[CentOS] HA with CentOS

Mon May 14 20:14:05 UTC 2007
Ruslan Sivak <rsivak at istandfor.com>

Scott McClanahan wrote:
> On Mon, 2007-05-14 at 10:55 -0400, Ruslan Sivak wrote:
>   
>> Steve Huff wrote:
>>     
>>> On May 14, 2007, at 10:25 AM, Ruslan Sivak wrote:
>>>
>>>       
>>>> Steve Huff wrote:
>>>>
>>>> If you set up a third box to be the shared storage, doesn't that now 
>>>> become the single point of failure?
>>>>         
>>> Short answer: maybe. :)
>>>
>>> Longer answer: If you set up your shared storage according to 
>>> upstream's guidelines, as described in the documentation 
>>> (http://mirror.centos.org/centos/4/docs/html/rh-cs-en-4/ch-hardware.html#TB-HARDWARE-NOSPOF), 
>>> then you provide at least two channels of communication between each 
>>> component in the cluster.  In addition, you choose a platform for 
>>> shared storage that provides some redundancy of its own, whether it's 
>>> multi-controller HW RAID, or multiple storage nodes on a SAN, or what 
>>> have you.
>>>
>>> CS/GFS operates under the assumption that your shared storage is 
>>> fault-tolerant; its job is to make your services fault-tolerant.  Is 
>>> the recommended "no single point of failure" configuration proof 
>>> against your data center burning down, or against a madman with an 
>>> axe?  Unlikely.  Will it allow you to host services in a way that is 
>>> considerably more robust and flexible than hosting them on a single 
>>> box?  Yes.
>>>
>>> -Steve
>>>
>>>       
>> I am currently running a redundant environment on windows by having 2 
>> boxes with apache and having the data (images) be synced over 
>> automatically between servers using FRS (File Replication Service).   
>> This works well most of the time, except for when it breaks, at which 
>> point I need to resync the two servers, which usually takes days. 
>>
>> I would like to set up something similar using linux.  I don't have the 
>> budget for a SAN/NAS, and even having a third server as storage would 
>> probably not be worth it, although we can possibly go with this.  The 
>> problem, is that it would be a single point of failure. 
>>
>> Is there some service/filesystem in Linux that allows for the automatic 
>> replication of files to make a fault tolerant environment possible with 
>> only 2 servers?  Basically whenever there is an update of a file on a 
>> certain file system (certain folder), the file gets synced over to 
>> another system. 
>>
>> Russ
>>
>>
>> _______________________________________________
>> CentOS mailing list
>> CentOS at centos.org
>> http://lists.centos.org/mailman/listinfo/centos
>>
>>     
>
> DRBD and Heartbeat seem pretty solid together for cheap affective high
> availability.  We've been using them for our production FTP servers
> which handle hundreds of thousands transactions a day both
> uploading/downloading.  We fail over between the two every 6 months and
> haven't had any problems on CentOS 4.3, they've actually been up for
> several hundred days now.  There is actually a yumgroup named
> drbd-heartbeat in the CentOS extras repository but I don't see that it
> is available in CentOS 5.0.  Does anyone know if these packages will be
> available in any of the CentOS 5.0 yum repositories?
>
> _______________________________________________
>   

Looks interesting.  I will have to try them out once they're in the 
stable repo.  Looks as of DRBD-8.0.0, you can use it together with GFS 
and run both nodes as primary.  Would heartbeat still be needed?  Can 
heartbeat work with VM boxes?

Russ