[WBEL-users] WBEL, Clustering, GFS, and redundancy

[Sam Hillaire]

Some answers about GFS:

GFS is specifically a clustered file system, meaning that it is designed to have multiple hosts
connect to it and perform read and write operations to the disk device in an orderly manner.  The
file system itself (at least the quorum portion) allows for voting for which host gets to perform
the next write function to the disk device.

>From my experience (Oracle DBA), the OP is looking at some way to create and maintain a standby
database directly from the OS level, which GFS is not designed to do.  With my environment, I just
use Oracle functionality to create and maintain my standby databases.

For GFS links, I think this is the best starting link (http://www.redhat.com/software/rha/gfs/) at
RedHat's site.

(Just my opinion)  My skewed version of history:  a portion of Oracle's Maximum Availability
Architecture (RAC) relies on a cluster of database hosts connecting to a shared disk device (SAN,
NAS, SCSI, etc).  With database file sizes in the GBs, a file system had to be designed to allow
for concurrent writes to the same file but different blocks, allowing near optimal performance
from the multiple hosts.  Oracle devised a clustered filesystem (OCFS) but it had some severe
limitations as a filesystem.  Hence, I think Oracle went shopping for a better solution and turned
to RedHat for it, since Oracle was RedHat's launch partner for GFS, but don't quote me on that.

Sam Hillaire
samhillaire at yahoo.com


--- Benjamin Smith <lists at benjamindsmith.com> wrote:

> Does anybody here have experience with clustering on WBEL/RHES/AS? 
> 
> I have a web-based application hosted on WBEL/PostgreSQL and am looking for 
> the best way to provide even higher uptimes. (So far, 30 minutes of unplanned 
> downtime in over a year) 
> 
> I'm having some difficulty fully understanding what GFS is actually all about. 
> Is my understanding of GFS such that I could have two machines, each with 
> their own HDD, and have it set up with GFS as a sort of cross-machine RAID1? 
> 
> If I could save data like this, and then replicate the PostgreSQL database 
> with Slony, I *think* I'd have a good solution for redundant servers at the 
> colo with minimal downtime in the event of a failure of either system... Am I 
> right? 
> 
> Ideas/thoughts/pointers? Has anybody here set up stuff like that? 
> 
> Currently, we have two servers, a "live" server, and a hot backup which 
> synchronizes its filesystems periodically to the live server with rsync. This 
> gives a good "worst case scenario" recovery plan, but doesn't give much 
> support for lighter-grade failures. 
> 
> -Ben 
> 
> -- 
> "The best way to predict the future is to invent it."
> - XEROX PARC slogan, circa 1978
> _______________________________________________
> Whitebox-users mailing list
> Whitebox-users at beau.org
> http://beau.org/mailman/listinfo/whitebox-users
> 


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com