zookeeper-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Alexander Shraer (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (ZOOKEEPER-3408) Improve information about risks for forceSync config option
Date Mon, 03 Jun 2019 06:22:00 GMT

    [ https://issues.apache.org/jira/browse/ZOOKEEPER-3408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16854253#comment-16854253
] 

Alexander Shraer commented on ZOOKEEPER-3408:
---------------------------------------------

There were some discussions about this on the mailing list, eg., [http://zookeeper-user.578899.n2.nabble.com/forceSync-no-td7577568.html]

IMHO a single reboot and leader re-election could lead to data loss if data isn't synced to
disk - imagine 5 servers. A transaction committed to 3 out of 5, and one of those 3 reboots.
Now 3 out of 5 servers don't have that transaction and could cause that transaction to be
truncated from the logs of the other 2 servers. Maybe someone can correct me if I'm wrong
here.

> Improve information about risks for forceSync config option
> -----------------------------------------------------------
>
>                 Key: ZOOKEEPER-3408
>                 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-3408
>             Project: ZooKeeper
>          Issue Type: Improvement
>          Components: documentation
>            Reporter: Dmitry Konstantinov
>            Priority: Minor
>
> [https://zookeeper.apache.org/doc/r3.5.5/zookeeperAdmin.html#Unsafe+Options]
> {quote}
> The following options can be useful, but be careful when you use them. The risk of each
is explained along with the explanation of what the variable does.
> {quote}
> {quote}
> _forceSync_ : (Java system property: *zookeeper.forceSync*) Requires updates to be
synced to media of the transaction log before finishing processing the update. If this option
is set to no, ZooKeeper will not require updates to be synced to the media.
> {quote}
> The risks for this option are not very clear. Does it only the risk of loss some recent
committed transactions if all Zookeeper instances from an ensemble are crashed and restarted
at almost the same time or some other problems are also possible? 
>  



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Mime
View raw message