[
https://issues.apache.org/jira/browse/PHOENIX-542?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15287909#comment-15287909
]
James Taylor commented on PHOENIX-542:
--------------------------------------
Yes, for both UPSERT SELECT and DELETE when run with auto commit true, or when
run over transactional tables. In these cases, we batch the mutations up based
on the number of rows specified by {{phoenix.mutate.batchSize}} or
{{UpsertBatchSize}} value in the URL. It'd be better if we had a config
parameter to specify the number of *bytes* instead, as row sizes vary widely
based on the schema and values. We can easily determine the byte size based on
the mutations we're sending over. We also don't include the size of immutable
index data which would be sent over with the table data.
> Batch size calculations should take into account secondary indexes
> ------------------------------------------------------------------
>
> Key: PHOENIX-542
> URL: https://issues.apache.org/jira/browse/PHOENIX-542
> Project: Phoenix
> Issue Type: Task
> Affects Versions: 3.0-Release
> Reporter: mujtaba
> Labels: enhancement
>
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
