cloudstack-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Santhosh Kumar Edukulla (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (CLOUDSTACK-5957) [Automation]: Log Correlation and analysis for easier identification of issues
Date Fri, 21 Mar 2014 14:25:42 GMT

     [ https://issues.apache.org/jira/browse/CLOUDSTACK-5957?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Santhosh Kumar Edukulla updated CLOUDSTACK-5957:
------------------------------------------------

    Description: 
1. Provide a facility where, for every automation run ( regression or bvt  ), we should have
logs collected, stored at a common location and analyzed.
2. The logs include the following :
a) All logs collected through cloudbug tool,  test logs( both test run and test results etc
)
3. All logs once uploaded to a remote server post the automation run, will be allowed for
analysis part, where a given test case pass\fail is mapped to corresponding logs of management
server, api log or client logs etc.
4. The process should be automated. 

Phase1: 
1. Collect and store the logs at a given location post  every regression\BVTrun.
2. The logs are collected post automation run and uploaded,   we will  add a new job to CI
facility and this job will be added as downstream project for every regression or bvt run
post the run.
3. The job is run takes few configurable parameters viz., management server, path for logs,remote
location,path etc and stores for every release as
//<release version>/xen/Build_run_logs.zip
4. The remote copied folder ( compressed ) post the copy, will have test_logs, cs_logs etc
along with metadata like last commit-id, build information,cs version information and any
other data required.

Phase2:
1. provide few simple interfaces to correlate logs from both test and product like management
server logs and provide a mapping for easy analysys and extract this information to user.
2. Provide a UI where user can search for a given build, given test failure, and show what
happens during that time on management server or agent etc. for possible and easy reconciliation
of logs.
3. Extract if possible for a given jobid in test run log with jobid in management server log
etc.

  was:
1. Provide a facility where for a given regression or bvt run,where we should have logs collected
and stored at a given place.
2. The logs include the following :
 Management server logs, test logs( run,results etc )
3. The process should be automated. 

Phase1: 
1. Collect and store the logs at a given location post  every regression\BVTrun.
2.  We will  add a new job to CI facility and this job will be added as downstream project
for every regression or bvt run post the run.
3. The job takes few configurable parameters viz., management server, path for logs,remote
location,path etc.
4. Using the configuration, the script copies the files to the relevant location.
5. Initially, a vm with few specific setup will be used to store the logs. 
6. The logs will  be remotely copied in a compressed format, by specific build and commit
id.
7. The remote copied folder ( compressed ) post the copy, will have test_logs, cs_logs etc
along with metadata like last commit-id, build information,cs version information and any
other data required.
8. The folder name can be used along with metadata to identify the relevant logs.

Phase2:
1. provide few interfaces( through web ) to analyze and extract this information to user.
2. Provide a UI if possible for easy reconciliation of logs.


> [Automation]: Log Correlation and analysis for easier identification of issues
> ------------------------------------------------------------------------------
>
>                 Key: CLOUDSTACK-5957
>                 URL: https://issues.apache.org/jira/browse/CLOUDSTACK-5957
>             Project: CloudStack
>          Issue Type: Improvement
>      Security Level: Public(Anyone can view this level - this is the default.) 
>          Components: Automation, marvin
>            Reporter: Santhosh Kumar Edukulla
>            Assignee: Santhosh Kumar Edukulla
>
> 1. Provide a facility where, for every automation run ( regression or bvt  ), we should
have logs collected, stored at a common location and analyzed.
> 2. The logs include the following :
> a) All logs collected through cloudbug tool,  test logs( both test run and test results
etc )
> 3. All logs once uploaded to a remote server post the automation run, will be allowed
for analysis part, where a given test case pass\fail is mapped to corresponding logs of management
server, api log or client logs etc.
> 4. The process should be automated. 
> Phase1: 
> 1. Collect and store the logs at a given location post  every regression\BVTrun.
> 2. The logs are collected post automation run and uploaded,   we will  add a new job
to CI facility and this job will be added as downstream project for every regression or bvt
run post the run.
> 3. The job is run takes few configurable parameters viz., management server, path for
logs,remote location,path etc and stores for every release as
> //<release version>/xen/Build_run_logs.zip
> 4. The remote copied folder ( compressed ) post the copy, will have test_logs, cs_logs
etc along with metadata like last commit-id, build information,cs version information and
any other data required.
> Phase2:
> 1. provide few simple interfaces to correlate logs from both test and product like management
server logs and provide a mapping for easy analysys and extract this information to user.
> 2. Provide a UI where user can search for a given build, given test failure, and show
what happens during that time on management server or agent etc. for possible and easy reconciliation
of logs.
> 3. Extract if possible for a given jobid in test run log with jobid in management server
log etc.



--
This message was sent by Atlassian JIRA
(v6.2#6252)

Mime
View raw message