You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@pinot.apache.org by "lrao-stripe (via GitHub)" <gi...@apache.org> on 2023/07/28 22:06:12 UTC

[GitHub] [pinot] lrao-stripe opened a new issue, #11211: Make segment upload to deep store misconfiguration fail fast

lrao-stripe opened a new issue, #11211:
URL: https://github.com/apache/pinot/issues/11211

   We ran into an issue on one of our tables to the following misconfiguration:
   
   - Table config had upload to deep store option set
   - Streaming servers did not have the segment s3 uri configuration set
   
   This resulted in failures to upload the segments to the deep store. However, the failures only manifested during/after segment commit time. 
   
   It'd be great to have a way to fail faster when there's a misconfiguration possibly:
   -  By adding a check when the consuming segment is created and checking if the s3 uri property is configured 
   -  By attempting to do this as a table level validation (this is potentially harder given that tableConfig + server config both need to be matched)


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@pinot.apache.org.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@pinot.apache.org
For additional commands, e-mail: commits-help@pinot.apache.org


[GitHub] [pinot] jadami10 commented on issue #11211: Make segment upload to deep store misconfiguration fail fast

Posted by "jadami10 (via GitHub)" <gi...@apache.org>.
jadami10 commented on issue #11211:
URL: https://github.com/apache/pinot/issues/11211#issuecomment-1660865136

   for another of our clusters that we set this table config but the servers didn't have https://github.com/apache/pinot/pull/10136, it was just ignored. so that also seems like acceptable behavior


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@pinot.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@pinot.apache.org
For additional commands, e-mail: commits-help@pinot.apache.org


[GitHub] [pinot] navina commented on issue #11211: Make segment upload to deep store misconfiguration fail fast

Posted by "navina (via GitHub)" <gi...@apache.org>.
navina commented on issue #11211:
URL: https://github.com/apache/pinot/issues/11211#issuecomment-1661534323

   @jadami10 Yeah. ideally, the segment store uri should never be null in the server. However, for whatever reason, we allow that to be null in the server instance. 
   Falling back to controller will overwhelm controller which is what we were trying to avoid with this feature. 


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@pinot.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@pinot.apache.org
For additional commands, e-mail: commits-help@pinot.apache.org


Re: [I] Make segment upload to deep store misconfiguration fail fast [pinot]

Posted by "suddendust (via GitHub)" <gi...@apache.org>.
suddendust commented on issue #11211:
URL: https://github.com/apache/pinot/issues/11211#issuecomment-2104236477

   >controller deep store uri to the ZK (or open a rest api), and server can use the same uri if it is not configured
   
   @Jackie-Jiang I am in favour of this. We shouldn't fail when we already have the information in the system. 


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@pinot.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@pinot.apache.org
For additional commands, e-mail: commits-help@pinot.apache.org


[GitHub] [pinot] Jackie-Jiang commented on issue #11211: Make segment upload to deep store misconfiguration fail fast

Posted by "Jackie-Jiang (via GitHub)" <gi...@apache.org>.
Jackie-Jiang commented on issue #11211:
URL: https://github.com/apache/pinot/issues/11211#issuecomment-1661521773

   @jadami10 You mean falling back to regular controller upload? That will be more smooth (no failure), but can potentially overwhelm controller because that is the reason why we implemented the feature.
   Alternatively, @navina also comes with an idea that we can write the controller deep store uri to the ZK (or open a rest api), and server can use the same uri if it is not configured


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@pinot.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@pinot.apache.org
For additional commands, e-mail: commits-help@pinot.apache.org