incubator-cassandra-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jonathan Ellis <jbel...@gmail.com>
Subject Re: problem about bootstrapping when used in huge node
Date Tue, 23 Feb 2010 16:33:30 GMT
For Michael, I think raid5 is not a bad choice:
 - if you're going to have multiple TB of data, using JBOD gives a
pretty harsh limit to your compaction/anticompaction scenarios
 - he is concerned about his data set size, so so raid 1 or 10 "wastes" space
 - raid0 is potentially painful since you'd have to transfer a lot of
data to repair every time a single disk dies

-Jonathan

On Tue, Feb 23, 2010 at 10:26 AM, Brandon Williams <driftx@gmail.com> wrote:
> On Tue, Feb 23, 2010 at 7:31 AM, Jonathan Ellis <jbellis@gmail.com> wrote:
>>
>> > (2)     How to use node has 12 1TB disk??
>> You should use a better filesystem than ext3. :)  We use xfs at rackspace.
>
> Also, don't use RAID5.  Let Cassandra's replication handle disk failure
> scenarios instead, and supply multiple DataFileDirectory directives to
> unique mount points.  If you must use RAID, RAID0, 1, or 10 would be better.
> -Brandon

Mime
View raw message