cassandra-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From aaron morton <aa...@thelastpickle.com>
Subject Re: idea drive layout - 4 drives + RAID question
Date Wed, 31 Oct 2012 21:27:54 GMT
Good question. 

The is a comment on the DS blog or docs somewhere that says on EC2 running the commit log
on the raid-0 ephemeral is preferred. I think the recommendation was specifically about how
the disks are setup on EC2.

While the commit log will be competing with logs and everything else on the OS volume, it
would be competing with C* reads, Memtable flushing, compacting and repairing on the data
volume. 

The only way to be sure is to test both setups. 

Cheers

-----------------
Aaron Morton
Freelance Developer
@aaronmorton
http://www.thelastpickle.com

On 31/10/2012, at 1:11 PM, Ran User <ranuser99@gmail.com> wrote:

> Is there a concern of a large falloff in commit log write performance (sequential) when
sharing 2 drives (RAID 1) with the OS (os and services writing their own logs, etc)?  Do you
expect the hit to be marginal?
> 
>  
> On Tue, Oct 30, 2012 at 7:58 PM, aaron morton <aaron@thelastpickle.com> wrote:
>> We also have 4-disk nodes, and we use the following layout:
>> 2 x OS + Commit in RAID 1
>> 2 x Data disk in RAID 0
> +1
> 
> You are replicating data at the application level and want the fastest possible IO performance
per node. 
> 
>>  You can already distribute the
>> individual Cassandra column families on different drives by just
>> setting up symlinks to the individual folders.
> There are some features coming in 1.2 that make using a JBOD setup easier. 
> 
> Cheers
> 
> -----------------
> Aaron Morton
> Freelance Developer
> @aaronmorton
> http://www.thelastpickle.com
> 
> On 30/10/2012, at 9:23 PM, Pieter Callewaert <pieter.callewaert@be-mobile.be> wrote:
> 
>> We also have 4-disk nodes, and we use the following layout:
>> 2 x OS + Commit in RAID 1
>> 2 x Data disk in RAID 0
>>  
>> This gives us the advantage we never have to reinstall the node when a drive crashes.
>>  
>> Kind regards,
>> Pieter
>>  
>>  
>> From: Ran User [mailto:ranuser99@gmail.com] 
>> Sent: dinsdag 30 oktober 2012 4:33
>> To: user@cassandra.apache.org
>> Subject: Re: idea drive layout - 4 drives + RAID question
>>  
>> Have you considered running RAID 10 for the data drives to improve MTBF?  
>>  
>> On one hand Cassandra is handling redundancy issues, on the other hand, reducing
the frequency of dealing with failed nodes is attractive if cheap (switching RAID levels to
10). 
>>  
>> We have no experience with software RAID (have always used hardware raid with BBU).
 I'm assuming software RAID 1 or 10 (the mirroring part) is inherently reliable (perhaps minus
some edge case).
>> 
>> On Tue, Oct 30, 2012 at 1:07 AM, Tupshin Harper <tupshin@tupshin.com> wrote:
>> I would generally recommend 1 drive for OS and commit log and 3 drive raid 0 for
data. The raid does give you good performance benefit, and it can be convenient to have the
OS on a side drive for configuration ease and better MTBF.
>> 
>> -Tupshin
>> 
>> On Oct 29, 2012 8:56 PM, "Ran User" <ranuser99@gmail.com> wrote:
>> I was hoping to achieve approx. 2x IO (write and read) performance via RAID 0 (by
accepting a higher MTBF).
>>  
>> Do believe the performance gains of RAID0 are much lower and/or are not worth it
vs the increased server failure rate?
>>  
>> From my understanding, RAID 10 would achieve the read performance benefits of RAID
0, but not the write benefits.  I'm also considering RAID 10 to maximize server IO performance.

>>  
>> Currently, we're working with 1 CF.
>>  
>>  
>> Thank you
>> 
>> On Mon, Oct 29, 2012 at 11:51 PM, Timmy Turner <timm.turn@gmail.com> wrote:
>> I'm not sure whether the raid 0 gets you anything other than headaches
>> should one of the drives fail. You can already distribute the
>> individual Cassandra column families on different drives by just
>> setting up symlinks to the individual folders.
>> 
>> 2012/10/30 Ran User <ranuser99@gmail.com>:
>> > For a server with 4 drive slots only, I'm thinking:
>> >
>> > either:
>> >
>> > - OS (1 drive)
>> > - Commit Log (1 drive)
>> > - Data (2 drives, software raid 0)
>> >
>> > vs
>> >
>> > - OS  + Data (3 drives, software raid 0)
>> > - Commit Log (1 drive)
>> >
>> > or something else?
>> >
>> > also, if I can spare the wasted storage, would RAID 10 for cassandra data
>> > improve read performance and have no effect on write performance?
>> >
>> > Thank you!
>>  
> 
> 


Mime
View raw message