[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: [Linux-cluster] Can i use 2 nodes from a cluster?



On 23/10/13 01:33 PM, merc1984 f-m fm wrote:
On Wed, Oct 23, 2013, at 11:13, Elvir Kuric wrote:
If not what are LVS' components?
More here http://www.linuxvirtualserver.org/
Thanks, but I've already read through the whole site and could not find
a clue of LVS components.

In other words, I want to implement LVS in Fedora, without using all the
disjointed, one-off, disparate, individual apps.



It's been a while since I looked at it but the LVS site was never organized in a very friendly manner.
Way back in 2008 or so we setup a pair of LVS servers on CentOS4 using Ultramonkey 3.  We are still using those servers now with our local and remote data center hosting although we are soon going to finish phasing those out as we move to Amazon EC2. The Ultramonkey site (http://ultramonkey.org) was extremely helpful in giving example configurations and concepts when we first set this up.  The bigger problem actually turned out to be file replication.  We attempted to use DRBD+GFS but it was buggy locally and inept across WAN links, then moved to asynchronous rsync (which is simply a horrible idea with any significant amount of files/data), then asynchronous csync2 which works decently well, and now finally gluster which is for the most part a dream.
What happens in your configuration when (not if) your centralized storage goes down? Is that designed to be HA?

You may also want to check out http://linux-ha.org


----
dcmbrown shaw ca


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]