phoenix-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Samarth Jain (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (PHOENIX-3811) Ignore index write failure and let client deal with failure
Date Wed, 26 Apr 2017 18:59:04 GMT

    [ https://issues.apache.org/jira/browse/PHOENIX-3811?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15985357#comment-15985357
] 

Samarth Jain commented on PHOENIX-3811:
---------------------------------------

In case of a failure, the client would need to make sure that the retry of writes happen in
the same order as before. Another option would be introduce a new index state, something like
ACTIVE_REBUILD_NEEDED, which would tell our query optimizer to still use the index for querying
purposes. The index rebuild path will check for this state and continue with the rebuild of
the index to backfill the mutations that were missed since the index was disabled.

> Ignore index write failure and let client deal with failure
> -----------------------------------------------------------
>
>                 Key: PHOENIX-3811
>                 URL: https://issues.apache.org/jira/browse/PHOENIX-3811
>             Project: Phoenix
>          Issue Type: Bug
>            Reporter: James Taylor
>         Attachments: PHOENIX-3811-wip1.patch, PHOENIX-3811-wip2.patch
>
>
> We should provide a way to configure the system so that the server takes no specific
action when an index write fails. Since we always throw the write failure back to the client,
the client can often deal with failures more easily than the server since they have the batch
of mutations in memory. Often times, allowing access to an index that may be one batch behind
the data table is better than disabling it given the negative performance that will occur
while the index cannot be written to.



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

Mime
View raw message