On Sat, 13 Oct 2012, Ricardo Auada wrote: > Hi, I?m just starting to look at CEPH and some questions crossed my mind. I > wonder if someone can help me with them: > > 1- Does Ceph use any kind of cache to improve speed access to data in OSD?s ?
The ceph file system and RBD both have caches. The low-level librados does not, and radosgw (RESTful S3/Swift access) caches only certain metadata, but not object data. Caches can be easily layered on top of these. > 2- Because Ceph is usually used for big amounts of data, how can we backup / > snapshot this data ? How fast can we recover this big backup ? Or we wouldn?t > need a backup at all, due to the redundance features ? Here I mean only > disaster > recover backup. This is a big question for any large storage system. The goal is that for most applications, there would be no backup, since Ceph is masking individual component failure. Ultimately this is a question for you, based on the data you are storing and your level of trust. > 3- I see Ceph has an S3 API. Also, we are studying some S3 filesystems and > S3backer seems to be a very nice one. What would be the differences in using > one > or the other ? If the goal is a file system, I would use RBD or the ceph fs directly, not layer it on top of radosgw; you'll get a cleaner, simpler stack and better performance. sage -- To unsubscribe from this list: send the line "unsubscribe ceph-devel" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
