cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Hervé Toulan (JIRA) <>
Subject [jira] [Commented] (CASSANDRA-4782) Commitlog not replayed after restart
Date Thu, 29 Oct 2015 08:58:27 GMT


Hervé Toulan commented on CASSANDRA-4782:

Hi Robert,

1) Thank you, I did know that, I almost never encountered issue with Cassandra (even with
the 1.1.0)
2) That's what I thought. I plan to upgrade to 1.2.9.
3) Now, I lose data on all 1.1.0 platforms if I insert data and reboot servers (ring composed
by 2 servers)... 
I can't believe we never saw that before. 
I am not sure I understand how Cassandra works with System.nanoTime() and replay position
but couold we imagine current timestamp reaches a value for wich this bug is now reproducibke

Anyway you're right, and upgrade is mandatory.


> Commitlog not replayed after restart
> ------------------------------------
>                 Key: CASSANDRA-4782
>                 URL:
>             Project: Cassandra
>          Issue Type: Bug
>    Affects Versions: 1.1.0
>            Reporter: Fabien Rousseau
>            Assignee: Jonathan Ellis
>            Priority: Critical
>             Fix For: 1.1.6
>         Attachments: 4782.txt
> It seems that there are two corner cases where commitlog is not replayed after a restart
>  - After a reboot of a server + restart of cassandra (1.1.0 to 1.1.4)
>  - After doing an upgrade from cassandra 1.1.X to cassandra 1.1.5
> This is due to the fact that the commitlog segment id should always be an  incrementing
number (see this condition :
> But this assertion can be broken :
> In the first case, it is generated by System.nanoTime() but it seems that System.nanoTime()
is using the boot time as the base/reference (at least on java6 & linux), thus after a
reboot, System.nanoTime() can return a lower number than before the reboot (and the javadoc
says the reference is a relative point in time...)
> In the second case, this was introduced by #4601 (which changes System.nanoTime() by
System.currentTimeMillis() thus people starting with 1.1.5 are safe)
> This could explain the following tickets : #4741 and #4481

This message was sent by Atlassian JIRA

View raw message