Matt Lawrence wrote: > On Tue, 27 Apr 2010, Richard Chycoski wrote: > > >> If might help if you could be less vague on a few points - what kinds of >> things are you sharing, and what were the challenges with NFS? >> > > We are having failures that are causing application outages. > > >> If you hadn't mentioned NFS as an issue, it would be one of my recommended >> choices. Also, what NFS implementation(s?) have you been using? Is the >> problem with NFS in general, or with specific implementations? >> > > Older RHEL HA configuration. Like many HA implementations, it is not > fault tolerant, when something fails, the client tends to know it. The > applications here are not well behavied in those cases. > > >> It helps to know what you are sharing - i.e., if someone tells me that they >> have this product that uses Btrieve databases as its underlying mechanism >> (and yes, I have to work with a product that does exactly this ;-), I would >> say 'run away from NFS immediately' - the getattrs calls will kill your app. >> Other file sharing mechanisms have different profiles and issues. >> > > Mostly large and small text files. Generally not accessed by more than > one client at a time. > > Performance is not a big issue. Total storage is in the TB or 10s of TB > range. Not huge by today's standards. > > -- Matt > Ah - then a 'better' NFS HA implementation might be a possible answer. Next - do you have the budget/learning-curve-bandwidth to look at something like NetApp?
- Richard _______________________________________________ Tech mailing list [email protected] http://lopsa.org/cgi-bin/mailman/listinfo/tech This list provided by the League of Professional System Administrators http://lopsa.org/
