jiajunwang commented on pull request #1072:
URL: https://github.com/apache/helix/pull/1072#issuecomment-641539518


   I'm not convinced that this is a good idea.
   
   For the non-Helix user, ZNRecord is not meaningful. How do you plan to 
document this config and let them configure 
ZK_SERIALIZER_**ZNRECORD**_WRITE_SIZE_LIMIT_BYTES for the general ZK usage?
   Secondly, we shall not block the write in the ZkClient if the write size is 
smaller than the jute.maxbuffer configuration. That is not very meaningful. On 
the other hand, if we have a large 
ZK_SERIALIZER_ZNRECORD_WRITE_SIZE_LIMIT_BYTES configured, but the 
jute.maxbuffer is small, what we shall do? The write will still fail.
   
   My suggestion,
   - The raw ZkClient shall check size based on jute.maxbuffer, and maybe we 
leave some buffer like the default limit is current doing.
   - If jute.maxbuffer is not configured, then we apply the default value 
according to the ZK document.


----------------------------------------------------------------
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.

For queries about this service, please contact Infrastructure at:
[email protected]



---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to