Project

General

Profile

How Can I Give Ceph a Try » History » Version 1

Jessica Mack, 06/03/2015 02:30 AM

1 1 Jessica Mack
h1. How Can I Give Ceph a Try
2 1 Jessica Mack
3 1 Jessica Mack
h3. Resolution
4 1 Jessica Mack
5 1 Jessica Mack
Follow our "Quick Start":http://docs.ceph.com/docs/master/start/ guides. They will get you up an running quickly without requiring deeper knowledge of Ceph. Our "Quick Start":http://docs.ceph.com/docs/master/start/ guides will also help you avoid a few issues related to limited deployments. If you choose to stray from the Quick Starts, there are a few things you need to know.
6 1 Jessica Mack
 
7 1 Jessica Mack
We recommend using at least two hosts, and a recent Linux kernel. In older kernels, Ceph can deadlock if you try to mount CephFS or RBD client services on the same host that runs your test Ceph cluster. This is not a Ceph-related issue. It’s related to memory pressure and needing to relieve free memory. Recent kernels with up-to-date %{color:grey}glibc% and %{color:grey}syncfs(2)% reduce this issue considerably. However, a memory pool large enough to handle incoming requests is the only thing that guarantees against the deadlock occuring. When you run Ceph clients on a Ceph cluster machine, loopback NFS can experience a similar problem related to buffer cache management in the kernel. You can avoid these scenarios entirely by using a separate client host, which is more realistic for deployment scenarios anyway.
8 1 Jessica Mack
 
9 1 Jessica Mack
We recommend using at least two OSDs with at least two replicas of the data. OSDs report other OSDs to the monitor, and also interact with other OSDs when replicating data. If you have only one OSD, a second OSD cannot check its heartbeat. Also, if an OSD expects another OSD to tell it which placement groups it should have, the lack of another OSD prevents this from occurring. So a placement group can remain stuck “stale” forever. These are not likely production issues.
10 1 Jessica Mack
11 1 Jessica Mack
Finally, "Quick Start":http://docs.ceph.com/docs/master/start/ guides are a way to get you up and running quickly. To build performant systems, you’ll need a drive for each OSD, and you will likely benefit by writing the OSD journal to a separate drive from the OSD data.