This is our current configuration (with 3.0.1). We are running up against an upper limit with cache read-busy failures when we have many simultaneous requests for the same object that I was hoping that using a raw disk might fix (or give us a higher ceiling).
Are there other things to consider when getting read-busy failures? From: Leif Hedstrom [mailto:[email protected]] Sent: Friday, June 22, 2012 8:49 PM To: [email protected] Cc: [email protected] Subject: Re: Cache hardware and config questions On Jun 22, 2012, at 3:53 PM, "Earle, Erik" <[email protected]<mailto:[email protected]>> wrote: For actual hardware, our in-house datacenter only supports certain configurations of hardware, none of which include being able to have a dedicated physical disk as a raw device for cache configuration. My two options are: 1) A logical partition of the main disk to use as the raw disk 2) Move to VMs to get the raw disk on a separate virtual physical device (which may end up on the same actual physical device) Why not just put the cache on the file system then? It will be less bad than going through a hyper visor... -- Leif Are there any observations/recommendations regarding either of my options?
