sounds like a job for couchdb to me.
masterless replication, you do need to figure out some kind of topology
though, you don't want everything replicating with everything as that
would be a huge number of connections. Maybe some kind of broadcast
system could be used to have nodes advertising that they have an
interesting update that is worth replicating with. Alternatively a
random topology might work. Each node connecting at random to another
node every x minutes, tune to meet your requirements.
Alan.
Chris Simmonds wrote:
> Adrian Bridgett wrote:
>
>> On Tue, Sep 15, 2009 at 17:15:18 +0100 (+0100), Chris Simmonds wrote:
>>
>>> One option I have considered is using, say, MySQL with one master node
>>> replicating to all the others and some mechanism to elect a new master
>>> if the original went down. But, that sounds messy. There must be a
>>> neater solution?
>>>
>> MySQL + LinuxHA + DRBD is the typical solution. You could use MySQL
>> with replication and failover (or even multi-master if you are brave).
>>
>> A proper clustered filesystem (e.g. OCFS2 ontop of DRBD8) maybe better.
>>
>> However be aware that all of these aren't simple configs and you
>> really want to do thorough testing beforehand - I've seen some setups
>> be incredibly brittle - so any theoretical improvement in uptime may
>> in practice become a major drop.
>>
>> For example when I was looking at GFS last year:
>> http://www.smop.co.uk/blog/index.php/2008/02/11/gfs-goodgrief-wheres-the-documentation-file-system/
>>
>> Adrian
>>
>
> Hi and thanks to everyone who replied. I'm busy researching some
> possibilities at the moment. However, just to clarify, the issue is high
> availability among the 50 or so nodes so that any node can go down and
> come back up again without impacting any of the others. There are no
> servers as such, each node is potentially both a server and a client so
> that if using a traditional database there needs to be some mechanism to
> elect a master server if the current master goes down. The trickiest
> case is if the network gets disrupted such that you get two separate
> segments for a while - each with their own master - which then get
> joined together again. On the other hand the amount of data is quite
> small - maybe a thousand rows in SQL terms. Actually it doesn't have to
> be SQL at all, I'm just using that as an example.
>
> Anyhow, thanks again. I'll post again when I have had time to mull over
> the various suggestions and tried a few things out.
>
> Bye for now,
> Chris.
>
>
--
Alan Bell
The Open Learning Centre
Web:
www.theopenlearningcentre.com
Mob: +44 (0)7736 778908
Tel: +44 (0)8444 3576000
The Open Learning Centre is a trading name of Bell Lord Ltd,
a company registered in England and Wales #05868943.
VAT Registration #GB 901 4715 55
