You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@ozone.apache.org by "Rajesh Balamohan (Jira)" <ji...@apache.org> on 2022/07/12 09:52:00 UTC

[jira] [Updated] (HDDS-6831) Ozone need not set pipeline/blockToken detail when doing listKey/lookup

     [ https://issues.apache.org/jira/browse/HDDS-6831?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Rajesh Balamohan updated HDDS-6831:
-----------------------------------
    Description: 
{color:#172b4d}This is based on a pure lookup test done based on "warp" client ([https://github.com/minio/warp).] Lookup test can be performed via "stat" option in warp.{color}

{color:#172b4d}Issue: {color}

{color:#172b4d}===={color}

{color:#172b4d}For lookup/list calls, ozone need not call SCM to get the containerIDs and refresh the pipelines (or set the pipeline details). {color}

{color:#172b4d}This will be pure overhead when someone just wants to list the keys. Also, it need not set "blockToken" information during lookup calls.{color}

{color:#172b4d}Example: During split generation, it is not necessary to know the pipeline details. FileStatus doesn't expose the block locations as well.{color}

 

 

!Screenshot 2022-06-06 at 12.20.32 PM.png!

  was:
{color:#172b4d}This is based on a pure lookup test done based on "warp" client ([https://github.com/minio/warp).] Lookup test can be performed via "stat" option in warp.{color}

{color:#172b4d}Issue: {color}

{color:#172b4d}===={color}

{color:#172b4d}For lookup/list calls, ozone need not call SCM to get the containerIDs and refresh the pipelines (or set the pipeline details). {color}

{color:#172b4d}This will be pure overhead when someone just wants to list the keys. {color}

{color:#172b4d}Example: During split generation, it is not necessary to know the pipeline details. FileStatus doesn't expose the block locations as well.{color}

 

 

!Screenshot 2022-06-06 at 12.20.32 PM.png!


> Ozone need not set pipeline/blockToken detail when doing listKey/lookup
> -----------------------------------------------------------------------
>
>                 Key: HDDS-6831
>                 URL: https://issues.apache.org/jira/browse/HDDS-6831
>             Project: Apache Ozone
>          Issue Type: Improvement
>            Reporter: Rajesh Balamohan
>            Priority: Major
>              Labels: performance
>         Attachments: Screenshot 2022-06-06 at 12.20.32 PM.png
>
>
> {color:#172b4d}This is based on a pure lookup test done based on "warp" client ([https://github.com/minio/warp).] Lookup test can be performed via "stat" option in warp.{color}
> {color:#172b4d}Issue: {color}
> {color:#172b4d}===={color}
> {color:#172b4d}For lookup/list calls, ozone need not call SCM to get the containerIDs and refresh the pipelines (or set the pipeline details). {color}
> {color:#172b4d}This will be pure overhead when someone just wants to list the keys. Also, it need not set "blockToken" information during lookup calls.{color}
> {color:#172b4d}Example: During split generation, it is not necessary to know the pipeline details. FileStatus doesn't expose the block locations as well.{color}
>  
>  
> !Screenshot 2022-06-06 at 12.20.32 PM.png!



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@ozone.apache.org
For additional commands, e-mail: issues-help@ozone.apache.org