|2446|destroying VCL::new object, address: 2ebbc08
> 2017-04-19 19:54:36|14129|3149948|3258996|new|State.pm:DESTROY|1098|VCL::new process duration: 3 seconds
> 2017-04-19 19:54:37|14129|3149948|3258996|new|Module.pm:DESTROY|2446|destroying VCL::Module::Provisioning::Lab object, address: 3e4ec88
> 2017-04-19 19:54:37|14129|3149948|3258996|new|Module.pm:DESTROY|2446|destroying VCL::Module::OS::Linux::UnixLab object, address: 3e29600
> 2017-04-19 19:54:37|14129|3149948|3258996|new|Module.pm:DESTROY|2446|destroying VCL::Module::OS::Linux::ManagementNode object, address: 3e4eac0
> {noformat}
> After this process exits, the user is not presented with a Connect button.  vcld.log does not reprocess the request then the start time is reached.



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