Mahadev konar updated ZOOKEEPER-596:

    Attachment: ZOOKEEPER-596.patch

this patch adds a test case for using memory based zkdatabase most of the time. 
The test checks to see that a server who has corrupted database cannot join the 
cluster. I am still thinking if we should just start with a empty database in 
such a case or just shutdown and let the admin figure it out. This way if the 
disk is corrupt, an admin can take care of it. For now, I have left the 
quorumpeer to exit if it finds its database is corrupt on and upto the admin to 
sanitize the database (by just deleting the database and starting all new on 
that node).

this patch includes :

> The last logged zxid calculated by zookeeper servers could cause problems in 
> leader election if data gets corrupted.
> --------------------------------------------------------------------------------------------------------------------
>                 Key: ZOOKEEPER-596
>                 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-596
>             Project: Zookeeper
>          Issue Type: Bug
>    Affects Versions: 3.2.1
>            Reporter: Mahadev konar
>            Assignee: Mahadev konar
>             Fix For: 3.3.0
>         Attachments: ZOOKEEPER-596.patch, ZOOKEEPER-596.patch, 
> ZOOKEEPER-596.patch
> It is possible that the last loggged zxid as reported by all the servers 
> during leader election is not the last zxid that the server can upload data 
> to. It is very much possible that some transaction or snapshot gets corrupted 
> and the servers actually do not have valid data till last logged zxid. We 
> need to make sure that what the servers report as there last logged zxid, 
> they are able to load data till that zxid.

This message is automatically generated by JIRA.
You can reply to this email to add a comment to the issue online.

Reply via email to